Issuu

SOUND ON SOUND THE WORLD’S BEST RECORDING TECHNOLOGY MAGAZINE

THE WORLD’S BEST RECORDING TECHNOLOGY MAGAZINE

CELEBRATING 40+ YEARS 1985 — 2026 TM

MUSIC PRODUCTION TECHNIQUES / INDEPENDENT IN-DEPTH PRODUCT TESTS / ENGINEER & PRODUCER INTERVIEWS / LIVE SOUND

SOFTUBE FLOW STUDIO THE CONTROLLER FOR THE BIGGER PICTURE

JULY 2026

www.soundonsound.com

Oeksound Soothe 3

Resonance control becomes creative

ASM Leviasynth VOLUME 41 • ISSUE 9

“Immense, powerful, deep, confusing and astounding”

How AI music works

And how to protect your own music

30+ IN-DEPTH REVIEWS INCLUDING: MQA / BALCONY LEFT / BAT / ZERO-G / IZOTOPE / VINTAGE TOOLS / NI / SOURCE ELEMENTS / AKAI / SHURE / JOSEPHSON / MACKIE / BAREFOOT / TEMECULA DSP / SONIBLE / UB DSP / CHERRY AUDIO / GINTRONIC / HAMMOCK HILL

TECHNIQUE: MAKING YOUR MIXES WORK ON MOBILE DEVICES

ARTURIA MEGA-BUNDLE WORTH €1330

www.soundonsound.com

July 2026 £7.99

Small Box.

Big Bang. Analog. Ultra-portable. Pro-grade. Modular. Balanced stereo I/O - CV Control MIDI sync. +26dBU headroom Optional class compliant USB audio interface - 6 IN, 4 OUT Studio, stage and everywhere in between

SEE US AT

rodec.com AVAILABLE PRODUCTS

MiniStyler

SideStepper

M-101

M-102

M-108

M-104

M-501

MX Modular

Sherman's cult ﬁlter, reborn.

Analog gater + step sequencer.

Dual discrete phono preamp.

Classic 3-band musical EQ.

Discrete stereo line preamp.

Dynamic ﬁlter with envelope.

500-series mic/line pre + EQ.

The full 500 / 100-series console.

Analog Love from Belgium Distributed in the UK by Source Distribution - www.sourcedistribution.co.uk

@rodec_audio

LE ADER

MACRO ECONOMICS A

s music production environments become more complex, the challenge of making those environments accessible grows. Learning a modern DAW is very different from operating a four-track cassette recorder. Luckily, there are now vastly more resources available to learners. Where once we had printed manuals, we have YouTube tutorials, Discord servers, Subreddits and Facebook groups. We can even ask an AI chatbot. And developers themselves are putting ever greater resources into smoothing out learning curves. From a design perspective, macros are an excellent tool for this purpose. An experienced synth programmer might simultaneously adjust pulse width, filter cutoff modulation depth and resonance in order to achieve a specific effect. A skilled user interface designer can assign all of those to a single physical control. The result: immediate hands-on gratification for all users, and an opportunity for the inexperienced to get results without needing to understand the underlying structure. This approach is not new, but this month’s cover product takes it to another level. Literally so, in fact. The Omni knob on Softube’s Flow Studio is a macro to rule all macros, allowing users to move seamlessly between multiple layers of interaction. When you’re diving in for the first time, you can get close to what you want with a single hand movement. When it’s time to apply the finishing

ALLIA BUSINESS CENTRE KING’S HEDGES ROAD CAMBRIDGE CB4 2HY T +44 (0)1223 851658 sos@soundonsound.com www.soundonsound.com

touches, you can dive in and tweak individual parameters. New users need not be intimidated; old hands need not feel patronised. The kicker, perhaps, is that this is possible only because Flow Studio is a closed system. This level of integration is possible only where a single developer has complete control, and that makes some people uneasy. It’s surely a price worth paying in this case, but I do wonder whether it’s part of wider trend. Forty years ago, synth manufacturers could agree on what MIDI needed to do. Today, the idea that DAW developers might commit to a universal session interchange format seems far-fetched. The AES72 ‘standard’ for analogue audio over Ethernet isn’t really a standard at all. HUI and MCU are dated and limited, but EuCon remains proprietary. And how many audio-over-IP formats, or immersive audio configurations, do we really need? In today’s world, successful formats and platforms are valuable assets. Companies like Dolby, Audinate and XMOS have thrived not by selling things to consumers, but by persuading manufacturers to license their technology. In fact, Atmos and Dante have succeeded despite the existence of open alternatives. Does this mean that market forces are proving their superiority? Is it even possible that adhering to open standards can be a brake on innovation? I don’t know, but I don’t think we should automatically view proprietary systems as bad.

ADV ER T ISIN G

admin@soundonsound.com

sos.feedback@soundonsound.com

adsales@soundonsound.com

Managing Director/Chairman Ian Gilby

Editorial Director Dave Lockwood

UK Media Sales Manager Chris Cope

Editorial Director Dave Lockwood

Executive Editor Paul White

Marketing Director Paul Gilby

Editor In Chief Sam Inglis

Finance Manager Keith Werthmann

Technical Editor Hugh Robjohns Reviews Editor David Glasper Reviews Editor Matt Houghton

P R ODUC T I ON

Production Editor Chris Korff

graphics@soundonsound.com

News Editor Luke Wood

Nick Humbert

Printing Warners Midlands plc Newstrade Distribution Warners Group Distribution Ltd, The Maltings, Manor Lane, Bourne, Lincolnshire PE10 9PH, UK.

support@soundonsound.com

ISSN 0951‑6816

Digital Media Director Paul Gilby

A Member of the SOS Publications Group

www.soundonsound.com/subscribe Circulation Manager Luci Harper

July 2026 / www.soundonsound.com

International Business Development

Designer Andy Baldwin

subscribe@soundonsound.com

marketing@soundonsound.com

O N LIN E

S UB S CR I P T I ON S

NORTH AMERICA

M AR K E T IN G

Designer Mick Reilly

Design Andy Baldwin

UK/WORLD

“This level of integration is possible only where a single developer has complete control, and that makes some people uneasy.”

EDI T O R I AL

Production Manager Michael Groves

WOR LDW I D E E D I T I O NS

Editor In Chief

A DM I N I S T R AT I O N

Reviews Editor Chris Korff

WWW.SOUNDONSOUND.COM/SUBSCRIBE

Sam Inglis

Administrator Nathalie Balzano

Web Content Editor Callum Hall Web Editor Adam Bull Digital Media Manager Atheen Spencer www.soundonsound.com twitter.com/soundonsoundmag facebook.com/soundonsoundmag instagram.com/soundonsoundmag

The contents of this publication are subject to worldwide copyright protection and reproduction in whole or part, whether mechanical or electronic, is expressly forbidden without the prior written consent of the Publisher. Great care is taken to ensure accuracy in the preparation of this publication but neither Sound On Sound Limited nor the Editor can be held responsible for its contents. The views expressed are those of the contributors and not necessarily those of the Publisher or Editor. The Publisher accepts no responsibility for the return of unsolicited manuscripts, photographs, or artwork. © Copyright 2026 Sound On Sound Limited. Incorporating Music Software magazine, Recording Musician magazine, Sound On Stage magazine, SPL magazine, Sound Pro magazine and Performing Musician magazine. All rights reserved. All prices include VAT unless otherwise stated. SOS recognises all trademarks.

Unlock Your Best Mixes If you can’t trust what you’re hearing, you can’t trust your mix. The truth is your room is likely holding you back. Introducing ORIA Mini, the room correction powerhouse that sits between your audio interface and speakers. Powered by Sonarworks, ORIA Mini is a dedicated hardware solution that analyses and corrects your room’s acoustic imperfections in minutes, transforming your studio monitoring.

56 INSIDE TRACK

IN THIS ISSUE

www.soundonsound.com

July 2026 / issue 9 / volume 41

WIN

FEATURES 30 How AI Music Works AI music generation services are trained on a vast library of existing music. We explain how the process works, and what you can do to keep your own tracks from being assimilated.

48 How I Got That Sound

ARTURIA MEGA BUNDLE WORTH €1330

Massive Attack producer and collaborator Neil Davidge chooses the band’s own ‘Teardrop’ as a career standout.

56 Inside Track: Gabe Simon Rediscovering the human has helped producer Gabe Simon to enjoy colossal hit albums with Noah Kahan and Dermot Kennedy.

64 Mixing For Mobile

PAGE 90

Most listeners discover music on phones, where attention spans are short. Should that affect how we approach mixing?

72 Modular Ferry Island Modular’s Justin Goney wanted some synth sounds to go with his drums; now the company he founded make a four-dimensional wavetable oscillator...

84 Peter Tägtgren: Producing Metal

122 Spotlight: Expandable

Eight-channel Interfaces

ADAT-equipped audio interfaces are a blessing if you need more channels, but the choice can be overwhelming. We round up some of the best eight-channel offerings.

Peter Tägtgren’s Abyss Studio has recorded some of the most influential metal music of the last 30 years — including his own projects Pain and Hypocrisy.

136 Q&A

100 Talkback

138 Why I Love... Shoegaze

Xiu Xiu main-man Jamie Stewart explains why you should always take things too far.

In an age of pristine audio, Sam Mortley finds joy in the washed-out sound of dream pop.

July 2026 / www.soundonsound.com

Your studio and recording questions answered.

ON TEST

114 Akai MPC Sample

108 iZotope RX12 Advanced

Sampling Workstation

70 ALM/Busy Circuits

Quaid Gigaslope Eurorack Module

92 ASM Leviasynth Polyphonic Synthesizer

118 Bad Cat Purepath Passive Analogue Speaker Simulator

54 Balcony Left Little Sir Rekko FET Compressor Plug-in

Dual-channel Microphone Preamplifier & EQ

Audio Restoration & Repair Software

119 Karanyi Cloudmax 3 Granular Reverb Plug-in

50 Mackie Thump Go & ThumpSub Go Portable PA System

500-series Microphone Preamplifier

22 Cherry Audio DS-2 Software Synthesizer

126 Edu Prado Sounds Zither Sample Library

MQA Labs Inspira & Endura Impulse Response Correction Plug-ins

96 Native Instruments Absynth 6 Software Synthesizer

24 Oeksound Soothe 3 Resonance Suppression Plug-in

Otter Mods OTTeR

127 Quiet Music SacredTones Sample Library

44 Shure KSM44MP, KSM40C

Sample Library

Capacitor Microphones

ginTronic transparentSea Desktop Multi-effects Unit

Sonible pure:level Level Riding Plug-in

40 Source Elements

Source-Connect 4.1 Remote Recording Software

20 Temecula DSP DEEP/4 Multi-effects Plug-in

120 The Struggling Musician’s Guide

To Earning Money Book

UB DSP Grit Blender Harmonic Distortion Plug-in

36 Vintagetools VT N95 PE Analogue Equaliser

127 VSL Big Bang Orchestra 2.0 Sample Library

52 Zero-G Morphology Evolved Virtual Instrument

Eurorack Module

126 Fracture Sounds

Lighthouse Piano

Guitar Effects Plug-ins

Softube Flow Studio Control Surface

119 MixWave EHX Classics Bundle

28 British Audio Technologies

BAT 1588

Josephson C706

Capacitor Microphone

104 Barefoot Footprint 02 Gen2 Three-way Monitors

Hammock Hill Audio PreQ-2R

C O V E R

28 BRITISH AUDIO TECHNOLOGIES BAT 1588

24 OEKSOUND SOOTHE 3

& KSM32C

SSL 1 USB Audio Interface

WORKSHOPS 128 Studio Pro 130 Pro Tools 132 Cubase 134 Logic

www.soundonsound.com / July 2026

ON TE ST

BOB THOMAS

any pro-audio product manufacturers today take their inspiration from the second half of the 20th Century and, while that can be wonderful, I always find it refreshing when I come across someone whose products march to the beat of a different drum. One such company is Hammock Hill Audio, a boutique manufacturer based in Austin, Texas. Owner Indri Schaelicke declares his passion for music proudly, along with a design philosophy of “capturing the full frequency profile of an incoming signal and allowing the user to shape the tonality of the result”. This aims to offer, he says, “unique combinations of features and sound characteristics to provide engineers and artists with new tools that help them turn their creative vision into reality”. Hammock Hill’s first product, the PreQ22 microphone preamp and EQ, appeared in 2022. A single-slot, 500-series unit, it features a discrete op-amp and output transformer topology, a switchable two-band sweep EQ, a fully variable high-pass filter, and a post-transformer output fader. The PreQ22 was followed by another 500-series mic preamp called the LOG (short for Lots Of Gain!), which offers up to a chunky 86dB boost, a high-pass filter, two input impedance choices, and a twin-tap, air-gapped output transformer. Their latest offering, reviewed here, is the PreQ-2R, and for this Hammock Hill have taken two PreQ22s, added a couple of additional features, and housed the result in a 1U 19-inch rackmount enclosure. As with all Hammock Hill products, the PreQ22 is designed and hand-built in Austin, using only full-size, through-hole components.

Silvered Star With a silver-coloured front panel, and logos that fade horizontally from left to right, the PreQ-2R is a stylish unit that will stand out in any rack. Its two sets of channel controls are identical, and lie on either side of a small central on/off toggle switch. This effectively acts as a standby switch, as the integrated

July 2026 / www.soundonsound.com

Hammock Hill Audio PreQ-2R Dual-channel Microphone Preamplifier & EQ

With two channels, ample gain, character on tap, and an EQ section thrown in for good measure, there’s plenty to commend this capable preamp. mains on/off switch and IEC socket with fuse sit on the rear panel, on the opposite end to the balanced audio input and output XLR connectors. Each channel also features a front-panel XLR/quarter-inch (6.25mm) jack combo connector, allowing you to connect a balanced, low-impedance microphone to the XLR, or to use the jack as a DI input for unbalanced, high-impedance instruments, such as guitars and keyboards. Inserting a jack plug bypasses the microphone preamplifier, and routes the signal to a JFET-based preamp. Next to the XLR/jack combination sits a thru (loop out) from the DI input, and this allows you to connect to an external amp, effects unit or other device, or even to feed the output to the PreQ-2R’s other channel, should you wish to record a parallel instrument track with a different EQ and/or gain profile. Next to the thru jack sits the switching triumvirate found on virtually every professional‑grade mic preamp in the known universe: 48V phantom power on/off (with indicator LED), a -20dB pad, and (correctly-labelled this time round) polarity reverse. The sole occupant of a channel’s gain section is a 12-position, sealed gold-contact switch by Grayhill, which selects up to 40dB of gain (between 26dB and 66dB) in steps of approximately 3dB. A bi-colour LED

illuminates green when the pre-fader output signal exceeds a level of -28dBu, and red at -3dBu to warn of the risk of clipping. Each channel’s equaliser comprises a continuously variable 18dB/octave 20-200 Hz high-pass filter (HPF); and a three-band semi-parametric sweep EQ, and both are equipped with in/out switches that allow you to make quick and easy A/B comparisons to judge the effects of your efforts. The EQ provides ±10dB of gain in continuously variable, overlapping bands of 40-400 Hz, 300Hz-3kHz, and 2-20 kHz. The final channel control is a post-transformer fader that attenuates the output signal from unity gain to minus infinity. The idea, of course, is that this allows you to drive the transformer as hard as you wish, while ensuring that the input to the next link in your signal chain isn’t overloaded by the PreQ-2R’s resulting output level.

Insider Information Internally, the PreQ-2R carries high-quality components that have been meticulously hand-soldered to the two beautifully laid-out channel PCBs. An indication of the care taken in assembling the PreQ-2R is that the shielded cables connecting the rear‑panel XLR balanced connectors to the channel PCBs are carefully spaced not only to avoid crosstalk along the cable run, but

also to steer clear of any interference from the output transformers. Power to the boards comes courtesy of an internal, in-house‑designed analogue power supply that’s fed by a custom-made toroidal mains transformer. Given the highly regarded sonic performance of the ancestral 500-series PreQ22, it should be of little surprise that Hammock Hill have adopted an “if it ain’t broke, don’t fix it” approach to the design of the PreQ-2R. Consequently, this rackmounted newcomer’s audio path retains not only the integrated front end, discrete op-amp and output transformer of its ancestor, but also operates at the 500-series voltage of ±16V DC. Similarly, the PreQ22’s high-pass filter and two-band semi-parametric sweep EQ have been carried over unchanged. The 300Hz-3kHz corner frequency range of the PreQ-2R’s mid band has an approximate half-octave overlap with the bands at either end. The PreQ-2R’s technical specifications are impressive. An overall gain range of +66dB (+6 at 26dB minimum gain with the -20dB pad engaged, +66dB at maximum gain) will cover most modern microphones. Its frequency response of 15Hz to 75kHz (±0.5dB); EIN (Equivalent Input Noise) of -129dBu; THD+N of 0.03% (1kHz, +20dBu output level, no load); and clipping level of +30dBu hold out the welcome prospect of a clean, precise and dynamic audio performance that can also deliver both edge and transformer-related warmth and weight at higher input and/or gain levels.

Hammock Vibes Once I’d started working with the PreQ-2R, I began to appreciate just how well it delivers on all the promises implied by its specifications. With every microphone I auditioned, this preamp delivered a clear, open sound

Inside the box, each channel, including its output transformer, is laid out on its own circuit board (top), and the internal power supply board (bottom) is fed from a custom toroidal transformer.

that never felt sterile. At lower gain/signal levels, to my ears there’s an inherent touch of colour and warmth that means the PreQ-2R always sounds organic, rather than overly clinical. As the levels increase, though, the PreQ-2R’s clarity starts to develop a harder, more

rather than attempting more radical surgery. When it comes to stereo operation, the PreQ-2R provides superb channel-to-channel balance. The accuracy of the 12-position Grayhill switches obviously lies at the heart of this. I was impressed with the positivity of their action, though I wouldn’t complain if the knobs themselves were slightly larger, to give

“As the levels increase, the PreQ-2R’s clarity starts to develop a harder, more potent edge — that’s something I couldn’t get enough of.” potent edge — that’s something I really couldn’t get enough of. This slightly forward edge means this preamp is in many ways ideal for bringing focus and presence to vocals, but it also brought a lovely definition and body to acoustic instruments, as well as electric guitars and basses. Similarly, I found that stereo drum stems also benefited from being run through the PreQ-2R, with this sense of ‘edge’ combining with the output transformer’s harmonic distortion to add both weight and definition. If you’re into more creative use of distortion, you also can wind up the input signal and/or the gain control to drive the PreQ-2R into delivering exactly the level that you require. You can then use the post-transformer rotary fader to manage the level of output signal reaching the next unit in your signal chain. I habitually high-pass-filter sources when I’m recording, as I prefer to record the low frequencies that the instrument in front of the mic or DI can produce, rather than those that it can’t. The PreQ-2R’s high-pass filter 20Hz-200Hz range is perfect for my needs as it covers the lower frequency ranges of the bowed and fretted stringed instrument families. The three-band sweep EQ’s smooth and relatively gentle 10dB cut and boost is best suited to tailoring the overall sound of the source,

my big fingers that bit more purchase! As boutique manufacturers, Hammock Hill are well-versed in custom builds for clients, so it won’t come as any great surprise that they also offer a range of customisation options for the PreQ-2R. These include modifying the EQ band frequency ranges, changing their curves from peak to shelving, and altering the slope of the high-pass filter to 12dB/octave. I found the defaults pretty well thought through, though.

King Of The Hill? The Hammock Hill Audio PreQ-2R is truly a high-quality, great-sounding microphone preamplifier, whose overall performance is up there with the best available in its price bracket. To me, the PreQ-2R’s greatest strength is that is has a sonic character that doesn’t owe anything obvious to the iconic giants of yesteryear — while being at least their equal, and some might argue better. This unique character makes the PreQ-2R an attractive value proposition for the first-time purchaser, for those looking to upgrade their current mic preamp, and perhaps also for those who have all the vintage bases covered and are looking to the horizon for something new. If you sit in any one of those categories, the Hammock Hill Audio PreQ-2R should be high up on your audition list. Looking around, it seems I have a 1U rack space available — so it seems unlikely that the review unit will be leaving me any time soon!

summary A truly great-sounding microphone preamplifier with a sonic character that distinguishes it from the pack.

£ $1600 plus import duty & VAT. E audiotools@hammockhillstudio.com W www.hammockhillaudio.com

www.soundonsound.com / July 2026

ON TE ST

Mac users won’t need to install a driver. SSL have enhanced the loopback functionality of their desktop interfaces since I reviewed the 2 and 2+; both those models and the 1 now appears to your computer as eight-in, six-out devices. Outputs 3+4 loop back to inputs 5+6, whilst outputs 5+6 loop to inputs 7+8. In terms of audio specifications, the design of the 1 closely follows that of the 2 MkII. Consequently, the mic preamp is super-clean, at least unless you engage 4K Legacy mode, and the headphone amp is pleasingly loud.

The 1 & Only?

SSL 1

USB Audio Interface SSL boil down the USB interface to its bare essentials!

SAM INGLIS

any companies offer both affordable and premium products, but few span the same range as Solid State Logic. At one end of the market, they’ll happily sell you an analogue console costing six figures. And at the other, they make audio interfaces costing less than the monthly electricity usage of that console. The latest, and most affordable yet, is the 1.

2s Become 1 The 2 and 2+ desktop USB interfaces have been mainstays of SSL’s line-up for more than six years now, and received a thoughtful overhaul back in 2024. The 1 is at heart a cut-down version of the 2 MkII, housed in a very similar but marginally smaller wedge-shaped box. Despite the name, the 1 does offer stereo output, with a large volume control that goes to 11, and also has two mono inputs that can be recorded independently. However, whereas the 2 and 2+ both accept mic, line and instrument signals on each input, the 1 adopts a ‘divide and conquer’ approach. Input 1 has a female XLR rather than a combi socket, and is designed primarily for plugging a microphone in. If you want it to accommodate another source, you’ll need a DI box, or perhaps just a funky cable and

July 2026 / www.soundonsound.com

an enquiring mind. Input 2, meanwhile, presents only a quarter-inch jack socket, and loses the phantom power, ‘4K’ and high-pass filter options. Instead, a single button toggles between instrument and line modes, shifting both the gain range and the impedance to suit different source types. Output signals emerge on a pair of line-level quarter-inch jacks, and a single front-panel headphone socket. SSL have also further simplified the 2’s already simple direct monitoring arrangements, reducing user control in the 1 to a single Mix button. When active, this feeds both inputs into the monitor path at a predetermined level. Both are centrally panned, and are routed to both the headphones and the main outputs. However, there is one 1 feature not available on the 2 or 2+: a second USB-C port. In a laptop recording context, it would be bus-powered as normal, but the ability to use a separate USB supply might prove valuable with phones and tablets. Like the 2 and 2+, the 1 is designed to be plug and play. There’s no software control panel, and it’s class compliant, so

SSL’s desktop interfaces have always been convenient problem-solvers for me. I can’t begin to count the number of last-minute overdubs, podcasts and impromptu performances my 2+ has captured when I’m out of the studio and all I have to hand is a laptop. When I first heard about the 1 I wondered if it would be a bit too limited to fulfil the same role, but looking back, there’s actually very little I’ve done with my 2+ that would be beyond the scope of the 1. In these situations I rarely need to drive two sets of headphones, record two mics, capture a stereo line out from something, or mess about with the monitor balance. What I need is the simple ability to record a mic and/or a guitar in a reliable, noise-free fashion, whilst hearing what I’m doing on cans. If anything, the 1 actually simplifies that process, and it’s even more affordable. What’s not to like?

summary SSL’s streamlined desktop interface is surprisingly capable.

£ £143.99 including VAT T Headline Audio +44 (0)1908 477142

W www.headlineaudio.com W www.solidstatelogic.com

STUDIO QUALITY SOUND ANYWERE YOU GO. THE BEST OF BOTH WORLDS Thanks to its 32-bit float recording technology and dedicated gain controls, the H5studio offers the best of both worlds. There is no risk of clipping and no need for post-processing. With its transparent sound quality, studio-grade preamps and smart synchronisation options, creative professionals can now record anywhere, anytime with the largest microphones of any handheld recorder to date.

ZoomUnitedKingdom

ZoomEurope

Exclusive distribution in D, A, CH, DK, GB, PL, EE, LV, LT, BG, HU, BE, NL, L, GR, CZ, SK: Sound Service U.K. Limited | www.soundservice.uk | sales@soundservice.uk

ON TE ST

ginTronic transparentSea

Desktop Multi-effects Unit

When game developers turn their hand to making studio effects, the results are bound to be highly playable! ROBIN VINCENT

he transparentSea multi-effects unit is the first hardware device from Dresden-based developers ginTronic. It draws on their experience in game design and interactive systems, and there’s an intentional and functional clarity about this box that makes it both beautiful and very playable: it eschews screens, presets and complexity in favour of hands-on, intuitive control with everything out on the table.

Soft Hues & Overviews On my desktop, the transparentSea sits in quiet contrast to everything around it. The soft hues of blue, red and pink against the off-white enclosure with Buchla-style knobs radiate a sense of assured calm amidst the chaos of my working environment. It’s beautifully designed, clear, ordered, well-labelled and solidly built. Those transparent, illuminated buttons are very

July 2026 / www.soundonsound.com

striking, and have a satisfying ‘clack’ to them, as they transition between blue and pink, and occasionally green. The knobs feel solid too, with a good amount of resistance to turning. All in all, then, first impressions of the hardware are that it’s top-notch. One thing that does raise an eyebrow is the choice of physical connections. On the back are four mini‑jack sockets, labelled MIDI Out, MIDI In, Out, and In. Mini‑jacks are, of course, very common in Eurorack modular gear, but the transparentSea is primarily designed as a line-level device. The audio mini‑jacks are stereo, which means many will need to access this awesome, high-quality effects box through a pair of headphone adaptors. I feel that full-sized jacks would have been a more robust choice for something designed for use with synthesizers and outboard gear. But, with the right adaptor, that’s a gripe more than a genuine difficulty.

The effects on offer here are a pair of delay lines, a chorus, a tremolo, saturation (what GinTronic call ‘Heat’), a DJ-style resonant low-pass/high-pass filter, a stutter effect, and a reverb. An enormous reverb, in fact! All the physical controls are on the front panel, and except for a couple of Shift+button combos, it’s almost all one-knob-per-function. In addition, there’s have a Clock button, a Bypass button, input and output level controls, and a wet/dry mix knob. First in the chain are the two digital delay lines, which are identical. On each you have controls for time, feedback and level, with a Shift function for panning, and a nice big on/off button. The button works as a ‘soft’ bypass, letting all the echoes continue to dissipate after you’ve turned it off. The delay time ranges up to around two seconds on the knob alone, or up to four seconds when you hold the Shift button, and it has a nice, slushy tape feel as you push the buffer through time changes. The feedback is clean, well-behaved and, when you turn it up to its maximum, infinite. There’s something very heavenly about being able to build a pad with a monosynth

through the first delay, and then let that play while you noodle some lead through the second one. Alternatively, you can build up two loops, disengage both delays, and play along. With another Shift+button combo, you can put the delays in series, so that one becomes the input for the other, for some more complex delay-on-delay action. Next, we have the chorus and tremolo. These are uncomplicated, doing exactly what you’d expect. The Shape control on the tremolo takes it from a regular sine wave to something sharper. You can pull your wobbles into sync (if you’re not using MIDI Clock) using a tap-tempo option: hold the Clock button and tap the Tremolo button. This also works for the delay lines, although they’re all independently geared (trying to get them all in sync can be a lot of fun!). A global, non-MIDI clock might have been nice, although the inaccuracies in my tempo tapping do produce some rather lovely inconsistencies. The filter is of the state-variable variety: turn left for low-pass and right for high-pass, via a little bit of band‑passing. Pile on the resonance, and it will just about self-oscillate if you give it something to bite on. If you could just borrow the LFO from the tremolo, over on the left, it could make for a rather nice wah effect — a missed opportunity, perhaps? Or maybe the DJ-style cutting and drops are more interesting in this context. Just beneath the filter is the Heat knob, and that adds some soft‑clipping overdrive. It’s nice. Slightly underplayed, but it does bring something to your waveforms. The reverb is vast. The website says it’s an algorithmic room/plate hybrid, with diffusion, decay and tone‑shaping. I’m not sure where all that is on the one knob, but it certainly takes you from gentle reflections to deep space adventures in less than a turn. In the context of just wanting to push out into space rather than control every aspect of a reverb, this is a very beautiful thing. At the end of the chain, we have the stutter. This is a little out of character for an otherwise sensible box of effects, but it makes you feel like you could get away with being an exciting and dynamic performer! Hold the button, and it grabs and repeats whatever is in the buffer to create a ratcheting, stutter effect. It has a Rate control knob, and if you hold Shift, it will do it backwards. When the tranparentSea is receiving MIDI Clock, the stutter will automatically sync to it, and then the Rate knob provides time divisions. It won’t respond to tap tempo, though, which seems an odd decision.

It’s not just the MIDI that’s on mini‑jacks — the stereo audio I/O is too.

When it comes to MIDI, the marketing says that it’s fully fleshed out in both directions — meaning everything can be controlled by MIDI, and the front panel can be used as a MIDI controller. MIDI out had not yet been implemented on the late pre-production model I was sent for review, so I couldn’t test that side of it, but the MIDI implementation does add an interesting level of control and automation that goes some way to counterbalance the lack of CV control.

Getting Hands-on In use, the transparentSea lends itself comfortably to all sorts of ambient noodling. The looping potential and interaction of the two delay lines sets you off on some very gratifying adventures. You can lose lots of time as you drift off into the arms of that lush reverb, before dropping in the gentle throb of the tremolo. The buttons and illuminations give you a certainty about your choices, as do the knobs that are almost always pointing to what they’re doing. The chorus and tremolo are brilliant on pads and organs, while the filter and Heat can liven up many a drum loop. The stutter effect is fun, and encourages you to see the transparentSea as a performance device rather than a set-and-forget box. The interface definitely pulls you in to play. The lights, buttons and controls are all inviting you into their playful arena of sound processing. What’s arguably missing is modulation: there’s no internal LFO available to mess with things, and there are no CV inputs to push parameters around in interesting ways. You can use MIDI, of course, but to me, at least, that always feels less transparent and more complicated to manage than analogue. Really, this box relies on hands-on interaction, and it does that very well, but while the bubblegum colour palette suggests fun

and playfulness, the overwhelming flavour here is vanilla. Not a crappy, supermarket own-brand two-litre tub, though! This is an exquisite serving, made with the finest Madagascan pods. What I mean to say is that the transparentSea isn’t edgy or oozing with character, but is solid, beautiful, clean and unruffled. That’s not to say it’s boring, either; rather, it has a clarity and versatility that will make it useful in any situation, and it doesn’t make it all about itself by overpowering your sound with flavour and eccentricity. I’m still slightly disappointed by the mini‑jack audio connections. If it were a Eurorack module, I wouldn’t have batted an eyelid, but as a desktop unit designed with versatility in mind, they feel out of place. Most of my synths have regular jack outputs, and if I were looking to use the MIDI control, I would tend to plug it into a computer that probably also has regular jacks on the audio interface. That said, with the right adaptors I have had no problem getting the full use out of transparentSea.

Verdict Setting those connectors aside, the transparentSea has more than lived up to its billing as a clear, concise, hands-on, interactive multi-effect processor. It looks beautiful. It stays calm under pressure. And it never fails to make other things sound amazing. When something works this well, it doesn’t need to be bursting with quirkiness.

summary The transparentSea is a refreshing raspberry ripple of a multi-effects box that combines beauty and clarity with sensible and elegant versatility.

£ £521.99 including VAT. T Juno +44 (0)20 7424 1960 W www.juno.co.uk W https://gintronic.io

www.soundonsound.com / July 2026

ON TE ST

Sonible’s latest ‘pure’ plug-in aims to take the pain out of level riding. JOHN WALDEN

here’s a lot to commend about Sonible’s ‘pure’ plug-in concept. The series, which now includes seven plug-ins, offers a streamlined take on the features found in the company’s ‘smart’ plug-ins, giving you power over various functions (EQ, compression, limiting, unmasking, etc) but in a format that makes all of these tasks incredibly simple for the end user. The latest addition, reviewed here, is pure:level, which is their take on labour-saving level riding.

On The Level Level riding does what it says on the tin: detailed volume adjustments (‘fader riding’) are used to achieve a more even signal level throughout a performance. Arguably, this results in a more natural handling of the dynamics than with typical compressors, but given that creating manual automation of level across multiple tracks in a mix is undoubtedly a tedious task, it’s perhaps not surprising that compressors are often employed instead. Of course, pure:level is not the first automatic level riding solution — Wave’s Vocal Rider is something of a classic and others have followed in those footsteps — but with Sonible’s keen pricing, it’s possibly one of the most affordable. For use on a single track such as a vocal (it can be used on any type of source), operation is very simple. Placed as an insert, you simply trigger the Re-Analyze Level button, play back a portion of your track, and pure:level will apply its AI-based ‘smarts’ to analyse the level variation. The central display provides real-time graphical feedback on the level adjustments being applied, and you can adjust both the Intensity (how strongly the plug-in constrains the average level) and the Speed with which adjustments are applied (slow, medium, fast or static). Rather neatly, the Target dB then allows you to adjust the overall level of the signal without any further change to the dynamics, letting you quickly bring a series of tracks back into relative volume balance without needing to adjust or disturb any existing automation data for your mixer’s faders.

July 2026 / www.soundonsound.com

Sonible pure:level Level Riding Plug-in Interestingly, pure:level also offers a Group feature. Individual instances of pure:level can be allocated to one of four groups via a simple Matrix View. You can then set a Group Target dB level to which all members of that group will be adjusted. This can be helpful with things like multiple sets of backing vocals, or stereo panned instruments that need to sit at a similar level to balance across the stereo image. And usefully, if you make changes elsewhere to a track (upstream of where pure:level is inserted in your signal chain) and need to re‑analyse the level in pure:level, the Matrix view lets you do that as a single operation for all tracks within a group.

Mixing Made Easy? Like other members of the pure series, pure:level is designed to do one key task in as straightforward a fashion as possible. I found that it worked really well with vocals (probably the most common application) but I obtained good results using it with dynamic bass and guitar parts also. Cubase reported about a 20ms latency added by an instance of pure:level, so I suspect there may be a little ‘look‑ahead’ going on, and

that obviously limits the applications for live use while tracking, but in a mixing stage of a project that would be a non-issue. I’ve found some other members of the pure series (pure:EQ would be an example) to be great tools when you’re just starting a new mix, and need to quickly get the sources into workable shape. The ease of use makes getting your basic EQ, de‑essing and compression moves off to a flying start and, if more control is eventually required, you might then turn to more fully featured options to finesse the details. Sonible’s pure:limit fits into that same mould and, as with all the pure series plug-ins, it performs tasks that can otherwise be complex or time‑consuming very effectively, very quickly, and super easily. Worth checking out! £ £35. Also included in pure:bundle

(seven plug-ins) £139. Prices include VAT.

W www.sonible.com

summary Sonible’s pure:level makes routine level riding very easy, reducing the need for often tedious manual level automation work with dynamic sound sources.

RME PREMIUM USB AUDIO INTERFACES RME interfaces for studio and live recording offer a comprehensive feature set with sonic excellence and intuitive control for Mac and Windows, including TotalMix FX for flexible routing and mixing. Unrivalled ultra-low latency operation combined with the legendary RME driver stability and maintenance for users who don’t want to make compromises.

Worldwide Distribution: Audio AG, +49 8133 9181-0, www.audioag.com • UK Distribution: Synthax UK, +44 1727 821870, www.synthax.co.uk www.rme-audio.com

ON TE ST

Capacitor Microphone

Josephson

C706 The C706 aims to shine a new light on one of the great capsule designs. SAM INGLIS

anufacturers of capacitor microphones fall into two categories: those who make their own capsules, and those who don’t. You could argue that buying in capsules is like bottling someone else’s wine, but capsule design and manufacture is a highly specialised art, and few people have the skill, experience and facilities to do it well. David Josephson is among those few, and his capsules are employed in quite a few highly regarded ‘boutique’ mics from other manufacturers. Naturally, they are also central to the Josephson Engineering range, where they are typically paired with strikingly original bodywork and innovative electronics.

C Suite The C706 is a fixed-cardioid, externally polarised microphone with a one-inch, edge-terminated capsule. Dubbed the J12 by the manufacturer, this is a painstaking recreation of the ‘brass ring’ AKG CK12. More than 70 years after it was designed, the CK12 is still regarded as the pinnacle of capsule design by many, and its reputation for sound quality is matched only by its reputation for being difficult to manufacture. AKG gave up on it about 50 years ago, and today the flame is kept alive mainly by Josephson and a handful of other skilled operators building CK12-alikes by hand. In fact, Josephson have been making their J12 since 1990, meaning that it has enjoyed a longer

July 2026 / www.soundonsound.com

production run than the original; and during that time, they’ve taken numerous opportunities to refine the design and the production process. The AKG and Telefunken mics that used the CK12 were all transformer-balanced, and many of them had valve-based electronics. By contrast, the C706 uses the same “discrete FET cascode input circuit and active balanced transformerless output circuit” developed for Josephson’s existing C716 model. This delivers a self-noise level of less than 15dBA, a comparatively hot sensitivity figure of 26mV/Pa, and enough headroom to accommodate sound pressure levels up to 130dB. There are no pads, filters or other controls, and the C706 is powered over a standard ±48V phantom supply, from which it draws 6mA. Mics in this price bracket usually come with cosmetically appealing accessories, such as an oversized attaché case, or a smart wooden box that holds the mic just a little too tightly for easy retrieval. Not so the C706, which ships in a foam-lined cardboard box. It is, however, supplied with tough fabric covers that can be placed over the mic when not in use, which are possibly more useful than the aforementioned pretty cases. In any case, if I was lucky enough to own a C706, I would want to keep it on display as much as possible, because this is an absolutely stunning microphone to look at. Measuring 261 x 60mm (about 10 x 3 inches) and weighing 1.2kg (2.5lbs), it’s an imposing beast, and also a fascinating one, because the headbasket is unusually transparent. Structural rigidity is imparted by a broad hexagonal grid, which is lined with a layer of gauze so fine as to be nearly invisible. Looking inside reveals the extensive measures that have been taken to optimise the acoustic properties of the microphone. The detriment made by internal reflections and headbasket resonances to the performance of a mic is sometimes under‑appreciated, but not here. The aforementioned dual-layer structure provides a very open acoustic environment, and Josephson’s attention to detail extends to offsetting the narrow bars that run up either side so that they don’t reflect sound towards each other. Meanwhile, the capsule sits atop a dome-shaped mount which disperses sound in all directions. This arrangement also incorporates effective internal

shockmounting, meaning there’s no need for a separate elastic suspension. Instead, the C706 is mounted using an old-school yoke, with thumbwheels at either side to fix it at your chosen angle. This mounting scheme would make having the XLR connector in the base of the mic problematic, since the cable and connector would tend to foul the yoke. Consequently, Josephson have positioned the socket on the back of the body. Any standard XLR can be used, but Josephson offer a special cable with a low-profile 90-degree XLR plug as a cost option, and this definitely helps to keep things neat.

Capsule Capers Some classic capsule types have a built-in high-frequency emphasis that needs to be tamed in the mic’s electronics. Others seem to benefit from the audible

mics. (I would love to compare the 706 to Nordic Audio Labs’ superb NU-24K, as well.) As a vocal mic, the C706 is also remarkably immune to plosive blasts, even when used without a pop filter. And, of course, it isn’t just a vocal mic. I loved it on acoustic guitar, and a pair would offer high-class capture of a drum kit or grand piano. Josephson don’t publish frequency response plots, and I don’t doubt that the C706 would measure as a ‘bright’ mic. But such is the quality of the capsule and its acoustic environment that it comes across as relaxed and true to life, rather than tizzy or artificial. What would probably be recorded as a broad presence boost in a test chamber in practice manifests itself as a subtle enhancement to the clarity of an otherwise delightfully balanced sound. This is particularly apparent

“What would probably be recorded as a broad presence boost in a test chamber in practice manifests itself as a subtle enhancement to the clarity of an otherwise delightfully balanced sound.” contribution of valves, transformers and other components with a slightly non-linear response. But the consensus among some of the world’s best mic designers is that a good CK12-type capsule shines in the most transparent possible context. For example, the Austrian Audio OC-S10 exemplifies a somewhat similar ‘purist’ approach to designing the mic around the capsule, albeit that Austrian Audio’s CKR-12 is a modern evolution of the CK12 design rather than an exact copy. Interestingly, the OC-S10 and C706 are almost exactly the same size and shape, although the latter is a much simpler affair, with no switchable polar patterns, filters, pads, or dual outputs for after-the-fact manipulation. But with the Austrian Audio mic set to cardioid, the sonic similarities between the two are obvious. On-axis, my comparisons suggested that the C706 and OC-S10 are closely matched through the upper midrange, with the Austrian Audio model exhibiting a small peak at 10kHz that is absent from the former, and a marginally thicker low midrange. On my own voice, I found the C706 somehow more natural, but both are extremely good

in its presentation of vocal sibilants, which are represented in a calm, matter-of-fact fashion, without smearing or added resonances. You’d never think twice about putting a C706 up on a breathy vocalist, or a jangly item of hand percussion. Being a fixed-pattern mic with no frills, the C706 represents an attempt by Josephson to make their designs more accessible to a wider market. However, these things are relative, and it remains one of the more costly solid-state mics on the market right now. Such is the level of engineering on offer that I don’t think anyone who tries it will feel that it’s overpriced — but I would recommend making sure that you have the funds available before you try it, because you’re not going to want to give it back. I certainly don’t.

summary A beuatifully constructed mic with a natural, articulate sound.

£ £3690 including VAT. T Studiocare +44 (0)151 236 7800 W www.studiocare.com W www.josephson.com

www.soundonsound.com / July 2026

ON TE ST

Cherry Audio DS-2 Cherry Audio’s take on the Crumar DS-2 could be better than the original...

GORDON REID

don’t view Gordon’s Rules Of Synthesizer Design as particularly onerous. One of these is, ‘No integrated monosynth shall cause me to puff when I carry it up two flights of steps.’ Another is, ‘No synthesizer shall greet me with a cloud of mildew when I take the lid off after an extended period of storage.’ A third is, ‘No synthesizer shall have an immovable knob, even if it hasn’t been twiddled in a while.’ On all three counts, my Crumar DS-2 failed miserably when I liberated it for this review. It occurred to me that there must be a better way to obtain its sounds and the experience of playing it. Let’s get one thing clear; the Cherry Audio DS-2 is not a slavish recreation of the strange combination of a dual-oscillator monosynth and a basic paraphonic synth that is the Crumar DS-2. In keeping with almost all modern soft synths, it offers far more than its inspiration. So, while Cherry’s engineers have modelled the unusual waveform generation and filtering that gave the Crumar its character, their version offers things such as an additional audio waveform, oscillator sync and the pitch contouring

July 2026 / www.soundonsound.com

of osc 2 that makes it useful, a 12dB/oct filter mode, filter cutoff key follow (a big improvement), hugely enhanced contouring and modulation, programmable split/layer for the Synth and Poly sections, an arpeggiator, and a powerful effects section. Oh yes... and the main synthesizer voicing is now (up to) 16-voice polyphonic when desired. If, in 1978, Crumar had tried to build in hardware what this offers in software, the result would have needed a shed for storage, a forklift truck to move it, and it would likely have had a mean time between failures shorter than an episode of Blake’s Seven. Nonetheless, many early synths thrived on their limitations. So a good question to ask (in addition to ‘does it sound any good?’) is whether the numerous enhancements have turned the Cherry DS-2 into just another all-singing, all‑dancing polysynth that overlaps the myriad Prophet-y, Oberheim-y, Roland-y, Korg-y and even Moog-y soft synths you already have. Alternatively, does it bring something new or special to your synthesizer party?

Comparisons The first place to look is in the oscillator section because, on the original, the oscillators generated stepped waveforms rather than shapes with smooth slopes. I set up my DS-2 and listened for the

Software Synthesizer high-frequency pollution generated by the stepping and, sure enough, there it was. The artefacts are most noticeable when using the triangle wave at low audio frequencies, but also exist when selecting the sawtooth and square waves, although they are lower in apparent level. Cherry’s engineers have captured the triangle and sawtooth waves correctly, but the soft synth’s square wave doesn’t appear to exhibit any overt weirdness. It’s not clear why any artefacts would exist on the original, but perhaps Cherry could have investigated this a bit more. Having said that, it’s not an egregious oversight. Staying with the oscillators for a moment longer, Cherry’s engineers have also emulated the weird stepping of the DS-2’s pulse width. When modulated, this can create a disjointed sound that you may or may not find useful. The Crumar’s 24dB/oct filter is based upon four OTA stages. It’s noisy and prone to distortion, its resonance can be harsh, and the onset of self-oscillation occurs suddenly. If, in the late ’70s, you were after a creamy, controllable Moog-ish sound (and who wasn’t) this wasn’t a synth to place high on your list. As I soon discovered, the engineers at Cherry have smoothed off the rougher edges, so one has to ask whether this enhances or detracts. In terms of programming a wider range of sounds and

obtaining useful results, the former. In terms of occasional unexpected wackiness, the latter. You decide. Now we come to modulation and shaping. Although the soft synth retains much of the spirit of the original, it offers six modulation slots rather than five, with inputs from three LFOs (two with stepped sawtooth outputs) rather than two, and 34 selectable destinations instead of five fixed ones. What’s more, the LFOs now offer clock sync, retrigger (key sync), one-shot and ‘reverse wave’ options, and individual modulation delays. In a similar vein, Cherry have replaced the original four-stage ADSRs with far more flexible velocity-sensitive, loopable, five-stage DADSRs with a negative polarity option on the filter contour. The combination of all manner of sync’ed modulation with looping contours can be extremely creative, taking you — dare I say it — into territory more normally associated with large, modular synths. In addition to monophonic and mono unison modes (the latter for super-fat bass and lead sounds) and its standard polyphonic mode, the main DS-2 synthesis engine offers a ‘Multi’ mode that allows you to create four, related versions of the same sound by varying the pitch, pan, filtering and contouring of each. From subtle ‘slop’ to wide, ensemble effects, the results of this can be very satisfying.

In Use Cherry’s engineers have also recreated the original’s paraphonic synth engine. This comprised a single divide-down, quasi-sawtooth waveform that passed through a two-band EQ (erroneously named high-pass and low-pass), before being presented to the main filter and amplifier. The soft synth duplicates this with 32-note polyphony and adds a wider range

of contouring plus three octave ranges. There’s no longer a dedicated LFO selector and amount fader, but you can replicate these in the LFO matrix, so that’s no loss. Combining the paraphonic section with the main synth in poly mode puts you firmly in three-osc/voice polysynth territories — sonically, if not strictly. In addition to the synth voicing, there’s an arpeggiator that offers all of the expected facilities. You can apply it to the main Synth section, the Poly section, or both simultaneously, allowing you to create and play along with a range of atmospheric soundscapes. I hadn’t expected this, and it’s good. Finally, we come to the Effects section, which offers three chains (Synth, Poly and Global), each with five slots into which you can insert your choice from 20 effects units, some of which are quite unusual and take you off in all sorts of interesting directions. There’s even an additional modulator for the effects in each chain, and you can control each of these (or not, as you choose) using the mod wheel. I must admit that I wasn’t overwhelmed with the Cherry DS-2 when I first played it but, as I experimented and learned to embrace its architecture and timbre, I realised that I was creating some excellent sounds that were subtly different from those that I would have obtained elsewhere. I also enjoyed playing it far more than the Crumar sitting alongside it, not least because its monosynth offers a last-note priority option that the original lacked. What’s more, you can map almost every control to MIDI CCs for automation and advanced performance using suitable controllers. But I’ve reserved my final compliment for an unexpected but very welcome addition — four dedicated modulation slots for aftertouch, with independent bipolar amounts and the same 34 destinations found in the LFO mixer.

Switching to the Effects view reveals five slots into which you can place your choice of effects from a pool of 20.

Cherry Audio DS-2 $59 pros

• There’s no mildew, and it’s vastly more usable than the original. • Its architecture might lead you to sounds that you wouldn’t otherwise discover. • To an extent, basing the GUI upon the original’s control panel hides the power within.

cons

• It’s less raw than my 48-year-old original — which you may or may not view as an improvement.

summary

Cherry have both extended and tamed the beast. It’s now a surprisingly powerful synthesizer, and reaches so far beyond the original that, in use, there’s little comparison between the two. I started out a bit ambivalent but, after a few days, I really liked it.

Both channel and polyphonic aftertouch are supported, so I set about programming interesting variations of my favourite ARP ProSoloist and Yamaha CS80 patches. Thank you, Cherry!

Conclusions Whenever designers add additional facilities to a quirky vintage synth there’s a risk that, rather than becoming a Super-Whatever-It-Was, the result is little more than another contender in a world already swamped with impressive polyphonic soft synths. But by retaining the slightly metallic waveforms of the original plus a few other quirks, much of the Crumar’s individuality has survived. Of course, many of the Cherry DS-2’s facilities now overlap those of other soft synths with powerful modulation, extensive effects sections and advanced performance capabilities but, if you compare its sounds against virtual Prophets, Jupiters, Oberheims and others, you’ll find that it can still occupy its own sonic space. It may or may not have the character you’re chasing, but there’s no denying that it has character, and that’s no bad thing. Even if you’re replete with soft synths of ‘the classics’, DS-2 could add an additional flavour that might lead you to something unexpected and interesting. At the time of writing, it costs just $59, which could be very good value if it triggers an idea or produces a sound that you might never have otherwise obtained. £ $59 including VAT. W www.cherryaudio.com

www.soundonsound.com / July 2026

ON TE ST

Oeksound Soothe 3

Resonance Suppression Plug-in

How much more soothing can Oeksound’s Soothe get? Quite a lot, as it turns out...

Oeksound Soothe 3

SAM INGLIS

£199

ince plug-ins started to become popular, some 30 years ago, many new types of processor have been introduced. A few have become universally successful, the most obvious example being real-time automated pitch correction. But many of these new ways of modifying recorded sound have remained niche products. Plug-ins that have become universal ‘must haves’ don’t often do something fundamentally new or original: rather, they offer a faster or better-sounding or more versatile way to implement existing concepts. Think, for example, of Soundtoys’ EchoBoy, or FabFilter’s Pro-Q. Then, back in the mid-2010s, a group of Finnish developers identified a problem that hadn’t previously been solved and

pros • Improved algorithms and new features mean Soothe 3 is both a more effective corrective tool, and now a creative tool too. • Simplified EQ-style interface and new Detail control make it easier to use than previous versions, yet more powerful. • Comprehensive multi-channel linking and weighting.

cons • Not backwards compatible with Soothe 2.

summary An imaginative and very well thought-out upgrade to an already impressive processor.

July 2026 / www.soundonsound.com

developed a tool to address it. Ten years later, that tool is well on the way to the ubiquity enjoyed by plug-ins like Antares’ Auto-Tune. Other companies have since released alternative implementations of the same idea, and it was arguably anticipated in some respects by products like Paul Frindle’s Dynamic Spectrum Mapper, but Oeksound’s Soothe is by far the most widely used processor of its kind.

Mellow Gold The original raison d’être of the plug-in is summed up well by its name. Soothe was initially launched as a tool for addressing harshness in recorded audio, and much of its success is down to the fact that the default preset worked remarkably well at doing exactly that, across a wide range of material.

Like many subjective qualities, harshness is hard to define clearly, or even to agree on, but Oeksound’s insight was to identify it with an excess of temporary resonances. For instance, when a vocalist belts out a high note, overtones of that note can become too strong, an issue which is only made worse by inappropriate mic choice or a compromised signal chain. But the key point is that these overtones are specific to the note in question. Other notes will produce different overtones, and quieter passages may well be completely problem-free. Consequently, straightforward EQ is not usually a satisfactory solution. The overtones are too narrow to target effectively, they aren’t always present, and they aren’t always at the same frequencies. So, although it presents an EQ-like interface, Soothe is not an equaliser. The filter, bell and shelving curves set up by the user don’t affect the sound directly; rather, they weight the plug-in’s algorithm to focus more or less on different frequency regions in its hunt for annoying resonances. The algorithm acts dynamically, so it’s continually identifying the resonances as they

come and go, and automatically applies attenuation only when they occur. Soothe’s Depth control determines how aggressively this is done, while the Selectivity and Sharpness settings let you choose whether the attenuation should be broad and smooth, or surgically targeted. Released in 2020, version 2 of Soothe greatly expanded the feature set. The original ‘hard’ algorithm was joined by a new ‘soft’ variant, which is typically more transparent in use. Compressor-style attack and release controls were added, along with the option to use an external side-chain input to trigger the processing. Soothe 2 added new powers to handle mud and congestion in low frequencies. And, in the background, the entire codebase had been rewritten, reducing latency by around 50 percent and making possible the development of a Live version targeted at front-of-house mixing.

Attention To Detail Soothe 2 felt like a pretty complete tool, and you could be forgiven for wondering whether Oeksound had left much room

for further improvement. Following its release, they went away and developed the superb Bloom, another dynamic tone-altering plug-in that complements Soothe without encroaching on the same applications. But now, in time for the 10th anniversary of Soothe’s launch, they’ve introduced another major update to their flagship product. Like version 2, Soothe 3 is not backwards-compatible, so any existing projects that you load will open with Soothe 2 active unless you manually replace it. The interface is now fully scalable, rather than just offering a choice of fixed sizes, and is rendered in a warmer and more friendly palette of blues. Everything is pretty much where it was before, but nothing is exactly the same! In use, perhaps the most immediately obvious change is that the Selectivity and Sharpness controls have disappeared. In their place is a single knob labelled Detail. The division of labour between the two was never very clear, and the new arrangement is both quicker and more intuitive to use. In

www.soundonsound.com / July 2026

ON TE ST

OEKSOUND SOOTHE 3

fact, I’d actually forgotten that there used to be two controls until I opened up an old project with Soothe 2 active; and if there’s anything you could do with both that can’t be done with the new Detail control, I haven’t found it yet. There are also major improvements to the EQ-style interface. Soothe 2 presented six bands, all active by default, and although these were colour-coded, they were otherwise not easily differentiated visually. Soothe 3’s default preset includes only high‑ and low-pass filters, now immediately identifiable by their lozenge-shaped icons, and a single bell band. Additional bands can be added by double-clicking, up to a total of eight, and there are several new types available including band‑pass and tilt. Soothe 2 introduced the option of working in Mid‑Sides mode as well as left-right stereo, and channel balance is an area that has received a lot of attention in Soothe 3. Each band can be individually weighted anywhere between 100 percent left and 100 percent right, or Mid and Sides respectively if the plug-in is operating in that mode. This is done using a new ‘focus’ setting which is operated in an intuitive click-and-drag fashion. Better still, this extends to surround too, with similar linking and weighting options available in multi-channel configurations up to 9.1.6.

Across The Board The other major addition is a new set of controls labelled Tilt. Again operated in a very immediate way by clicking and dragging, this in effect allows the Detail level and the attack and release times to be varied with frequency. For example, it’s relatively common practice with multiband dynamics processors to use shorter attack and release times at higher frequencies, where fast transients live, and slower time constants in the bass, where overly rapid adjustments can lead to distortion. The Tilt settings allow you to do the same with Soothe. As well as the Depth control, Soothe has always offered a wet/dry mix option, which provides a slightly different way of varying the overall amount of processing. This is now joined by a Wet Trim setting, which does what you’d expect, but perhaps my favourite feature in Soothe 3 — in fact, definitely my favourite feature in Soothe 3 — is the new Max Cut parameter. Unlike most of Soothe’s parameters, this is scaled in real-world

July 2026 / www.soundonsound.com

units (decibels), and allows you to impose a hard limit on the amount of attenuation that is applied at any given frequency. The Max Cut setting might not sound like a radical development, but it actually opens up all sorts of new possibilities. When Soothe is being used as a corrective tool, it makes it a lot easier to find settings that work effectively on the quiet bits without going too far on the loud bits. In a situation where you’re triggering Soothe from an external key input to open up space for a vocal in a wall of guitars, things become much more controllable. And, for me, it was the key to a new range of applications that might be called creative rather than corrective. Especially on drums, I have now started to use Soothe almost as an alternative to EQ; by dialling in aggressive settings but limiting the amount of gain reduction, you can effectively change the timbre of stick hits on hi-hats, for example. (It would be even better if Max Cut could also be tilted across the frequency spectrum, perhaps my one tiny wish for Soothe 4!)

Under The Hood The power behind the throne here comes from changes Oeksound have made to the actual algorithms. Improvements have been made to both soft and hard modes, and Oeksound say that there has been particular focus on making the algorithms contextually ‘aware’ of the overall tonality of the source. This, in turn, means that when you push Soothe 3 hard, it behaves differently from previous versions. I’m not sure that ‘transparent’ is quite the right

The new Max Cut parameter imposes a hard limit on how much attenuation Soothe can apply at any given frequency.

word to describe this, but it certainly allows you to go further in changing the timbre without telltale side-effects becoming apparent. And at the same time, the new band‑pass band type makes it much easier to limit Soothe’s processing to a specific frequency region when you do want to do something surgical. I’ve used Soothe 2 intensively ever since it was released, and if you’d asked me what features should be added in a version 3, I’d probably have scratched my head and struggled to answer. So full marks to Oeksound for coming up with a comprehensive update that not only makes it better at its core job, but also enables new applications. Previous versions of Soothe were invaluable tools for fixing badly recorded audio. Soothe 3 improves upon its predecessors in that role, but is also a genuine alternative to tools such as broad-brush EQ and multiband compression for creative manipulation. And thanks to the merging of the Selectivity and Sharpness controls, it’s actually easier to use than Soothe 2, despite being more versatile. I think every existing owner will want to upgrade, especially given the very fair cost of doing so. And if you’re not already a Soothe user, now is definitely the time to take a second look at this uniquely capable plug-in. £ £199; upgrade from v2 £45. Prices include VAT.

W www.oeksound.com

ON TE ST

British Audio Technologies BAT 1588

500-series Microphone Preamplifier

Clean gain isn’t exactly a problem for us these days, so we often look to dedicated preamps more for their tonal character and ease of control. The new 1588 offers both. NEIL ROGERS

ew company British Audio Technologies (BAT) are, as you’d expect, based in the UK, and they are just one string to recording and technical engineer Matt Sartori’s bow. In fact, this isn’t our first encounter with Matt’s work: he’s the brains behind the Periscope mic we reviewed in SOS February 2021 (https://sosm.ag/scope-labs-periscope), a design that’s as distinctive as it looks, since as well as the mic itself there’s an aggressive-sounding compressor built in. Matt has also been a friendly face at some of the UK’s leading pro-audio repair services for over 20 years, as well as looking after the technical and service needs of several commercial studios on both sides of the Atlantic. Like many boutique designers, he explains that he tends to develop his products when he feels he needs something that other people haven’t quite managed to offer. It’s an approach that leads us neatly to this review: Matt tells me that he has long been a fan of a particular 1960s solid-state preamp made by Altec, and that this project came about because he was keen to have one to hand in a convenient, reliable format, to take with him

July 2026 / www.soundonsound.com

to different studios or use for location recordings. So it comes as no surprise that he chose to build this new preamp as a 500-series module. It’s priced in line with the top end of preamps in this format, but is intended to deliver more than gain alone — it promises us the unique sonic properties of a classic but perhaps lesser-known vintage studio tool. So let’s find out if it delivers the goods...

From Muse To Modernity The Altec 1588C was introduced in the 1960s as a compact, modular ‘octal’ preamp — meaning it used the same eight-pin base as some of the valve-based modules that preceded it. This was a period in audio electronics that saw rapid change, with many designers and engineers moving away from the more complicated tube-based products, in favour of solid-state, transistor-based designs. At the time, the 1588C was respected as a good, practical option, but in the years that followed it came to be recognised as a device whose

British Audio Technologies 1588 £795 pros

• Great, characterful-sounding preamp. • Capable of subtle or more obvious saturation. • Trim and Fader controls make this a multi-trick pony! • Can be used as a line processor while mixing.

cons • None!

summary

Inspired by the 1960s Altec 1588C preamp, British Audio Technologies’ debut release is an excellent, flavourful addition to the 500-series preamp menu.

transformer-coupled input stage could impart a pleasing character and, when deliberately driven, could take on the role of a saturating, tone-shaping device where that was desirable. The 1588C had its pros, then, but its core circuitry was relatively simple — the design features no op-amps or ICs in the signal path — and it was never really conceived as a fully featured preamp. The original modules offered only a modest 34dB of gain, and they were very much intended to be part of a wider recording chain. Of course, those modules are now very old, too. So from the point of view of today’s recording practices, there are clearly also cons. And that, of course, explains why Matt decided to develop a modern, more reliable version.

Hitting The Metal Essentially, what Sartori has done is to add to the original module design (which, understandably, featured a transformer only on the input stage — the output stage was handled by the device into which it was plugged), to create a more complete preamp module. As well as the input transformer, in this case made by Cinemag, his 1588 features a transformer‑balanced output (also Cinemag), which means that, unlike the original, it can be used on its own or with other processing stages, as you require. In keeping with the original version, the gain structure is somewhat different from that of some more famous solid-state preamp designs like the Neve 1073 or API 312, but with a +25 to +55 dB gain range this version is much more versatile than the Altec. There are just three main controls, but Matt has clearly paid careful attention to the

Every BAT 1588 is individually hand-built, and carefully assembled using premium components.

control layout, which ensures that you can easily hit those different transformer stages to taste, in search of saturation. The top knob, Gain, sets the amplification of the input signal, so determines the signal level fed into a discrete amplifier stage, and thence the output transformer. But it’s not the only means of controlling those levels. Next comes a knob labelled Trim, which is a variable attenuator (0 to -20 dB). A simple control, but one that means you have ample control over the ‘analogue magic’ — you can drive the input stage hard while dialling back the output transformer saturation. Finally, we have an ‘Output Fader’ knob. A post-transformer attenuator, this one lets you optimise the level being fed to the next device in your recording chain, no matter how hard you drive the output transformer. Naturally, the BAT 1588 also offers the typical features you expect of any preamp today: switchable 48V phantom power, a polarity inverter, and a -20dB input pad to accommodate particularly loud sources.

Sat Tests The aesthetic styling of the BAT 1588 very much aligns with my personal tastes, clearly resembling a vintage device but without sacrificing convenience for purism. A pair or more of these in a 500‑series rack would look very inviting indeed! The more important thing, though, is how it sounds, and Sartori deliberately bills this device as a ‘character preamp’, yet while its controllable colour is very

much its USP, the quoted frequency response is within ±0.5dB from 20Hz to 25KHz — so it can also sound reasonably clean if you want it to. In my first recording session, for example, it worked superbly for capturing an acoustic guitar that wasn’t crying out for any additional saturation. With the single input LED only rarely turning red, the 1588 produced a lovely, ever-so-slightly ‘rounded off’ sound when the musician dug into his strumming on the chorus sections. I did notice, however, that I had to leave both the Trim and Fader controls fully up, to get enough level out of the preamp when used this way — ideally, I’d have had a little more final output level at my disposal to keep a cleaner tone. That’s perhaps less of an issue with passive ribbon and dynamic mics, for which you can use an in‑line preamp such as a Cloudlifter before the 1588. And on the subject of dynamic mics, my next test was pairing the 1588 with a beyerdynamic M 201, a hypercardioid mic, used as the top snare mic in a fairly minimal Glyn Johns-inspired drum recording setup. I knew this was a session where vintage-style saturation would be appropriate, and the thick, slightly overdriven sound that the 1588 helped coax out of the low-tuned snare drum was very satisfying indeed. In this application, the input LED was nearly always in the red, and I was able to play with the Trim and Fader controls to find the sweet spot both level- and saturation-wise.

Vocals were another source on which this preamp impressed. On a couple of sessions, I paired it with my old Neumann U87 capacitor mic to record vocals that I felt would benefit from a little thickening coloration. I do find that you have to be careful in such scenarios — it’s not always appropriate to have audible saturation or distortion on a vocal, and you really have to be sure that you’ll like how a more coloured preamp sounds if the singer suddenly decides to gets louder; something to bear in mind when the singer’s warming up. On one particular session, I was recording more of a scratch vocal using an RE-20 dynamic mic, with the singer alongside me in the control room. This time, I felt comfortable just letting the preamp go very ‘hot’, and we loved the more obviously overdriven, compressed vocal sound this presented. So much so, in fact, that I recreated that very same signal chain when it came time to cut the final vocal. Lastly, I also made a point of using the 1588 as a line-level insert when mixing. While the character is appealing, you might find it a bit too much if used on its own like this, but it sounded great when used as a kind of ‘parallel dirt’ channel to enhance a bass guitar and some of the drum mics.

Summing Up Most of us already have technically good preamps at our disposal in our interfaces, and when that’s not enough we’re fortunate to have a lot of mic/line preamps to choose from today, including lots of very tasty options at around the price of this new boutique release. In the 500-series format in particular, we’re really spoiled for choice! As both an engineer and a pro-audio gear reviewer, what I’m really looking for in a new preamp, then, is either an option that nails its sonic colours very firmly to the mast, or one that can do a few different things very well. In a recording session, ideally you want a nice zone where you can hear that colour being introduced, but with enough range in the controls that you can make adjustments easily, before it becomes too much — while knowing that if it does get a bit too much, it’s still going to sound good. BAT’s new preamp achieves all of that, which I find really impressive. This is one preamp that’s well worth checking out. £ £795 including VAT. T Studiocare +44 (0)151 707 4545 E sales@studiocare.com W https://studiocare.com W https://britishaudiotech.com

www.soundonsound.com / July 2026

FE ATURE

How AI Music Works And How To Protect Your Music

AI music generation services are trained on a vast library of existing music. We explain how the process works, and what you can do to keep your own tracks from being assimilated.

JOSHUA STEWART

hen you hear an AI-generated song for the first time — a convincing one, not the uncanny-valley output of a few years ago — the reaction is usually somewhere between fascination and dread. The technology is moving fast. Services like Suno and Udio can now produce broadcast-quality tracks in under a minute, complete with vocals, instrumentation and arrangement. But how does any of this actually work? How is our own music exploited in order to

make these systems work? And is there anything we can do about it? These are questions I’ve been grappling with for the past two years, both as a musician and as a researcher. The answers turn out to be more nuanced — and more hopeful — than most of the headlines suggest.

How AI Generates Music To understand how AI music generation works, you first need to understand that these systems don’t ‘compose’ in any meaningful sense. They don’t understand harmony, melody or rhythm

the way a musician does. Instead, they learn statistical patterns from enormous quantities of existing music, and generate new audio that follows similar patterns. The process begins long before anyone types a prompt. It starts with training. A raw audio file is, at its most basic level, a sequence of numbers: amplitude values sampled tens of thousands of times per second. At CD quality (44.1kHz), that’s 44,100 data points for every second of mono audio. Processing this directly would be enormously expensive in computational terms, so AI systems compress the audio into more manageable representations. The most common approach is to convert audio into a mel spectrogram — essentially a visual map of how the energy in a piece of audio is distributed across frequencies over time, weighted to approximate how humans perceive pitch. If you’ve ever seen a coloured heatmap of a song’s frequency content in a DAW or spectrum analyser, you’ve seen something similar. These spectrograms condense the raw waveform into a format that’s far more efficient for a neural network to process, while preserving the information that matters most. More recently, systems like Meta’s EnCodec have taken this a step further. EnCodec uses a neural encoder to compress audio into discrete tokens — short numerical codes, each representing a tiny fragment of sound. Think of it as converting audio into a kind of vocabulary, where each ‘word’ represents a particular combination of timbral and tonal characteristics. The encoder breaks the audio down; a decoder reconstructs it later. The compression is remarkably efficient:

The AI music pipeline starts with training from thousands of real-world examples.

July 2026 / www.soundonsound.com

use latent diffusion, where this denoising process happens in a compressed representation of the audio rather than on the raw waveform itself, making it far more computationally efficient. In practice, many systems combine elements of both. Suno, for example, appears to use autoregressive generation combined with diffusion-based upscaling for high-fidelity output. The details of these commercial architectures are proprietary, but the general principles are well established in published research.

The Sample Rate Question

At present, the most prominent commercial AI music generation services are Suno and Udio.

a few seconds of audio can be represented by a handful of tokens. EnCodec uses a technique called Residual Vector Quantisation, which employs multiple ‘codebooks’ that are essentially lookup tables of sound fragments. Rather than trying to capture everything in a single pass, each codebook refines the representation left behind by the previous one, progressively filling in detail. It’s not unlike how a painter might block in broad shapes first, then add finer and finer detail with each successive pass.

Two Approaches To Generation Once we have a system of representing audio using tokens or spectrograms, the actual generation can happen. There are two major approaches in use today, and most commercial services use some combination of both. Autoregressive models work much like the predictive text on your phone. Given

a sequence of audio tokens, the model predicts what comes next, one token at a time. Meta’s MusicGen is the most well‑known example. It uses a transformer — the same type of architecture that powers large language models like ChatGPT — to predict the next audio token based on everything that came before it. The conditioning can come from a text description (“upbeat jazz piano with brushed drums”) or even a melodic reference. The model doesn’t understand jazz or piano; it has learned the statistical relationships between millions of audio tokens that were labelled with similar descriptions. Diffusion models take a different approach. Instead of building audio token by token, they start with pure noise and gradually refine it into coherent audio, guided by the text prompt. At each step, the model removes a small amount of noise, nudging the output closer to something that matches the requested description. Many systems

One important technical detail that affects both the quality and the detectability of AI-generated music is sample rate. Most AI models operate internally at lower sample rates than we use in music production — typically between 16kHz and 24kHz. This is a pragmatic choice: lower sample rates mean fewer data points per second, which means smaller models, faster training and lower computational costs. But music production works at 44.1kHz as a minimum, and increasingly at 48kHz or higher. To bridge this gap, AI systems use neural vocoders — specialised networks that reconstruct a full-bandwidth waveform from the model’s lower-resolution output. HiFi-GAN is one of the most widely used, and it works using a generative adversarial network (GAN): one network generates the audio while another critiques it, and they iteratively improve each other. The result can be impressively convincing, but the high-frequency content above the model’s native sample rate is synthesized rather than generated by the core model. This is one reason many users of these platforms report a characteristic ‘shimmering’ or metallic quality in the output; the vocoder is interpolating detail that the model never actually produced.

How Our Music Feeds The Machine All of this generation capability depends on one thing: training data. And the training data is music — our music. Training a music generation model requires vast quantities of audio. Suno reportedly trained on over 50 million hours of music. Each track in the training set is processed through the feature extraction pipeline described above: raw audio is converted to spectrograms or tokens, and any available metadata (genre tags, tempo, mood descriptors, lyrics) is aligned with the audio representations.

www.soundonsound.com / July 2026

FE ATURE

HOW AI MUSIC WORKS

The phenomenon of psychoacoustic masking allows us to conceal imperceptible artefacts within an audio file that significantly disrupt the AI training process.

The ethical questions around how this training data is sourced remain contentious. Several major lawsuits are ongoing (the RIAA filed suits against both Suno and Udio in 2024), and while some companies now claim to use only licensed material, the reality is difficult to verify. What is clear is that the scale of data required is enormous, and the Internet’s vast catalogue of streaming audio and uploaded content has been the primary source.

What The Model Actually Learns When a neural network trains on this data, it doesn’t store copies of songs. Instead, it learns patterns — the statistical relationships between audio features. After processing millions of tracks, the model develops an internal representation of what ‘music’ sounds like: how certain chord progressions tend to follow others, how vocals sit in a mix relative to instrumentation, how a kick drum’s spectral profile differs from a snare’s. This is an important distinction, because it means the output isn’t a collage of existing recordings. It’s newly generated audio that follows the same statistical patterns. But the patterns were learned from real artists’ work without their consent, and the generated output can sometimes reproduce stylistic elements with uncomfortable specificity. The question of whether learning statistical patterns from copyrighted

July 2026 / www.soundonsound.com

material constitutes infringement is one that courts are still working through.

Feature Extraction: The Vulnerability Here’s where things get interesting for anyone wondering whether they can stop their music from being used to train AI models. The feature extraction stage — where raw audio is converted into the representations that the model trains on — is the critical vulnerability in this pipeline. If you can disrupt what the model ‘sees’ at this stage, you can disrupt what it learns. This is the principle behind a class of techniques known as unlearnable examples. The concept of unlearnable examples was first demonstrated in the image domain by Huang et al in 2021. The researchers showed that imperceptible perturbations added to training images could prevent deep neural networks from learning useful features from them. The data looked normal to humans but was effectively invisible to the model — it couldn’t extract the patterns it needed. It turns out that this principle works even better in audio, because we can take advantage of psychoacoustic masking. Anyone who has worked with audio compression will be familiar with the basic idea. MP3 and AAC codecs exploit psychoacoustic masking to discard audio information that humans can’t perceive — for example, a quiet sound that occurs at the same time and frequency as a louder one. Our auditory system has well-documented

limitations: a loud signal renders nearby time-frequency components inaudible, and sounds just before or after a loud transient are harder to perceive. These same principles can be used to shape adversarial perturbations. Rather than adding generic noise to an audio file, a psychoacoustic model analyses the specific content of a track — its spectral profile, its dynamics, its transient structure — and computes exactly where perturbation energy can be placed without being heard. A dense rock mix with heavy distorted guitars provides far more masking opportunity than a solo acoustic performance, so a dense mix would receive a different perturbation profile to a sparse one. Every protected track is unique. This is fundamentally different from adding filtered noise. The perturbation is adversarially crafted — mathematically optimised to target the specific representations that AI models rely on. Tape hiss, vinyl crackle and other natural noise sources would have no comparable effect, because AI models are generally quite robust to random noise. They’re trained on noisy data all the time. What makes adversarial perturbation effective is that it exploits specific vulnerabilities in how neural networks extract features from audio.

iCloak: A Working Implementation iCloak is a system I developed that applies this approach. It’s built upon

DR510 500 Series System features: High-power 500 Series rack, USB/ADAT interface, Dante AoIP interface, and D510mx 10x2 analogue summing mixer.

32Classic MS Mix Strip features: High-performance transformer-balanced mic pre, 32Classic console EQ with HP/LP filters, and integrated mixbus.

FE ATURE

HOW AI MUSIC WORKS

HarmonyDagger, an open-source psychoacoustic perturbation engine released under the MIT licence. The system processes audio files — masters, stems or individual tracks — and applies perturbations tailored to each file’s content. To validate the approach, I conducted formal tests at both 22,050Hz (the standard sample rate in machine learning audio research) and 44,100Hz (the industry standard for music production). The tests evaluated protection against three categories of AI model: • S peaker verification (Resemblyzer): used in voice cloning pipelines to capture vocal identity. • Audio feature extraction (Microsoft CLAP): a general-purpose model that learns audio-language representations. • Neural audio compression (Meta EnCodec): the tokenisation system used by music generation models like MusicGen. The results showed statistically significant disruption across all three model types. At 44.1kHz, EnCodec token divergence reached 43.9%, meaning nearly half of all tokens produced from a protected file differed from those produced from the unprotected original. CLAP embedding distance for music improved by 425% at 44.1kHz compared to 22kHz, suggesting that the protection becomes more effective at production sample rates due to the additional high-frequency bandwidth available for perturbation placement. Critically, perceptual quality remained extremely high. PESQ scores exceeded 4.5 (on a 1–5 scale where 4.5 represents imperceptible degradation), STOI (Short-Time Objective Intelligibility) measured above 0.999, and signal-to-distortion ratio exceeded 25dB for both speech and music. Listening tests across a range of studio monitors and studio headphones confirmed that the perturbation was inaudible. No protection system is perfect, and there are important caveats to acknowledge. The formal tests used synthetic audio corpora rather than commercial recordings, though the methodology (standardised metrics, statistical significance testing) is well established. Testing against a wider range of AI models, including diffusion-based synthesis and retrieval-based voice conversion, is an active area of research. Robustness against

July 2026 / www.soundonsound.com

Perturbation in action on a music file. The spectrograms of the raw and protected files are visually indistinguishable, and the added noise is completely masked by the wanted content.

lossy compression such as MP3 and AAC is another important question: anecdotal testing through commercial AI platforms using MP3 files suggests the protection holds up, but formal codec-specific testing is ongoing. And while psychoacoustic modelling provides strong theoretical guarantees about inaudibility, formal listening tests in high-SPL environments (such as large PA systems) have not yet been conducted.

From Protection To Detection Protection aims to stop AI from learning our music. A parallel challenge is detection: identifying music that AI has already generated. These are complementary problems, not competing ones. There are numerous reasons why we might want a reliable way to detect AI music, for example to stop streaming platforms being flooded with content, and to ensure fair distribution of royalties to human artists. However, the detection challenge is significant. As generation models improve, the spectral and temporal characteristics of their output increasingly resemble real recordings. But there are still telltale signs. AI-generated audio often exhibits subtle spectral artefacts, particularly in high-frequency regions where vocoders are synthesizing content above the model’s native sample rate. Unnatural periodicity,

inconsistent harmonic decay, and characteristic patterns in the noise floor can all serve as forensic markers. Recent research has made promising progress. The SONICS project (Rahman et al, 2025), accepted at the prestigious ICLR 2025 conference, developed a dataset of over 97,000 songs — including 49,000 synthetic tracks from Suno and Udio — alongside a novel detection architecture called SpecTTTra. The system analyses both temporal and spectral patterns in mel spectrograms, using a transformer to model long-range dependencies across entire songs. This is important because many authenticity cues only become apparent over longer timescales. The way a human performance develops and varies over the course of a song is difficult for current generation models to replicate convincingly. Detection is vital work. But part of the reason I believe protecting audio at source is so important is that it’s the only way to ensure these models don’t eventually become so good that detection becomes impossible. Prevention and detection are two sides of the same coin — but if we wait until the models are indistinguishable from real recordings, we’ve already lost that race.

Where Do We Go From Here? The music industry’s response to AI has so far been overwhelmingly reactive:

lawsuits, opt-out registries, lobbying for legislative protection. These are all necessary, but they address the problem after the fact. By the time a lawsuit is settled, the model has already been trained. What’s emerging now is a proactive approach that acknowledges the need to protect audio at the source, before it ever enters a training pipeline. This shifts the power dynamic. Instead of asking AI companies to respect opt-out requests (which relies on their compliance), artists can take direct action on their own masters and stems. The technology is still young. Applying protection in real time, as in a DAW plug-in or a broadcast workflow, remains a significant engineering challenge, because the psychoacoustic model needs to analyse the full audio content to optimally shape the perturbation. But the fundamental approach is sound, and it’s improving. For the first time, artists and producers have a tool that doesn’t rely on legislation, litigation or the goodwill of technology companies. It’s not a complete solution — nothing is — but it represents a meaningful

If it’s possible to successfully protect music from being used as AI training material, it’s likely to be more effective than retrospectively opting out or taking legal action after the fact.

shift in how we think about protecting creative work in an age of generative AI. Joshua Stewart is a musician, researcher and the developer of iCloak. He is currently completing a Masters

degree with research focused on the intersection of music and artificial intelligence. His iCloak app can be found at https://icloak.app, or see it in action at www.instagram.com/melodiq.io/reels/.

www.soundonsound.com / July 2026

ON TE ST

Vintagetools VT N95 PE Analogue Equaliser N I C K M AV R I D I S

or decades, an equaliser was seen less as a creative tool and more as a means of technical control: if you desired a different sound, you’d either alter the source itself, or put your microphone somewhere else. For the best audio quality, though, a crucial stage in the chain from the microphone to the consumer’s ear was the vinyl cutting process — carving the perfect groove helped record companies compete for the highest market share, and when it came to cutting equipment,

Vintagetools VT N95 PE €4268 pros

• Doesn’t sound like ‘an EQ’! • Superb build quality. • Stellar technical specifications. • Convenient to use. • Quickly recallable.

cons

• None, bar the price.

summary

Vintagetools have adapted the classic concept of Neumann’s famous vinyl cutter EQs, and adapted it to meet today’s production realities and technical possibilities. The result is, frankly, astoundingly good.

July 2026 / www.soundonsound.com

What if someone were to lavish money, time and considerable engineering know-how on improving Neumann’s highly revered cutting-lathe EQs — could they really be improved? The answer is a resounding “yes!” there was one company that stood head and shoulders above all others: Neumann. And this brings us back to the subject of equalisation, because an integral part of Neumann’s VMS cutting machines was an ultra-precise equaliser, the W495 (and its predecessors PEa and OEV). Needless to say that cost was not really a consideration when those machines and their equalisers were being developed...

More Than A Clone Today, there’s a small company in South Hesse, Germany, called Vintagetools, who have decades of experience in racking, servicing and repairing notoriously well-built old German recording and broadcast equipment, and now make their own recreations. Amongst those products are the famous V72 and V76 preamps but also, you guessed it, Neumann equalisers. Vintagetools are not simply a ‘cloning’ business, though. They set out to create the definitive German precision equaliser, with no compromise whatsoever. Yes, “no compromise” is one of those overused marketing terms, but when it comes to Vintagetools’ VT N95 PE, its use is entirely justified. This EQ contains over 800 components, most of which, due

to the tight tolerances needed, cannot be bought off the shelf — so testing and selecting involves lots of labour, long before the company’s soldering irons are warmed up! The design employs input and output transformers made by Haufe, and even the manufacture of the front plate involves a time-consuming and expensive dye-sublimation printing process, with several different colours in use (it looks really classy and three-dimensional!). The control knobs are from the original manufacturer, although they are bigger than the original models; Neumann’s cassettes were fairly small, and so too were the knobs. Nowadays, it’s possible to obtain much better technical figures than when those cutting machine EQs were built, and Vintagetools’ aim in designing the N95 was to surpass the original specifications and bring them to the absolute cutting edge. Accordingly, the frequency response is 20Hz to 120kHz (±1dB), which helps transients to pass through the EQ unaltered. THD+N is less than 0.004% — so there’s almost zero distortion. A left-right channel separation of greater than 96dB means crosstalk is vanishingly low, and so too is the

noise level, which, at -99dBA (without the A-weighting, the figure would look even better), allows for high boosts even of highly dynamic material. Vintagetools have also widened the usability of Neumann’s RC design not only by adding frequencies, but also a second mid band. This makes the VT N95 PE a two-channel, four-band EQ. The ±15dB shelving filters have three additional frequencies, and can now be set at 27, 40, 50, 60, 80 or 100 Hz for the low shelf and 7, 8.5, 10, 12.5, 14 or 22 kHz for the high shelf. Both ±10dB mid bands offer 11 selectable frequencies (from 100 to 800 Hz and 1 to 8 kHz) with three bandwidth settings (Q values of 1, 0.5 or 0.25). Because 2dB steps are considered too coarse for

when I put my glasses back on! So, having the EQ inserted is really nice for single signals, but it is also an instant upgrade to any stereo sum. EQ’ing with this unit really is a piece of cake. The frequencies you can select are those you’d need in most cases. You cannot super narrowly notch out resonances or high-pass filter, of course, but that kind of correction is really a job for a ‘to-tape’ channel EQ. This VT device is better viewed as a stereo bus or mastering equaliser. By the way, since I mentioned high-passing, cutting 15dB with the low shelf very often does just what you’d want from a pass filter — for example, too much bass in the guitar bus can easily make the mix sound muddy, and cutting

“Every mix gets ‘more highs’, of course, but it’s more than that — like the first sunbeams of the day illuminating all the details in a landscape.” mastering, a mastering version (VT N95 ME) is also available, with ±1dB steps up to 7dB, though omitting 6dB in both boost and cut. The mastering version comes with a creamy eggshell-coloured front plate, with all-black knobs and labels. It’s possible to order mixed versions of the standard and the mastering version if you like, both in terms of the aesthetics and the technical aspects. You can also specify 0.5dB steps, if you like. I know some engineers find meaningful colours helpful, but personally I’d probably opt for a black fascia and black knobs...

EQ Of The Utmost Quality Of course, SOS didn’t hire me to discuss my views about colour schemes! So, how does the N95 perform sonically? Just what is it that sets it apart from other equalisers? It is, of course, a very transparent-sounding processor — yet, it isn’t quite crystal clear and neither is it intended to be. The signal always passes through the Haufe transformers, even in bypass mode. These don’t thicken the sound much and don’t add a whole lot of harmonics to the signal, even when driven with hot signals. But they do serve to tame harsh elements in the signal; to round spikes and make them appear less ‘glassy’, less ‘spiky’. They add just a touch of ‘graininess’ and ‘grip’ to the signal, so that while the N95 never seems to lower detail, it seems to give everything more ‘contour’. To draw an analogy, working with this Vintagetools EQ feels like

the bass band of course pushes everything down to the very sub-bass, without ever letting the signal get ‘washy’. Both mid bands make it hard for me to believe there is an EQ in the signal path! Usually, I can quickly tell which is the cut or boosted signal and which is the unprocessed original one. With moderate gain settings, this isn’t easy at all when using the Vintagetools N95, so transparently does it work. Even with relatively narrow cuts or boosts, you don’t notice those typical ‘hollow-sounding’ artefacts at the edges of the EQ curve. Yet, the VT doesn’t sound ‘boring’, like the

linear-phase EQ you might use in your DAW. So carefully balancing between snare drum, guitars and vocals in a mix by adding a couple of dB here and removing a touch there is a job for which the VT N95 is the perfect tool. The absolute highlight for me, though, is the treble band. Please excuse the hi-fi-esque description that follows, but it really is nothing short of phenomenal. When boosted, yes, every signal and every mix gets ‘more highs’, of course, but it’s more than that — like the first sunbeams of the day illuminating all the details in a landscape, and bringing to life the rich tapestry of colours within. I’ve heard many EQs praised for such qualities and enjoyed using them — but the VT is simply breathtaking. For example, I’d even consider the Millennia STT-1’s famous treble band (that has garnered such a strong reputation that it’s been made available as a separate product, the NSEQ HT 500) less suitable for adding air to a very modern pop vocal chain than the N95’s high band. What’s also great is how a mix livens up with boosts starting as low as 7kHz with 2dB of gain. I really miss the smaller steps the ME version or bespoke N95s offer, though. Reducing gain in the high band simply makes all material darker, but never duller: it remains lively and agile. The rotary switches allow for simple and quick recalls, of course. Switching between different projects and altering the parameter settings takes mere seconds, and when actively EQ’ing, I could dial in what I desired much faster than with my beloved three-band Roger Schult mastering EQ, which consists of

Unlike the Neumann originals, the Vintagetools VT N95 PE features a four-band design (instead of three) with six selectable frequencies (instead of three) in the shelving filters.

www.soundonsound.com / July 2026

ON TE ST

V IN TAGE TO OLS V T N 95 PE

The positions of the front-panel switches are easily recognisable — even the blue-on-black, thanks to the superb dye-sublimation printing.

six Neumann-based W2377 modules. I noticed something strange too: I always seemed to need less gain when cutting or boosting with the VT than with any other equaliser — and when applying more gain, the N95 still sounded more natural. The Roger Schult is, overall, a great EQ, but by comparison it got a little more ‘raspy’ when boosting the trebles, while the H2 5011 Helios EQ you’ll find in the audio samples that accompany this review (https://sosm. ag/vintagetools-vt-95-pe) is, of course, a far more colourful-sounding unit. I was stunned to discover that the left and right channel are absolutely identical: when I swap left and right used on the same material in a blind test, I am completely unable to tell which is which. As you might expect, then, the VT N95 PE never has any negative audible effect

on the stereo image or the depth when the settings are identical on both channels — it’s a good ear-training exercise to listen for such subtleties with other EQs. It’s the job of a reviewer not just to gush, but also to point out the cons of any product. Yet I’ve had a genuinely hard time trying to find any for the VT 95 PE. I’m not overly enthusiastic about the look, the name and the logo, but that’s only me and it shouldn’t be a major consideration for anybody, especially in the professional context for which this EQ is designed. You might, perhaps, expect me to criticise blue labels on a black front panel, but thanks to the premium dye-sublimation printing this actually reads really well, even when in rather dim light. I wouldn’t say no to bigger bandwidth switches. And then, of course, there is the asking price, which is a considerable amount of money. But for a small-batch analogue processor,

The controls are all stepped, of course, and inside you can clearly see the resistors that are connected to the switches — and their the values can be customised to the customer’s specifications.

July 2026 / www.soundonsound.com

designed and crafted by people with utmost experience and skill, in a high-wage economy and without any bean-counters attempting to ‘cost-engineer’ the design, this is not only what a product like this costs: it’s what it’s worth.

Verdict I’m fully aware this article reads less like a review and more like an anthem; a celebration. But I mean every word of it: the Vintagetools NT 95 PE is absolutely phenomenal. It’s an equaliser that’s capable of improving literally any signal, and apparently without any drawback. That’s something only a select few EQs can claim. It’s an ultra-precise tool, but one with neither a clinical nor dull character, nor a short-term hit of excitement that wears thin over time. The VT N95 is just ‘right’. The price seems hefty at first, and yes, it’s a shame that many of us can only dream. But this is the type of gear you can pass on to your children and grandchildren. So if you’re looking for the EQ you will work with for decades, without ever looking back, get one. But watch out: If you leave the VT N95 PE unattended in the studio overnight, your other equalisers might well murder it out of sheer envy! £ €4268.10 including VAT and shipping. T Vintagetools +49 6105 2713915 E equipment@vintagetools.de W www.vintagetools.de

ON TE ST

Source Elements Source-Connect 4.1 Remote Recording Software

When it comes to remote Atmos collaboration Source-Connect 4.1 is currently the only game in town. JULIAN RODGERS

ource-Connect, from Source Elements, is a long-established tool for high-quality, low-latency audio connections over the Internet. It’s a staple in post-production, where remote voiceover sessions and attended reviews are routine, but it has a lower profile in music production. In this review we’ll look at version 4.1, which introduces a capability that is unique. Remote collaboration for mix approvals is common in stereo and surround work, but Dolby Atmos introduces practical challenges. While it’s technically possible to share a lossless Atmos ADM file, the large size of ADM files makes iterative remote work slow, while smaller MP4 exports rely on heavy compression. Source-Connect

Source Elements Source-Connect 4.1 From $35 per month pros • First and only metadata-connected remote Atmos workflow. • Accurate spatial rendering for the receiver, independent of their monitoring layout. • Removes need for repeated ADM exports during lossless collaboration. • Stable, high-quality streaming built on the established Source-Connect ecosystem. • Professional workflow features (Auto Restore/Replace, secure transfers, sub-accounts)

cons • Setup complexity for first-time users; requires familiarity with external renderer workflows. • Facility version is expensive.

summary Source-Connect 4.1 adds Dolby Atmos metadata streaming and makes remote mix approvals faster, more accurate, and considerably more practical for professional users.

July 2026 / www.soundonsound.com

4.1 avoids these disadvantages by allowing an Atmos session to be streamed live, meaning that you can send up to 128 channels of audio, along with the associated panning and object metadata, to a remote collaborator. Their system renders the Atmos mix locally in real time, rather than relying on a pre-rendered ADM. The result is a far more agile workflow where approvals can happen interactively. If you’re unfamiliar with Source-Connect, its core purpose is to provide high-quality, real-time audio streaming for professional work. It’s a mainstay in post‑production for voiceover and ADR sessions and is widely used for remote mix approvals. An AAC codec is used up to 7.1 and in high‑channel‑count applications like Dolby Atmos the Opus codec is used. Both are lossy codecs, but are widely regarded as transparent in professional use. The system is designed to be resilient. Source‑Connect is frequently used for remote recording. In recording workflows, any dropouts in the monitored stream are automatically replaced afterwards by transferring the original full-resolution audio in the background. In practice, you monitor a compressed, but still high-quality stream. It’s a clever approach. Encryption and the fact that nothing is stored online are important considerations. If ADMs are shared online they can be disassembled without permission. Source-Connect also offers some thoughtful workflow touches, including a system of sub-accounts that lets users assign temporary licences to collaborators. And, critically for many, no iLok is required.

For those who’ve worked with earlier versions, Source-Connect 4 represented a significant modernisation. The interface was overhauled, but the more important changes were under the hood: automatic networking replaced the old manual configuration, the AAC ELDv2 codec brought low-latency high-quality streaming, multi-channel support was expanded, and the system gained integrated automation for the high-resolution background audio replacement that underpins its resilience. Version 4.1, released in 2026, builds directly on this foundation with a complete Dolby Atmos workflow. Other products can stream large numbers of audio channels, but in Source-Connect 4.1 not only are up to 128 bed and object channels transmitted, but their positional metadata travels with them, allowing the destination renderer to behave exactly as it would in the originating studio. The remote participant isn’t listening to a pre-rendered Atmos mix; they’re rendering it locally, in real time, using their own Dolby Atmos Renderer. That preserves spatial accuracy and monitoring integrity, making the experience at the receiving end identical to the source. The practical implications are that real-time approvals, remote mix checks, and even sessions with multiple reviewers all become possible — something that previously required exchanging large ADM files, reliance on lossy MP4s or attending sessions in person.

New In 4.1 The focus of this review is specifically the remote renderer metadata workflow

introduced in Source-Connect 4.1. To test this properly, you need an Atmos-capable DAW that supports this workflow. At present, Pro Tools is the primary option. DaVinci Resolve is also reportedly compatible, though I didn’t test it. You also need a standalone copy of the Dolby Atmos Renderer. You cannot use Pro Tools’ internal renderer for this workflow. The external renderer setup is more involved: Pro Tools must be configured to use the Dolby Audio Bridge as its playback engine, and linear timecode has to be routed between Pro Tools and the Atmos Renderer to maintain local sync. The specifics of this configuration are beyond the scope of this review, but it’s worth emphasising that anyone who has only ever used the internal renderer will find the external workflow different and less automatic. Naturally, this all assumes you have someone to collaborate with — remote workflows don’t function in isolation. This is where Source-Connect enters the picture. Source-Connect 4.1 is available in Studio and Facility versions, with Facility necessary for surround and immersive work, with a more limited Talent version designed purely for receiving streams. Because I wanted to test in both directions of a two‑way session, only Source-Connect Facility was used at both ends in this test. In practice, audio is routed from Pro Tools via the Dolby Audio Bridge into Source-Connect, which transmits it to the receiver. You must also choose the correct routing to the remote renderer so that object metadata is sent alongside the audio. A sync configuration must be enabled in the Sync tab to ensure both systems behave consistently. Monitoring can vary depending on

your environment: binaural monitoring for headphone users or loudspeaker monitoring for those with Atmos rooms. If this sounds complex, that’s because it is, at least compared to turnkey remote-listening tools. Source-Connect 4 isn’t unmanageable, but it is a professional tool designed for professional use. It requires a clear understanding of the workflow to configure correctly.

Preparing The Test Session I asked my friend New York-based engineer Nathaniel Reichman, a very experienced and Grammy-nominated Atmos specialist, to join me for a session. By contrast, while I’m very experienced with Pro Tools and comfortable with Atmos, the overwhelming majority of my Atmos experience has been with the integrated renderer. My early encounters with the external renderer, back before the internal one existed, were not especially enjoyable, and I hadn’t spent time with it since the internal renderer became available. Before scheduling a test call, I spent some time reacquainting myself with the external renderer workflow. Virtual routing has always had plenty of scope for confusion. In my experience, it’s easy to lose track of signal flow when working with virtual routing. We shared our setup procedure via a Zoom call and we were able to route audio without difficulty, but metadata proved less cooperative. If you’re not familiar with the way the Dolby Atmos Renderer displays object activity, each object has a ring that illuminates when metadata is present. Ours remained stubbornly dark. Another source of uncertainty was the distinction between Source-Connect’s global

settings and its call-specific settings. As long as you remember both exist, it’s manageable, but it’s entirely possible to think you’ve changed something when in fact you haven’t. With everything appearing correct yet metadata still missing, the time-served solution (stop the call and reconnect) got things moving. As soon as we reconnected, metadata flowed as expected and the system behaved exactly as it should. At this stage, the call was set so that I was receiving Nathaniel’s test session. My local renderer handled the incoming audio and metadata correctly, rendering the mix binaurally to my headphones. With that working reliably, we repeated the exercise in the opposite direction: my test session fed into Source-Connect, sent to Nathaniel, and rendered on his 7.1.4 monitoring system. And, predictably enough, once you’ve done something once, the second attempt is far less involved.

Real-world Test Session In actual use, the connection was solid. My home fibre connection is only 100Mb/s, whereas Nathaniel has gigabit, and it became clear that my bandwidth wasn’t comfortable with a full 128-channel stream (and a Zoom call). I had an alternative gigabit-equipped venue available but we thought we’d try with a fast, but not very fast, connection to see how it went. Reducing the send channel count to 64 solved the issue. It was a useful reminder that while the system supports up to 128 channels, many Atmos sessions use far fewer objects in practice, and our test session was no exception. The fact that it took us some time to get everything working on our first attempt reinforces the earlier point: this is a professional tool with a professional workflow, and you do need to understand the moving parts. But once running, it was very dependable. Nathaniel was particularly pleased to have tried it without the pressure of a client waiting on the other end.

How Well Does It Work?

Object tracks are positioned using panning metadata to control the Atmos Renderer.

Source-Connect 4.1 works exactly as intended. It’s solid, it sounds excellent and, based on the long track record of Source-Connect, none of that is particularly surprising. What is significant is that this is the first product to take the step into audio-plus-renderer-metadata streaming for Dolby Atmos. That capability is still unique, so the obvious question is: how useful is it, and who actually needs it?

www.soundonsound.com / July 2026

ON TE ST

S OU R C E E LE M E N T S S OU R C E- CON N ECT 4.1

The blue circles around the objects in the renderer indicate the presence of panning metadata. Linear Time Code keeps metadata in sync with the streamed audio.

Source-Connect and its competing tools have been used for years for remote mix checks and approvals. In Atmos remote approvals often rely on compressed Dolby Atmos MP4s because if a full‑resolution export is required the alternative is a large ADM export. If every revision requires exporting, uploading, downloading and auditioning a new ADM, the whole process becomes slow and disruptive. Metadata-connected streaming sidesteps this completely. It turns Atmos approvals into something much closer to a real-time attended session, where feedback is immediate and iterative changes don’t require generating new files. For commercial mix work, where rapid, repeated adjustments are standard, this alone is a substantial advantage. But it isn’t only about speed or convenience. Because object metadata is transmitted, the renderer at the receiver’s end can reproduce the mix correctly for their monitoring layout, which might be a different channel layout altogether.

Remote Speaker Checks & Practical Use An interesting possibility is the idea of remote speaker checks for headphone-only Atmos creators. With Atmos now available in most mainstream DAWs, including Logic Pro, there are far more people producing Atmos projects binaurally. Experienced Atmos mixers are unanimous: you can create an Atmos mix in headphones, but at some point it must be checked on speakers. For anyone unable to travel to a suitable room, Source-Connect 4.1 offers a pragmatic alternative. While the creator won’t hear the loudspeaker system directly,

July 2026 / www.soundonsound.com

Pricing There are three main ways to get Source-Connect, each with monthly subscription and one-time purchase options. Be aware, though, that the subscriptions also have a $105 setup fee, repayable if your subscription should lapse and you don’t renew within four months, and that the permanent licence requires an additional annual payment. Subscription Buy-in Source-Connect 4 Talent $35 per month + $105 one-time setup fee $650 + $144 per year after first year Source-Connect 4 Studio $105 per month + $105 one-time setup fee $1495 + $275 per year after first year Source-Connect 4 Facility $175 per month + $105 one-time setup fee $2490 + $344 per year after first year

they can participate in a meaningful, real-time review with an engineer in a calibrated space, identifying issues and agreeing on fixes collaboratively rather than outsourcing the decision-making entirely. Another benefit is that Source-Connect 4.1 still supports multiple remote listeners. Multi-endpoint review sessions were possible before, but the key difference now is that each endpoint can be an Atmos-equipped listening environment rendering the mix accurately. This is a first, and it expands the range of remote approval workflows significantly.

Conclusion With v4.1 Source-Connect remains the professional tool for professional workflows that it has always been. It requires deliberate setup and a certain level of technical understanding, and it isn’t aimed at an enthusiast audience. That isn’t to say that any part of the process is especially difficult; it simply assumes familiarity with the concepts involved. Once configured, it is stable, predictable and sounds excellent. As expected for a professional tool, it is not inexpensive. The Facility

version, at nearly $2500, isn’t priced for casual use, but for mix approvals there is a short‑term licence option available at $100 for two days. The workflow it enables won’t be important to everyone, but it will be relevant to an increasing number of people as Atmos continues to spread across music and post-production. Any product that arrives first with a genuinely new capability has added significance, and that’s the case here. The new features don’t solve a universal problem, but for the people who do need this workflow, it removes a great deal of friction. And at the moment, if you want to work this way, this is the only option available. The value of streaming object metadata isn’t simply that it makes the process faster. It makes it better. It preserves the format as intended, allows accurate monitoring at the receiver end, and opens up workflows that weren’t realistically achievable before. That is why this release is significant. £ See ‘Pricing’ Box. W www.source-elements.com

the next evolution of soothe

Since releasing Soothe in 2016, our goal has been to create original and useful tools that help audio engineers focus on what matters. With each plug-in and iteration, we’ve deepened our understanding in DSP and UX. Soothe3 continues that evolution with our most transparent processing and refined user experience yet. Learn more at oeksound.com

ON TE ST

SAM INGLIS

he KSM32 and KSM44 have been mainstays of Shure’s studio microphone range for a quarter of a century. And in that time, the world of affordable studio microphones has changed beyond all recognition. At launch, the Shure models faced limited competition from a handful of European and US manufacturers, with Chinese-made mics just beginning to penetrate Western markets. Today, pressure on prices is far more intense, and the majority of mics in this price bracket are wholly or partly manufactured in the Far East, nearly always using some variant on the ubiquitous Neumann K67 capsule design. This intense competition has, however, highlighted Shure’s unique selling point: their KSM-series mics are wholly designed and manufactured in North America, and the capsules are original Shure designs rather than copies. These are made in a clean-room environment, and tolerances are tight enough that any two examples of the same model can be considered matched. If anything has dated the KSM32 and 44, it has been their styling rather than their performance; there’s an unmistakable whiff of ‘ADAT project studio’ about their

July 2026 / www.soundonsound.com

Shure KSM44MP, KSM40C & KSM32C Capacitor Microphones

Shure’s popular family of studio mics gets a dark reboot — and a new sibling. curved shells and shiny silver finish. From a cosmetic point of view, therefore, a refresh was perhaps timely. And in the process, Shure’s engineers have found ways to improve both mics, leading to the release of the KSM32C and the KSM44MP. Although the uniform styling of the two mics means that the KSM32 is often thought of as a fixed-cardioid version of the multi-pattern KSM44, there are actually deeper differences between the two models. The KSM44 is an externally polarised or ‘true’ capacitor microphone with a one-inch, dual-diaphragm capsule, operating on the standard Braunmühl-Weber principle. By contrast, although the KSM32 is also a side-address mic with a very similar look and feel, it uses

a permanently polarised or back-electret capsule that is intermediate in size between typical ‘large’ and ‘small’ designs. It is consequently more comparable to models like the Neumann U89 and Sony C-80 than to more traditional studio mics. So, with this reinvention, Shure have also introduced a new model called the KSM40C — which really is the fixed-cardioid version of the KSM44 design that some expected the KSM32 to be.

Clean Lines Once again, the models look almost identical, the only obvious differences being that the 32’s body is slightly shorter than that of the other two, and the 44 bears an extra switch for selecting the polar pattern.

All three mics retain the 15dB pad and high-pass filter options that were featured in earlier iterations. However, the metallic sheen and bulbous curves are gone, replaced by a very smart and businesslike black cylindrical body. The robust headbasket is formed from a single piece of metal, into which a pattern of circular holes is cut; behind these is visible a finer mesh that protects the capsule from air blasts, stray particles and other foreign objects. It’s an attractive, no-nonsense design in its own right, and also brings the mics much closer in appearance to existing Shure classics like the SM7B. The XLR socket is, as usual, located at the base of the body, and surrounded by a threaded ring. This can be used to attach either an elastic shockmount or a simple standmount, both of conventional design. The KSM44MP and KSM40C ship with both as standard, along with a pop shield that fixes magnetically to the upper ring on the shockmount; this is a new accessory, which was not available for the older models. The KSM32C is available in either a ‘studio kit’ which includes the shockmount and pop shield, but not the hard mount, or as a slightly cheaper ‘stage kit’ with the standmount and a foam windshield, but no shockmount or pop filter. All three models ship in the same, good-quality padded case, which is large enough to hold the mic without removing it from its shockmount. All in

all, it’s a comprehensive package that ensures almost all recording scenarios are covered.

The Quiet Life Shure publish refreshingly comprehensive specifications and diagrams for all of their mics, and it’s interesting to compare these latest releases with the earlier variants. The previous KSM44A, for example, boasted some astonishing measurements, of which the most eye-catching was a self-noise figure of 4dBA in cardioid. The KSM44MP matches its performance in this respect, but a change in Shure’s measurement process now means the self-noise figure is reported as 5dBA. This is still state of the art, and I know of no other production mic that outperforms it on this front. In other respects, the 44MP measures pretty much exactly like you’d expect a well-designed large-diaphragm studio mic to do. At the low end, its frequency response is not curtailed in any way, and it’s largely flat through the midrange. Above 3kHz or so, it exhibits a presence lift that varies with polar pattern in a way that is typical of dual-diaphragm mics. A 10kHz peak is apparent in cardioid, more prominent in omni, and almost disappears in figure-8, while a broader boost from 3-8 kHz is most apparent in figure-8 and almost absent in omni. The polar pattern is well behaved at all frequencies in

Shure Shure KSM44MP, KSM40C & KSM32C £1059/£849/£709 pros • A well thought-out update that retains the many good qualities of the original KSM44 and KSM32. • The new KSM40C will hit a sweet spot for many studio owners looking for an affordable, high-quality large-diaphragm mic. • Practical cases, mounts and pop shields make for an attractive overall package.

cons • None.

summary Shure’s new KSM family is a suite of high-quality studio mics, with excellent accessories, at a tempting price. What’s not to like?

figure-8, whilst the cardioid setting tends to hypercardioid from about 2kHz upwards. The 44MP is marginally more sensitive than the preceding 44A, generating 32mV/Pa in cardioid and 18mV/Pa in the other patterns. Depending on the load presented by the mic preamp, the 44MP can accept sound pressure levels up to 150dB with the pad engaged for a 1% distortion level. The KSM40C uses a single-sided version of the same capsule. Shure told me that it should sound very similar to the KSM44MP in cardioid mode, but with slightly greater proximity effect. Specifications are similar, All three of these Shure models can cope with huge sound pressure levels.

www.soundonsound.com / July 2026

ON TE ST

S H U R E K S M 4 4 M P, K S M 4 0 C & K S M 3 2 C

but not identical, with self-noise rising to a still-very-impressive 7.5dBA and sensitivity falling to 21mV/Pa. The published frequency response closely resembles that of the KSM44MP in cardioid, but with a slightly more prominent lift from 4 to 7 kHz. The KSM32C has a similar extended low-frequency response and flat midrange, and although there’s a modest peak around 12kHz, the treble response is somewhat flatter than that of the KSM44MP in cardioid. The polar pattern also retains a recognisably cardioid shape over a wider frequency range, as you’d expect given the smaller capsule size. Self-noise is quoted as 11dBA, which is excellent considering the dimensions of the capsule, whilst sensitivity is 16mV/Pa, and 1% distortion is reached at sound pressure levels of up to 155dB depending on the load. All three mics use transformerless electronics, and share the same high-pass filter design. The two options are an 18dB/octave slope turning over at 80Hz, for removal of subsonic noise, and 6dB/octave with the -3dB point at 115Hz, which is designed to compensate for proximity effect.

Fresh Ears I’ve never had the good fortune to use the previous iterations of the KSM32 or KSM44, so was approaching the new mics without preconceptions. To take the KSM44MP first, most of the large-diaphragm mics I encounter could be described as belonging to one of two families, in terms of their subjective sonic character. Mics based around the K67 capsule and its derivatives tend to have a noticeably forward character in the midrange, even when this isn’t apparent from their measured frequency response. This is often paired with a presence boost that can be either exciting or a bit shrill and harsh depending on how far it’s pushed. On the other side of the divide, mics based around AKG’s large-diaphragm capsule designs tend to have a smoother and less prominent midrange, sometimes combined with an airy, sparkly high end. The capsules Shure use in the KSM44MP and KSM40C are edge-terminated, in the AKG tradition, but to my ears they belong sonically in the former camp. In fact, in side-by-side comparisons with my Neumann U87, the 44 often sounded very similar, albeit with an on-axis sound that seemed more consistent across polar patterns. The KSM44MP is also a lot quieter and is much better able to cope with loud sounds than the U87, and is a smidge brighter.

July 2026 / www.soundonsound.com

The 44MP, 40C and 32C Studio Kit include an effective shockmount and magnetically attached pop shield.

Subjectively, where a difference was apparent, I’d say the Shure mic was more analytical and less forgiving; on sources such as vocals, I sometimes missed the richness and the slightly soft quality that the U87’s transformer brings. Either way, the KSM44MP more than holds its own against much more expensive mics. The introduction of the KSM40C is welcome, as there are many buyers out there who want a high-quality large-diaphragm studio mic but don’t need multiple polar patterns. In a direct comparison with the KSM44MP, the additional brightness and bass tip-up were audible, but the core character is the same, and you could easily drop in with one of them onto a track recorded with the other. The most obvious difference I could detect was in the response 180 degrees off axis, where the KSM40C seemed to offer quite a bit more rejection at low frequencies. The KSM32C, meanwhile, is more of a sonic tabula rasa. It’s hard to say that it really has much of a sound of its own: it’s the sort of mic that stays out of the way and allows the source to sound like itself. I liked it a lot on drum overheads, for example, where it gave a balanced picture of the whole kit, with plenty of low end, without emphasising gritty resonances in the drums or bringing out harshness in the cymbals. I’d be more than happy to use the KSM32C to record almost anything, from loud guitar cabs to delicate winds.

Clean Sweep Overall, I don’t get the impression that these new variants sound significantly different to their predecessors, and I don’t think Shure

ALTERNATIVES Most of the major mic manufacturers offer competitors to the KSM44MP and KSM40C. Obvious examples include AKG’s C414 and C214, Neumann’s TLM103, Lewitt’s LCT540S, AIM Audio’s Inspire and Essence and Audio-Technica’s AT4047 and AT4050. So-called ‘medium‑diaphragm’ mics like the KSM32C are less common, but include the Sony C-80 and Neumann TLM102.

mean them to. In terms of the bits that matter, they got pretty much everything right first time around, and the level of attention to detail in the design is impressive. For instance, the two high-pass filter settings are both genuinely useful in different contexts, and make clear that Shure’s engineers have thought hard about the real-world applications for these mics. At the same time, there are genuine improvements here that go beyond the mere look and feel of the mics. To take a small example, as well as being a neat addition in itself, the inclusion of the magnetic pop filter has allowed Shure to dispense with a third layer of mesh inside the headbasket. The removal of anything that comes between the source and the capsule usually benefits the sound. It’s a fool who tries to predict the future, but if Shure’s aim is to ensure that the KSM range remains relevant for another quarter of a century, I’d say they’ve succeeded handsomely. £ KSM44MP £1059; KSM40C

£849; KSM32C £709 (Studio Kit) or £639 (Stage Kit). Prices include VAT. T Shure UK +44 (0)1992 703058 W www.shure.com

ON TE ST

With the launch of their new ThumpSub Go, Mackie’s off-grid PA goes full‑range! PAUL WHITE

oday we have a huge choice of ‘box’ or column‑style compact PA systems, but when it comes to battery operation, the choice is rather more limited — and all the more so if you also need a decent amount of power. The Mackie Thump Go and its companion subwoofer, the ThumpSub Go, form a robust but portable PA system for small to medium-sized venues such as pubs, bars and smaller outdoor events. Both units can run on mains power or removable lithium batteries and, when paired with the also battery-powerable Mackie ProFX 10 Go mixer, form a complete PA setup that’s just as usable off-grid as it is indoors. We reviewed the ProFX 10 Go separately in June 2025 (www.soundonsound.com/ reviews/mackie-profx10-go), so please refer to that for more detail on the mixer side. In addition to the speakers and mixer, Mackie offer their SPM400 speaker poles for mounting the tops above the

Portable PA System

Mackie Thump Go & ThumpSub Go subs, plus rechargeable batteries for each speaker (these are an optional extra when buying the speakers individually). Additional spare batteries and custom cases are also available. The system is suitable for mobile DJs, live events, corporate presentations, outdoor ceremonies and garden parties — with or without mains power.

Thump Day Battery operation inevitably means the maximum SPL may be slightly lower than a similarly priced mains-only system, but the Thump Go top and

The Thump Go and ThumpSub Go can both run for several hours on batteries. The subwoofer can even house two batteries, and supports hot-swapping.

July 2026 / www.soundonsound.com

sub combination still delivers a solid performance. A single Thump Go can be used alone as a compact full-range speaker with a frequency response of 57Hz to 20kHz (-3dB), or paired with the ThumpSub Go for extended low reach (-3dB at 39Hz). It is powered by a 200W (peak) Class-D amplifier driving an eight-inch woofer and a one-inch compression driver in a ported, moulded cabinet that includes a pole-mount socket. Its drivers are protected by a perforated metal grille, while the cabinet’s angled side allows it to be used as a floor wedge. The cabinet measures 457 x 230 x 285mm and weighs 8kg. Maximum SPL is quoted at around 115dB — which, while lower than some similarly sized speakers, is still subjectively loud. Battery life is impressive at up to 12 hours, though this is reduced at higher output levels. Connecting to the mains via the IEC inlet charges the batteries. Four voicing options (Music, Speech, Monitor and Sub) are selectable with a push-button, though the setting isn’t retained after power-down — it defaults to Music. An Outdoor mode optimises performance where there are no reflective boundaries. There’s also effective automatic feedback suppression

that deploys narrow notch filters, which can be fixed when ringing out during soundcheck or left to ‘free roam’ in use. A Ducking mode lowers channel two when signal is present on channel one, which is useful for DJ applications. The speaker includes a basic two-channel digital mixer with XLR/TRS combo line inputs. Input 1 can be set to accept passive mic signals (phantom power is not available), and input 2 has an additional mini‑jack input. A thru output passes a mono sum on to another speaker. Control is also available via the Thump Connect 2 app over Bluetooth, which also supports audio streaming and stereo linking between speakers. The app provides access to EQ, voicing options, mode selection, level setting and save/recall of presets. As expected, Bluetooth introduces slight latency and works most reliably over short distances. Sonically, the Thump Go sounds clean and full for vocals and instruments, though pairing it with the sub improves performance significantly for DJ use or bass-heavy material.

and indicate overload protection. Bluetooth streaming is supported, even when using non-Bluetooth tops, and the front‑panel LED can be switched off if the speakers need to remain visually non-intrusive.

In Use

The Thump Go features a two-channel mixer, with the first input accommodating passive mic or line signals and the second able to accept a Bluetooth stream.

Sub Thumping The ThumpSub Go subwoofer measures 582 x 249 x 531mm and houses a 400W (peak) Class-D amplifier driving dual eight-inch drivers in a ported enclosure. It weighs just under 19kg with one GB20 8880mAh battery fitted, and features a carry handle. Battery recharge time is around three hours. Maximum SPL is quoted at around 123dB, with a frequency response of 39Hz to 100Hz (-3dB). Batteries charge from the IEC input or an optional external charger. Two battery slots are fitted, providing up to six hours from one battery or up to 12 hours with two, with

The ThumpSub Go can handle crossover duties using its own DSP, or can pass a full-range signal on for filtering further down the line.

hot-swapping supported. A single battery is provided with the sub as standard and, as with the Thump Go, running at a higher output reduces the battery run time. The sub’s low-profile, squarish design allows horizontal or vertical placement, though pole-mounting requires it to be laid flat for stability. Vertical use requires a separate stand for the top. Stereo XLR/jack line input pairs feed the sub, then the signal is passed from the output XLRs to the tops, which should be set to their Sub voicing, and the sub set to full‑range output. Alternatively, the sub can be set to high-pass filter mode and the voicing on the top speaker set to Music. Stereo and mono output modes are available from button switches, as is polarity inversion. Status LEDs show when the sub is powered

The system is easy to transport and quick to set up. In testing at a typical pub gig for a rock trio, a single stack handled vocals, percussion mics and DI feeds from acoustic and electric guitars. Despite limited placement options due to the layout of the bar, we achieved strong, clean levels, with the feedback suppression proving effective — though it works only for the tops, so won’t deal with feedback in the very low frequency range. Switching between mains and battery operation had no noticeable impact on sound quality or level, and while high output levels inevitably reduce battery life, a single battery was enough for a typical two-set performance. Bluetooth streaming is useful for DJs and for playing music between sets, while remote control from the app is handy if adjustments need to be made from out front during performance. Note that as with most PA systems adopting a similar format, what you hear when standing behind them is mainly bass and lower midrange plus reflected sound, so adding a small monitor to the system really helps. Since we were playing in a venue with mains power, we used a Mackie SRM150, but an additional Thump Go would work perfectly and maintain the mains-free option. If you need a full-range PA system that can run on batteries as well as mains power, the All Systems Go makes a lot of sense. And if you only need a mono setup, or if extreme bass isn’t essential, two tops can be run from a single sub. Battery capability adds cost compared with a mains-only PA, and slightly limits maximum SPL, but for off-grid use, this system offers a practical balance of performance, flexibility and portability.

summary If you need serious sound without mains power, the Thump Go and ThumpSub Go combo is the system to beat. Throw in Mackie’s ProFX 10 Go mixer and you have a complete off-grid sound solution.

£ Thump Go £425, ThumpSub Go £559. Prices are per speaker, including VAT.

T Source Distribution +44 (0)20 8962 5080 E sales@sourcedistribution.co.uk W www.sourcedistribution.co.uk W www.mackie.com

www.soundonsound.com / July 2026

ON TE ST

Balcony Left Little Sir Rekko FET Compressor Plug-in

We check out a new FET compressor emulation based on producer Greg Wells’ favourite vintage unit. M AT T H O U G H TO N

ou might not have heard of Balcony Left or Little Sir Rekko before, but you’ve almost certainly heard of some of the names behind this plug-in: A-list producer, engineer, musician and... (well, a lot of things!) Greg Wells is behind the project, and invited me to review it, and the presets have been created by Greg and a handful of other reputable and respected engineers. Little Sir Rekko runs on Mac and Windows machines in VST3, AU or AAX plug-in hosts, and is activated on your machine using your Balcony Left online account: you get three ‘seats’ per licence, and seats can be deactivated to allow transfer to different machines.

Overview Plainly, this plug-in sets out to model one of the world’s best-known compressors but, specifically, this is an emulation of Wells’ own vintage hardware unit. Perhaps inevitably in 2026, its creation involved machine-learning techniques but, as I understand it, these were used as part of a hybrid process to measure and analyse the hardware’s behaviour, largely to identify aspects of the specific unit that might be overlooked with traditional component-level modelling, and the resulting plug-in was carefully tweaked until extensive listening tests confirmed that Wells was happy with the sound and behaviour. I have to say: what a result it is! I’ve used countless 1176 and 1178 emulations over the years, and this has to be up there with the best I’ve heard. A statement that, by the way, applies equally to the sound, the compression action, and the ease of operation.

July 2026 / www.soundonsound.com

In terms of controls, it’s pretty conventional. A skeuomorphic GUI facilitates the usual 1176/1178 operational approach: an input gain knob sets the signal level relative to an internal fixed threshold, while an output level knob determines the level feeding whatever’s next in the chain. Smaller attack and release knobs adhere to the UREI/Universal Audio paradigm of clockwise for faster/more aggressive, rather than ‘more milliseconds’. The expected four-button stack selects the ratio (4, 8, 12 and 20:1), while a dedicated button engages the famous ‘all buttons in’ mode, and another simulates pressing in just the 4 and 8:1 buttons. Nearby, another nice touch is the Off button, which bypasses the gain reduction, turning the plug-in into a line amp, or ‘saturator’ if you like, though there’s also an actual bypass on the right. The drag-to-scale GUI is dominated by a virtual moving-coil meter that can display average input or output levels or, of course, gain reduction. It has a more generous range than most (-24 to +12 dB), reflecting the reality of how engineers often actually use these devices. To its right are a couple more features that nod in the direction of modernity: a Mix knob to dial in parallel compression is joined by a ±12dB Headroom control. Acting rather like a dedicated threshold control, I found this really handy for quickly fine-tuning presets, or when using Little Sir Rekko in line-amp mode. Other modern features behind the scenes include oversampling and an anti-aliasing process that, together, might explain why this plug-in seems to sound so smooth and effortless — hardware-like, if you like — even when you’re spanking the bejeezus out of things. Despite that, and the

fact that there’s very little latency, this plug-in didn’t seem particularly CPU-hungry. For all the knobs, a Cmd- or Ctrl-drag facilitates finer control, while an Opt/Alt-click or double-click resets them to the 12 o’clock position, with the exception of the Mix knob, which resets to 100 percent wet. All good, though I occasionally wished for a modifier key to return individual controls to the initial settings of a preset, rather than ‘12 o’clock’.

Sound & Action The first thing to say about this plug-in is that if you’ve ever used an 1176-inspired compressor, there’s zero learning curve: you can cut straight to evaluating what it does to your audio. Or perhaps I should say ‘appreciating’ what it does, because there are a couple of things about this plug-in that set it apart from similar tools. First, and most important, is the sound. Yes, in broad terms, it’s similar to other decent FET compressor emulations. During my tests, I substituted it into projects where I’d used the Universal Audio or Pulsar Audio equivalents, and there’s a definite family resemblance. But it’s not the same. There’s a depth and richness to the sound from Little Sir Rekko that I found enticing, and notably

Balcony Left Little Sir Rekko $118 pros

• First and foremost, a good-sounding analogue modelling compressor • Classy, hardware-like performance, even at ‘abuse’ settings • Great collection of presets.

cons

• Side-chain filter might be nice?

summary

A superb analogue-modelling compressor that sounds right, with an uncomplicated controls set that gives you just enough ‘extra’ compared with the hardware.

W E N a lovely, full-sounding bottom end and flattering smoothness in the highs; not a softening one that compromises definition, but rather a pleasing sheen that just sounds classy. There’s plenty of range in the attack and release too, and the curves feel ‘right’; very easy to dial in. Whether all that’s down to the hardware muse or the modelling approach, I don’t know — and if I’m honest, as an end user, I don’t really care because either way it does nice things. Speaking of which, the other thing that really impressed was the preset collection. Now, I don’t need presets to operate this thing and neither, I imagine, do you. But taken together, these offerings form a great learning resource for those just embarking on their engineering journey. Each one has a meaningful name (eg. “Cymbal Sustainer” or “Bring The VocalForward”), and they’re all categorised by the artist (Wells himself, Joe Chiccarelli, Josh Gudwin, Chris Tabron, Mike Avenaim and Sage Skolfield), sometimes with sub‑categories for instrument/source material type. It’s easy to get a feel for why these engineers might reach for this style of compressor, or the different things you can achieve on the same material using compression alone. You might need to tweak the time constants to suit different material and tempos, of course, but I often I found that the Headroom knob was all I needed to get a feel for what the preset was designed to deliver. Was there anything lacking? Not really. That said, I felt that a few more ‘mod cons’ might be nice, and top of my list would be some sort of side-chain HPF or EQ facility. You don’t need that to replicate the hardware, of course, but it should be trivially easy to add this sort of thing in a plug-in, to extend the versatility without intruding on ease of use.

Major Mojo? There are some other very usable FET compressor emulations out there that cost rather less, so the full asking price means that Little Sir Rekko may not be an impulse buy for every hobbyist. But you can ‘rent to own’ monthly for a pocket-change figure, and there’s a free 15-day demo too — so you have the opportunity before purchasing to hear for yourself what earned Little Rekko his knighthood. I can’t vouch for the precision of the emulation or the ‘mojo’ the marketing material claims the modelling processes identified as unique in Wells’ particular unit, simply because I’ve never used it. But what I can tell you is that this is an utterly delightful‑sounding plug-in, and one that conveys what I think are the most important aspects of analogue hardware: you’ll find ‘warmth’ here when you want it, but also an effortless, aristocratic sort of class when you don’t. Importantly, although the sound and action of this compressor are musically satisfying in the sort of applications a gear designer might have in mind, lots of plug-ins do ‘warmth’ and ‘musical’ in those circumstances. Yet very few seem to exhibit this hardware-like ‘grace under fire’ — a performance that seemingly isn’t compromised, even as you push the gain reduction or saturation well into ‘abuse’ territory. It means you tend to think less about how you’re operating Little Sir Rekko than about all the creative and musical decisions that are what recording and mixing is all about!

USB Solo mk2 from

High spec MIDI to CV converter with both USB and 5 pin DIN MIDI V/Oct & Hz/V – Gate & S-trig – 3 aux outputs – clock & DIN Sync Can be used as a USB MIDI Interface (while still converting to CV) All of the features of the Pro Solo mk3 but with the addition of USB See website for full information

SPLIT, MERGE, EXTEND OR CONVERT All the MIDI you need, right where you need it

Thru-5, 12 & 25

Merge-4 & 8

Split a single MIDI source into 5, 12 or 25 identical copies

Combine the data from 4 or 8 MIDI sources into a single output

LNDX & LNDR

MIDI USB Host mk3

Extend MIDI up to 1000 metres LNDX (shown) has etherCON connector

MIDI IN & OUT on DIN for your USB-only keyboards, controllers & synths – without using a computer

£ $117.60 including VAT. W www.littlesirrekko.com

WE MAKE MIDI WORK FOR YOU Contact the MIDI Specialists www.soundonsound.com / July 2026

kenton.co.uk

INSIDE TRACK

Gabe Simon • Noah Kahan The Great Divide Rediscovering the human has helped producer Gabe Simon to enjoy colossal hit albums with Noah Kahan and Dermot Kennedy. PAUL TINGEN

abe Simon’s production and writing career began in EDM and pop. In 2015, he co-wrote Dua Lipa’s song ‘Begging’; this was followed by writing credits for Maroon 5, Hailee Steinfeld, Anderson.Paak, Rag’n’Bone Man and Lana Del Rey. However, Simon’s big breakthrough came with Noah Kahan’s third solo album, the largely acoustic Stick Season (2022). Kahan’s first two albums had not been particularly successful, but TikTok and a performance of the title track by Olivia Rodrigo on BBC Radio 1 propelled ‘Stick Season’ and its parent album to the top of charts around the world. This success also gave Simon’s career a big push, and his writing, multi-instrumentalist and production credits since then include Koe Wetzel, James Bay, Halsey, Jonas Brothers and more. He also enjoyed another huge hit album with Dermot Kennedy, The Weight Of The Woods.

Going Without The Flow Photos: David O’Donohue

July 2026 / www.soundonsound.com

Raised in Simpsonville, South Carolina, Gabe Simon began his musical career as an 11-year old, playing brass. “I did that all the way up to high school. I also did a lot of choral work, and had resigned

The Great Divide was a much more collaborative affair than Noah Kahan’s previous album Stick Season. This photo shows, left to right, Kahan, Simon, Carrie K and Noah Levine, outside at Guildford Sound.

myself to the idea that I was never going to be a cool kid. Then during high school my buddies and I would sit around playing crappy little guitars on crappy guitar amps, and we had a lot of fun. We jammed in our garage, where we learned to play Radiohead, Chili Peppers and Weezer songs. I started exploring other instruments as well, and taught myself bass and piano. Then I went to Nashville for college and started a group called Kopecky Family Band.” The Kopecky Family Band released two albums for ATO Records, Kids Raising Kids (2010) and Drugs For The Modern Age (2012). Simon meanwhile obtained degrees in music business and entrepreneurship at Nashville’s Belmont University in 2011, where he developed an aversion to the raison d’être of SOS. “I took a class in audio engineering, and was like, ‘This is the worst! I do not want to follow signal flow!’ I liked creating music and guiding people’s musical visions, but I didn’t want anything to do with the technical side of recording music.” The birth of Simon’s oldest daughter in 2016 changed his perspective. “When I had my first daughter, I decided that I did not want to be on the road

any more. I wanted to be home. I’m a curious person, and I think it’s fun to explore new sounds, so I started playing unfamiliar instruments and I went more deeply into producing. My other daughter was born in 2019, and having kids made me the producer that I am today. Before, I had the luxury of having the time to try everything. But when you have kids, you don’t, so you get really good at being discerning and understanding what’s a waste of time and a meaningful use of time. I went from second guessing myself all the time, to being more efficient and confident. “I remain very grateful for my early experiences, because for example my choral work made me appreciate the voice as the most powerful and most unique instrument that we have. As a result, both the Dermot and Noah albums have tons of backing vocals. And there’s a lot of choir on Dermot’s record. We brought in the community choir that he grew up with in college, and recorded it with them walking around a SoundField MkV Ambisonic microphone, creating chords. We did that for every song on his record, to create human dynamics, which adds a whole new layer of emotion.

“I feel that the human aspect gets downplayed in popular music lately, with people opting for Splice effects and sonic perfection, instead of, for example, going into the chorus using a choral bed that goes ‘ooaaahh’. It does not only create some unique sonics, but also provides this immense humanity that listeners don’t even consciously know is there, but that gives them a euphoric feeling that makes them throw back their heads and sing along. It harks back to learning to play guitar in that garage with my friends. All I want is to be free as a kid to make music, and particularly when I produce, I want others to have that same experience.”

Pedal Power As a songwriter, Simon was regularly working with “producers staring at a laptop the whole time”, and being a guitar player in a band involved knowing his way around pedals. Slowly, but surely, music tech was creeping up on him. “By the time of the pandemic something in my brain clicked. I was at home a lot, and thought, ‘I don’t have any outboard, but I have all these bizarre

www.soundonsound.com / July 2026

INSIDE TRACK

GABE SIMON • NOAH K AHAN

Drums were tracked using a more minimal setup than is the norm in Nashville.

go to a farmhouse in Only, Tennessee, and we’re riding dirt bikes and going fishing every day, and with the pressure off we wrote and recorded ‘Deny, Deny, Deny’, ‘We Go Way Back’, ‘Haircut’, and ‘All Them Horses’. I later went with Noah back to Long Pond, and we wrote ‘Downfall’, ‘Lighthouse’ and ‘A Few of Their Own’. Justin Vernon added some beautiful parts and sounds. By the summer of 2025, we were back in Gold Pacific and pulled the record together. It became our sanctuary in Nashville.”

In The Machine “When we’re writing we like to go straight for takes that can be final. So we brought tons of gear to the farmhouse, like a pair of ATC monitors, outboard racks, and great microphones. We threw blankets all over the walls to make things deader and stuff like that. Noah sounds great on my vintage Neumann FET 47, which

July 2026 / www.soundonsound.com

I bought in 2023. I was blown away by it, especially when I was making James Bay’s Changes All The Time [2024]. It was one of the few mics that can handle James’ big dynamics without breaking up. I find that a lot of males do really well on that mic. Dermot sounds amazing on it, as does Myles Smith. “Sam had the idea to use the Microtech Gefell CMV 563 at Guilford Sound, and Noah sounded so intimate and close, we ended up buying that one. I now use it all the time. I have a studio right behind my house, and Noah loves cutting vocals here, so we did a few of the vocals for The Great Divide here. On Stick Season, we used a [Telefunken] ELA M 251 primarily for his lead and ran it through a Neve 1073 mic pre, and a compressor like a Fairchild 660, 670 or a UREI 1176 Rev F. On The Great Divide, we ended up messing around with quite a few of the Heiserman mics like their H251 and Type

19. I love the Type 19. Looks weird, sounds sick. We were more adventurous with mic pres too, used the Tree Audio Branch tube channel strip a bit, some older API 212s and the Chandler LTD1. It’s so funny to think that I once had no interest in this, but once I found all the flavours and the colours these things give, I started nerding out, and now I have way too much shit, like tons of crazy synths, a 1960s LA‑2A, my 1970s 1176, Chandler Limited REDD.47 tube mic pre, APIs, Neves, spring tanks and more mics and guitar pedals then anyone needs. “I love it now and no longer fight the machine. I work in Logic for writing and programming, but when we track in a studio, we use Pro Tools, because it’s more functional for that process. I had the immense privilege of working with great engineers during the sessions for Noah — Maddie Harmon, Phillip Smith, Sam Westhoff, Alberto Sewald, Bella Blasko —

INSIDE TRACK

GABE SIMON • NOAH K AHAN

and I’m very involved in the spirit of how it goes down. For example, for the band recordings, we liked quite a live sound. I love atmosphere. I love to hear the room as much as I love to hear close. I’m a big fan of Atmos mixing, and it’s really fun to have the atmosphere, because you can do a lot more with it. “At the same time, I don’t go for the big Nashville mic setup, particularly with drums, where they’re running 24 mics on a kit. I don’t know if it’s laziness or just the confusion I experience when I look at all these drums mics, but I’m in the Glyn Johns spirit, with just four or so close drum mics and making them count. We did many of the band recordings at Gold Pacific, and it took us a while to get the best drum sound there. In the end we went for ribbons, and an AKG C24 or a pair of Neumann M49s for the room sound. We did the same kind of room vibe at Guilford Sound in Vermont but that room we kinda set up in the same way we did Stick Season, which felt really cathartic. “I was the main bassist, and I like flatwound strings. I played a Fender Mustang and a Reverend, which is like a Fender Coronado, with a hollow body and a thumpier sound. The Mustang is the main tone on the album, and it has a nice drive to it, so we can get dirtier if needed. I played through an Ampeg Venture VB-112 amp, recording that with a FET 47, and DI going via an ACME Audio Motown WB-3 or an A Designs Reddi box. “For electric guitars, on a lot of songs I was playing through a Fender Super Six 6x10 in the main room, and a smaller amp in an iso booth like a Princeton. We’d do [Neumann] KM86s or something on those. On acoustics I’m a sucker for the Neumann KM56, which we also used on Noah’s guitar for Stick Season. The banjo is a bizarre instrument that sounds very different in every room but we loved the Coles 4038 on that. The Heiserman Type 19 was great for tons of random stuff like hurdy-gurdy or pump organs. We also used an RCA 44 for a darker tone, for instruments like the mandolin or mandocello. We used the same vintage 44 on all backing vocals, so they did not get in the way of Noah’s lead vocals and could keep that high end for him.”

Human Factors The band recordings were often recorded without a timing reference. “Noah likes the ebb and flow, and it’s more fun.

July 2026 / www.soundonsound.com

‘All Them Horses’, ‘Haircut’, ‘23’ and ‘Dashboard’ don’t have click. We’re all in the room, and we’re playing off each other. We did a little bit of timing fixing in post-production, but only if it was really bad, and it was a core part that was oscillating too much. When Noah was performing the song with the band, we used the Sennheiser MD 441 on his voice, for example on ‘The Great Divide’. That song is one take, without tuning on him. ‘All Them Horses’ also was one take, with the whole band playing in one room together at the farmhouse at Only. That’s a quieter tune so we could sneak the CMV on his voice. “‘Doors’ was a tough song to record. We tracked it with Noah’s live band, and it was fun and had energy, but it wasn’t right. Then we tried a hybrid, with overdubs, and it still did not feel right. We tried a different ending that pivoted somewhere else, and we were still banging our heads against the wall as to what to do. Then we went up to Guildford Sound in Vermont, and with Sam engineering, we just decided to forget everything that had been done, and went on the floor as a band, and played it like we had never played it before. It was the first time that we felt it clicked. “When Carrie K, Noah and I play together, sometimes with Sam or Dylan Jones [Noah’s touring multi-instrumentalist], we feel like a unit, reading each other’s flow, and finding ways to reinforce each other. Carrie does a fill, and I move my bass part, and Noah does a hammer-on at the same time. That’s stuff you can’t do when you’re overthinking things, it only happens when you’re jamming. Also, you can hear my bass on the room mics, and that was a really badass sound. We also felt that it needed a guitar solo on ‘Doors’ so I had a Danelectro baritone, a Fulltone Tube Tape Echo, straight into the 6x10, just screaming loud. I was hitting the guitar so hard, it was going out of tune, but every time I hit, it sounded cooler and cooler. By the time I had finished, we realised we had cracked the code on that song.”

Pulling Together By the end of the process, Simon and his engineers Maddie Harmon and Alberto Sewald were pulling all recordings from different places together, including those produced by Aaron Dessner and recorded by Bella Blasko. This process involved them unifying tracks recorded

Several different mics were used on Noah Kahan’s voice, including Guildford Sound’s Telefunken ELA M 250.

in countless different ways to create a cohesive album sound, integrating the many nature recordings that had been done at Long Pond and at the farm house in Only, Tennessee, and Simon and crew adding finishing touches. “One of the big things with Aaron’s stuff that I really enjoyed was bringing in Carrie K on drums. She really tied together a lot of the tones and rhythmic energy that we had. If things got too cinematic, we tried to dial it back to make sure that it still felt organic and not too thick. We wanted to make sure that we weren’t doing things that Noah would not be able to do live. Like he plays live with a fiddle player, but does not have a full string section. “We had to find a balance and did use a lot of nature sounds, sometimes

recorded using an iPhone. For ‘Headed North’ we had mics in a field, including some Microtech Gefell UM70s, to capture the transition from birds and bugs to silence at the end of the day. I did some programming here and there in Logic, usually creating some weird ambience. Like taking an Ebow on an electric guitar, running it through amps, chopping the recording up, and then running it through some other stuff. On ‘Doors’ I did that with a hurdy-gurdy. I used Soundtoys’ AlterBoy to process that and chop it up to turn it into this aggressive drone. Or in ‘We Go Way Back’ I did some Noah vocal layering to bring in some weird soundscapes. “We also finalised the reverbs on the album. We usually had real big plate sounds. I like the warmth and the little

bit of buzz that you get from plates, so we were using the EMT 140 a lot. I used this weird OTO BAM verb on a lot of the guitars and for weird atmosphere stuff. Also, Guilford and Gold Pacific have really cool chambers. I remember engineer Phillip Smith printing the chambers in real time when tracking drums, so there’d be twice as many tracks. These sessions got huge! I got Phillip in for a few songs because I worked on an album for a British band called Picture Parlour, which is almost Black Sabbath-like. I loved Phillip’s rock tones, and wanted to have those on Noah’s album. As a producer you’re trying to get the right cocktail of tones and flavours, and engineers play a huge part in that. “Ryan Hewitt mixed the record, he always adds something amazing. He

mixed Stick Season as well. He and I work together a lot. We started off with our cornerstone tracks, like ‘The Great Divide’ and ‘American Cars’, and then went to the opposite, the more intimate tracks like ‘Paid Time Off’ and ‘23’. In that last two weeks of finishing tracks, Maddie and I were shipping off songs to him every day! By the time we reached Thanksgiving 2025, we had most of the record mixed. The first week of December, Noah and I went over to Ted Jensen at Sterling Sound for mastering, and then we sat together to create the track listing for the whole thing. After almost two years of making the record, after all the stress, the ups and downs, the studios, the locations, we looked at each other and said, ‘That’s pretty damn good!’”

www.soundonsound.com / July 2026

TECHNIQUE

Most listeners discover music on phones, where attention spans are short! JUSTIN SHERIFF

istening habits have changed considerably over the last decade or so, and in a couple of important ways. First, perhaps most obviously, a great deal of music is now first encountered not through studio monitors, nice hi-fi systems, or even car stereos, but rather through a pair of consumer earbuds (whether wired or wireless), or perhaps even a phone’s tiny built-in speaker. Some estimates put headphone/earbud listening as high as 80 percent, maybe more. Second, and maybe more importantly, whatever the playback system we’re listening on, many of us now encounter new music more passively — for example through short-form videos or preview clips, or in a scroll-based feed. These are all contexts in which our attention span is limited, and very easily diverted. For a significant portion of listeners, then, their first impression of our music is formed within a few seconds, often in a fragmented context, and often when listening on devices with limited low-frequency extension and reduced dynamic range. While none of this should change the main goals of mixing (plenty of people still enjoy listening on full-range systems, at home, on good headphones, in cars or in the cinema, and they won’t appreciate compromises), it does change the conditions under which our mixes are most often judged. Consequently, some of the mix decisions that might, in traditional listening environments, have felt secondary — questions of midrange balance, transient clarity, or arrangement density, for instance — could now be far more important in determining whether a mix will be perceived by most listeners as clear and engaging, or unfocused and lacking impact.

Consider Your Priorities Obviously, of those two outcomes, ‘clear and engaging’ is our goal! So what, exactly, might we consider changing about how we approach our mixes, to make that the more likely first impression? Well, the aim certainly isn’t to optimise mixes exclusively for phones or consumer earbuds, and what I said just above bears repeating: if we simply cater for the lowest common denominator, we lose the opportunity to impress more discerning listeners.

July 2026 / www.soundonsound.com

Mixing for Mobile Practical Strategies For Short-form & Small-speaker Listening

Rather, we need to aim to achieve two things: 1 . Ensure our mixes translate consistently across a wider range of playback systems, including those with a restricted bandwidth and limited headroom; and 2. Address the problem of passive listening and short attention spans. In order to achieve (1) we must understand how different playback systems reshape what the listener actually hears, and how we can shape our mixes to take account of that. For example, small speakers and many compact earbuds tend to emphasise the midrange, and limit both sub-bass reproduction and spatial depth. To achieve (2) we have to be very deliberate about how we choose to present the core elements in the mix. Ear-grabbing

intros have always been a thing, but for listeners with shorter attention windows the initial presentation of a mix now carries more weight than ever; if the core elements are not immediately clear, listeners are more likely to casually scroll on. Remember, they’re probably not actively trying to listen; they’re encountering your music while doing something else. So it pays to adopt a more intentional approach to prioritisation. Which elements must be clearly understood within the first few seconds? Which parts of the arrangement are essential? Which simply contribute density or texture, without improving clarity?

Midrange: The Main Event Let’s now dive a little deeper into how we might cater for the most common playback systems. On small speakers and many earbuds, the midrange effectively carries

Photo: cottonbro (Creative Commons)

the mix. To put some numbers on that, a frequency range of approximately 500Hz to 4kHz dominates our perception, and for vocals the 1-4 kHz region is particularly important for intelligibility; it’s where the sounds that allow us to distinguish between consonants tend to lie. Too little energy in these regions, and the mix feels distant or muddy, and will lack clarity. Too much, and it becomes harsh and exhausting, especially on earbuds that may be ‘voiced’ to put that range ‘forward’. The challenge, though, is not simply to boost the midrange content, or even judge the right overall level, but rather to manage the overlap between the various sources that sit in this part of the spectrum. Vocals,

that feels present and intelligible without becoming abrasive. (For more detailed advice on how you might approach that side of things, check out Matt Houghton’s SOS September 2022 article on achieving better separation when mixing: https://sosm.ag/ mixing-for-better-separation). Of course, not every instrument naturally sits in the midrange. If the airy highs of a mix can’t be conveyed very well on a phone or earbuds, that’s probably not a problem — nice to hear but not essential to the music. But what about the low end? Small speakers can’t reproduce true sub-bass and even where they can deliver low frequencies, if there’s lots of energy in that region it will very quickly eat up the headroom and result

“Ear-grabbing intros have always been a thing, but listeners’ shorter attention windows mean that the initial presentation of a mix now carries more weight than ever.” guitars, synths, pianos, snare presence and more already jostle for space here, in just a few octaves, and simply panning them for better separation won’t work, as the playback system may well be mono or at least have little by way of stereo imaging (a phone speaker, common kitchen or bathroom radios, and many portable Bluetooth speakers, for example). Careful EQ and level balancing and conscious arrangement decisions to ensure good separation tend to be more effective than aggressive boosting and cutting, or even more sophisticated mix techniques like ducking with side-chain compression. When all this is done well, the result is a mix

in distortion at normal listening volumes — which, of course, reduces clarity and impact. This doesn’t mean the low end becomes irrelevant. Far from it, in fact. But it does make it more important that we cater for listeners’ perception of low-end power than aim for its direct reproduction. Naturally, this requires tight control of low frequencies, such as through filtering, balance decisions and dynamics processing. But it also means considering how best we might reinforce the low-frequency elements so they’re noticeable on a range of small playback systems. For example, we might use distortion or a dedicated processor to add harmonics to the sound, to shift some

For a mix to have clarity and impact, it’s important not to squeeze the life out of every signal’s transients. Parallel compression can often be an effective means of increasing average levels without too much collateral damage.

of its energy higher up the spectrum. Or we might make creative artistic decisions when mixing, such as choosing to lean more on the beater click of a kick drum than its extreme low end, so that it makes the desired rhythmic contribution while giving bass instruments more space lower down the spectrum. Such tactics can allow the listener to perceive bass energy and groove, even when the fundamental frequency is not fully reproduced.

Reality Checks: Secondary Monitoring Reference-quality nearfield monitoring remains essential for mixing. Yet, such speakers no longer represent the majority of people’s listening experience: a mix that feels balanced in a treated room

Most of your audience will first encounter your music while scrolling through social media, where attention spans are short — so you have very little time to persuade them to listen.

www.soundonsound.com / July 2026

TECHNIQUE

MIXING FOR MOBILE

can behave very differently much of the transients/note when played back on a phone onsets are allowed through, speaker or consumer earbuds. for example. Or, for more It might seem tempting to assertive processing, perhaps overcompensate for this, look to parallel compression, or to ignore the problem. to allow some of the transient But constantly checking details ‘back through’. Either mixes on a phone can lead way, consciously working to to reactive decisions that maintain a sense of movement degrade the overall balance, and contrast as you compress while relying exclusively on will often translate better than a full‑range system won’t tell simply maximising the density you what you need to know in of a sound. terms of translation. Arrangement decisions An effective tactic is to have always played a major use small-speaker playback role in how well a mix only as a secondary, periodic translates, but I believe they’re reference — one that you now even more critical. turn to only at certain key Dense arrangements that feel stages of the mix, such as after exciting on full-range systems establishing an initial rough can often sound congested balance, perhaps when you’re when reproduced on smaller doing any radical reshaping of speakers. This is largely due the relationship between the to ‘masking’: as bandwidth instruments that occupy the and dynamic range are low end of the mix with EQ, reduced, elements compete and again before you finalise more in the parts of the the mix. In other words, you’re Masking problems can become more evident on smaller playback systems, where spectrum where they overlap the midrange takes on even greater importance. Some tools, such as this FabFilter EQ, using it to check for problems in the frequency and time can display a multi-channel ‘heat map’ that makes it more obvious where sounds are you may not immediately domains. The loudest sound competing in the same sonic territory. notice on a good, full-range dominates while the quieter system, but your nearfield monitors remain ones retreat, but it’s not just that some reduces punch and intelligibility. In most your primary mix reference. sounds get ‘buried’: the balance becomes cases, you’ll find that clarity will help Another thing you can use your inconsistent, as different instruments your mixes translate far better over such secondary small-speaker system for is come in and out of the mix. What felt rich devices than sheer average levels will. to assess loudness and dynamics. Small and layered in the studio can quickly Hook Me Now! playback systems naturally compress and collapse into something much less well distort, and pushing a mix too hard often defined in real-world playback. Cleaner In the sort of short-form contexts arrangements — not simpler music, but I discussed at the start of this article, more focused music — often translate the opening moments of a track carry more effectively because they side-step disproportionate weight. Some of that’s this problem. about the music itself, and about the So, to recap, leaving space, both arrangement (as I’ll discuss later), but there frequency-wise and arrangement-wise, is are also some important considerations a practical and powerful tool for improving when recording and mixing. clarity. If instruments seem to overlap, Transient information plays a key then, ask yourself why. Could you maybe role in perceived impact and clarity. separate them by putting them in different Overzealous compression and limiting octaves? Or by muting one of them at can smooth and soften these elements certain points, to leave the other the to the point where a mix comes across space it needs? Or by separating them in as ‘flat’, particularly on smaller playback time somehow? systems, and over streaming services that employ data compression. It follows The Mobile Loudness Myth that working to preserve or shape the transients can help elements come Small speakers often struggle for volume, across more clearly, without you having so it’s understandable that some people Limiting for loudness is often to rely on excessive level. assume that louder mixes will perform counterproductive, but it can also be problematic if This doesn’t mean you should avoid better on them. In practice, though, the your master is too quiet. Most streaming services won’t turn the music up past the point at which the compression. It simply means using opposite tends to be true. Yes, limiting highest peak hits 0dB TP, no matter their ‘LUFS it in a deliberate way: paying closer the peaks does buy you the headroom target’, while those that offer ‘loud modes’ may attention to the attack time and how to apply gain and increase the average apply limiting that robs the mix of clarity and impact.

July 2026 / www.soundonsound.com

TECHNIQUE

MIXING FOR MOBILE

level. But excessive limiting often reduces clarity and punch — partly because of the effect on transients, as I discussed above. Also, almost all streaming platforms now apply loudness normalisation on playback by default (few people change that setting), and that negates any advantage in pushing the mix/mastering levels aggressively. To put it bluntly, then, a heavily limited mix may well sound smaller and less dynamic when streamed on mobile devices. So what might you focus on instead of loudness? In a word, dynamics. If you maintain some dynamic contrast in a mix, it tends to sound more impactful across a range of playback systems. There are still some ‘gotchas’ to be aware of, though. For instance, if a mix is too dynamic, with very high peaks, most loudness normalisation systems won’t turn it up past 0dBFS True Peak, and those that do will apply limiting, causing distortion. So there’s a ‘sweet spot’, and if you want to explore this further and put some figures on it, I suggest you check out our interview with mastering engineer Ian Shepherd in SOS April 2023: https:// sosm.ag/ian-shepherd-loudness-dynamics. The key point to take away from this is that on small speakers, perceived ‘punch’ tends to owe more to transient clarity and a good balance than to sheer level.

Lost In Translation? Focusing on small-speaker translation is important today, but it does come with some potential trade-offs, and it’s certainly possible to take those too far. In much the same way that focusing exclusively on perfecting a mix for mono playback might compromise your ability to deliver a spacious stereo mix, decisions that improve clarity on phones — such as emphasising midrange content, reducing

low-end weight, or tightening dynamics — can, in some cases, make a mix feel less expansive or less engaging on full-range systems and higher-quality headphones. For example, if you focus the mix excessively on the midrange to improve translation on a phone speaker, the mix can become fatiguing over longer listening sessions on earbuds or headphones. Similarly, overly restraining the low-end energy to improve definition on small playback systems can reduce the sense of depth and impact when the track is replayed on larger systems that have greater low-frequency extension. So the goal is absolutely not to optimise the mix for one playback system at the expense of others. It’s simply to ensure you strike an informed balance — to create a mix that remains clear and communicative on limited systems, while still conveying depth, width, and dynamic interest when heard on more capable setups. Translation is ultimately about making sure that every element that’s musically/rhythmically important always comes across, not about absolute perfection in any single context.

Towards A Practical Workflow In practice, what this all boils down to is that you need very deliberately to check mix decisions across different listening systems/environments, but also remain alert to when adjustments made to solve one problem introduce other ones. For example, when checking on a small speaker, maybe you can’t hear a bass guitar that’s important to the groove. So you apply parallel saturation to give it more harmonics. Great: now you can hear it! Yet because the midrange is a small, crowded playing field, you then notice that the

Saturation plug-ins can be an effective way to make bass instruments heard on mobile devices, since the harmonics they add can be heard more easily on smaller speakers — but you must remember to listen out for the impact on other instruments!

July 2026 / www.soundonsound.com

bass’s added harmonics cause the electric guitars to come across less well. If they do, then you’ll need to make decisions about what to do differently — whether to the bass, or to the guitars. Such is mixing! As I explained above, your reference nearfield monitors remain hugely important. You certainly don’t want to be swapping speakers all the time and getting lost in rabbit holes like that bass/ guitar one. If you’re constantly chasing your tail like that, then you’ll lose sight of what gets listeners tapping their feet and singing along. The key to success is to establish a consistent workflow, in which you check translation at specific points in the mix process. This will help you avoid overreacting to individual playback systems. One such approach is to establish the mix primarily using your nearfield monitors. Then perform a focused check on a phone or earbuds to identify any major translation issues. Take notes, then make targeted adjustments to address them, while listening over your speakers. Then do a final check on your phone/ earbuds to confirm that the core elements now sound clear and balanced. Getting sound from your DAW to check on your phone can be an obstacle. You could just bounce your mix and load it on to your phone, and there are advantages to that approach. For instance, you can take your phone anywhere and listen in different, perhaps noisier environments than your studio. But for a quick reference check, there are now plug-ins that let you stream your mix to your phone in real time. Mix to Mobile by Sound on Digital, for example, sends lossless audio over your local Wi-Fi to a dedicated app on your phone, so you can tweak the mix in your DAW and hear the results on your phone straight away. Another tactic is to set up a DAW channel with a plug-in that simulates consumer playback devices: Sonarworks SoundID Reference has a Translation Check feature with over 20 device simulations, and Audified MixChecker Pro takes things further, with over 60 models, including specific phones and tablets, plus background noise and device distortion. So you can get a feel for how your mix holds up in the real world without leaving the studio. It needn’t be just about the playback device itself, though. Simply listening at lower volumes occasionally can be very revealing, because many of the

your workflow specific times when you want to turn things down, before returning to your standard listening level.

Awareness Without Compromise

Sound on Digital’s Mix to Mobile: one of several services to stream directly from your DAW to your phone.

issues that become apparent on small speakers — imbalances, masking, lack of clarity — become more obvious to us when monitoring levels are lower. You should be establishing a reference monitoring level in

your studio anyway (for more on that check out Hugh Robjohns’ SOS May 2014 guide to establishing reference monitoring levels: https://sosm.ag/reference-monitoring) so, as with your phone check, try to build into

Whether we like it or not, mobile devices are increasingly the first ‘place’ where listeners encounter our music. Treating that reality as an afterthought risks you delivering mixes that fail to convince in the environments where they’re most often heard. But, as I’ve discussed above, prioritising mobile translation need not mean compromise. Mixing has always been about translation. It’s simply that we need to stay aware of the range of systems that must be given consideration, and of how contemporary listening behaviours shape the way in which listeners experience music. Full-range systems and higher-quality headphones still reveal important details, and mixes should hold up under those conditions as well — follow the advice I’ve given above, and your listeners should be able to enjoy the best of both worlds!

www.soundonsound.com / July 2026

ON TE ST

ALM/Busy Circuits Quaid Gigaslope Eurorack Module

‘A

hands-on, menu-free, jam-friendly complex modulator’ is how ALM/Busy Circuits chalk up the Quaid Gigaslope, which takes the design principles of the preceding Quaid Megaslope and runs with it, well, a very long way. The Megaslope offered an elegant five-stage function generator that nodded toward the West Coast and could be used as a complex envelope, a very editable LFO or a diminutive step sequencer. The Gigaslope is similarly visually intuitive and tactile, only it takes the number of stages to a formidable 13 and moreover is endowed with four separate output channels to use any which way, separate or convergent. Don’t worry, all will become clear. Now, to state the obvious: the Quaid Gigaslope is massive. At 52HP wide it’s definitely up there among the largest modules of its — or any — type you’re likely to find; most in this size category are all-in-one modules of some sort, but the Gigaslope is not offering an all-in-one module: it is, so to speak, a one-for-all module, since it caters to more or less anything you can imagine

July 2026 / www.soundonsound.com

a multi‑stage function generator catering to. And in many ways this justifies its size, since it can all but assume the role of several modules at once. At the Quaid Gigaslope’s foundational level are sliders for adjusting its 13 stages’ voltage values, along with a Time and Curve knob for each. The Time knobs are essentially slew limiters to adjust how long any given stage takes to reach its maximum value; while Curve adjusts the shape of that movement through logarithmic, linear and exponential shapes. There are three primary slope modes that dictate how any sequence of stages behaves: Envelope mode, Loop mode and Step mode. Envelope mode expects a single trigger or gate to set a voltage function in motion, which will run from start to finish. Next is the free-running Loop mode, which does exactly that, and lastly Step mode advances the sequence by one stage with each trigger and automatically restarts after the last step, just like a traditional sequencer. There are four trigger inputs and four outputs for the four slopes. You might notice that these are not referred to as ‘tracks’, and this points toward something rather unique that I very much like about the Quaid Gigaslope. If you’re an owner of a module like the Intellijel Metropolix, you’ll be familiar with the workflow of a single row of sliders offering different variants

of a transport that all operate in parallel. The Quaid Gigaslope, meanwhile, offers a single track of sliders to essentially be cut up and used in sections. It’s a rather brilliant workflow, and incredibly intuitive. In practice, I quickly created a patch with Slope 1 occupying the first six steps as a sequence for an oscillator; Slope 2 acting as a three-stage envelope across steps 7, 8 and 9, working the filter; and Slope 3 creating a fairly complex LFO courtesy of steps 10, 11, 12 and 13. Sections can also overlap if desired, or even be placed directly on top of each other. It’s possible to derive quite a range of responses from a single set of voltage values, since the different modes elicit different behaviours. In Step mode, for instance, a stage will progress to the next step with a new trigger regardless of whether or not its Time value has reached its full duration, whereas in other modes the stage will wait to reach this before progressing. Some slopes can be stuttering and sliding around irregularly, while others, convergent or not, might be driving forward with regimented consistency. It’s also worth mentioning that the Sustain button allows any given stage to sustain indefinitely while a gate remains high; in other words if you’re of the type who favours ADSR or ASR envelopes, you’re more than looked after here. There’s a Quantiser button alongside a neat little set of chromatic keyboard-esque LEDs to indicate which scale (of the many on offer) is currently selected, and once again different settings can be applied to different slopes operating across the same voltage values. As such, in this particular patch I had Slope 4 as another four-stage cycling modulation source, using the voltage values of steps 8, 9, 10 and 11 — that is, straddling the values of two other slopes. Brilliant stuff. Thirteen sliders might seem a curious number since many such circuits in synthesis operate

in multiples of four, of course. But I like it, mainly because it encourages multiple uses of the Quaid Gigaslope’s stages; if you’re in a 4/4 time signature, you might default to four‑ or eight-stage patterns, which as a minimum leaves a remainder of five stages begging to be allocated elsewhere. Either ALM simply wanted to avoid the Quaid Gigaslope clearing 60HP, or it’s an intentional move to resist the more comfortable 16-step layout; whichever it is, in my opinion it works. Aside from a slew of useful additions — for instance end-of-cycle and end-of-slope trigger outputs, a manual

Otter Mods OTTeR Eurorack Module

he first time you play with the Oscillating Temporal Tessellation Repeater (OTTeR), it feels comfortingly intuitive. It has eight sliders that you can quickly flick into a sequence, a clock input, and a CV output. Set it running and you’ve got a range of voltages that you can use for modulation, or if you have a quantiser to hand, turn them into notes. It’s simple, easy, and the LEDs are a nicely subversive shade of green. However, there’s something odd going on. The steps are not necessarily running in a regular sequence; instead, they seem to dance about the place, shuffling here and there, repeating before setting off on another twisted cycle. As you adjust the three knobs, A, B and C, those green LEDs chase each other into all sorts of repeating patterns. I would love to say that I know what’s going to happen with a specific combination of A, B and C, but for the life of me, I can’t. On one turn, you get steps 1 and 2 alternating with 7 and 8, on another, it wanders through steps 1-4 a couple of times before doing the same to 5-8. Another turn halves the speed, another doubles it, and every now and again, you hit upon a classic

Reset button and a button to toggle the output range between a 0-5 V unipolar and bipolar ±5V — the Quaid Gigaslope’s final flourish comes in the form of four assignable CV inputs. Simply hold an input’s corresponding button and move one or more parameters to map it. It adds a wondrous dose of complexity and keeps things moving very nicely, and of course it can receive CV from one of the module’s own envelopes, loops or sequences if desired. It’s also compatible with ALM’s AXON-1 or AXON-2 expanders if you need even more CV inputs.

sequence of eight different steps in a row. It’s disconcerting but also brilliant in how it’s reworking the eight chosen pitches into different versions of itself. Having no idea what’s going to happen next doesn’t seem to take anything away from the enjoyment of the melodies and remixes that pour out of it. I made the mistake of asking Otter Mods about the purpose and intention of the knobs. What you don’t really have to understand (I think) is that the three knobs control the clock division of three bit-counters. The bit-counters are fed into a multiplexer where the binary sum dictates the order of the steps. For example, set A to 1, B to 2 and C to 3, and you get all eight steps in sequence because A is running at the clock/2, B is clock/4, and C is clock/8. Easy right? Well, no, not exactly, but if you understand a bit about binary, it all starts to add up. All you really need to know is that through different combinations, you get different patterns. It reminds me of the Triadex Muse, which generates patterns in a similarly baffling way. However, with the OTTeR, you have complete control over the pitch of those steps, which is something the Muse would never allow. There are other subtle moments of brilliance in the OTTeR that make it feel even more elaborate. You have a ‘Half CV’ output that gives you a scaled-down version of the ‘Full CV’ output, meaning you could be running two related sequences or use one for notes and another for modulation. A Transpose input lets you push the sequence up in pitch, while an Inhibit input makes it skip or trip into steps in response to a gate input. This produces some rather lovely rhythms. The three knobs have CV inputs that, when fully clockwise, allow external clocks to mess with the binary multiplexing. Once you start getting into unrelated clock divisions, the patterns become far less predictable. If blitzing things with audio rate modulation is your thing, then the OTTeR

As the creators of the iconic Pamela’s Workout module series, it almost goes without saying that ALM/Busy Circuits know what they’re doing when it comes to multi‑purpose voltage generators. In some ways, you might even say the Quaid Gigaslope feels like a kind of exploded, menu-less variant of Pam. It’s big, sure, with a price tag to match. But make no mistake: from its build quality to its excellent functionality and flexibility, this is premium hardware. Expect devotees. William Stokes £

£699.

www.busycircuits.com

will reward you with some rather great textures that you can shape with the step sliders. What I’m really enjoying about the OTTeR is how it starts off in a familiar place, and then, as you play, it runs off in all sorts of interesting directions. I like how the notes are mine, but the patterns are the OTTeR’s, and I’m just there making suggestions rather than being in charge. So, while the OTTeR may seem on first glance like a regular slider-based sequencer, it’s so much more interesting, more playful, and unexpected. Paired with the right quantiser, it’s a superbly musical experience. Robin Vincent £

$275.

www.ottermods.com

www.soundonsound.com / July 2026

ON TE ST

MQA Labs Inspira & Endura Impulse Response Correction Plug-ins Digital audio can sound wonderful today, and while it’s not quite technically possible to achieve ‘perfection’, this clever pair of plug-ins gets us that much closer! HUGH ROBJOHNS

e live in an imperfect world, and that applies to audio recording and reproduction technologies just as much as anything else. Of course, modern digital audio formats are immeasurably better across myriad technical parameters than all preceding analogue systems, but some critical listeners remain adamant that “even the best digital recordings fail to measure up to our best analogue experiences”. I’m not entirely convinced of that claim, personally, but agree nonetheless that current digital audio systems are not perfect — and probably never can be. Perhaps a better question is whether any of digital audio’s remaining imperfections are actually audible or practically significant. The words I quoted above come from MQA Labs’ white paper on the technology employed in their pair of closely related plug-ins that, as far as I’m aware, are unique in addressing a very specific artefact of digital recording that is never present in analogue recordings. Before looking at the plug-ins in detail, then, I think it would be helpful if we first consider the issue they are designed to correct...

Pre-ringing A noted imperfection of most current digital audio systems involves their time-domain response and, in particular, the pre-ringing artefacts typically introduced by the linear-phase anti-alias and reconstruction filtering employed in pretty much all A-D and D-A converters. Imagine a brief signal transient or audio impulse passing through a regular

July 2026 / www.soundonsound.com

Inspira on its ‘zero’ (top) and extended pre-ring (bottom) settings.

analogue filter. What comes out of the filter is the original impulse followed by a decaying resonance or ‘ringing’ that’s caused by the way energy is stored and released in all filtering circuits. (This natural phenomenon can be found in all forms of filtering, whether electronic, mechanical, acoustic or biological.) In the real world, an effect (the filter resonance or ringing) follows an event

(the transient impulse). The effect cannot pre-empt the event as that would go against the progressive unidirectional nature of time. Professor Hawking wrote a very good book about that! However, things can be different in the realm of digital signal processing, and the design of linear-phase digital filters inherently results in pre-ringing before the impulse, along with post-ringing after. I don’t have

space here to explain why that happens, so I’ll ask you to just accept it on trust. I explored this topic in some depth within an article about the MQA music format back in 2016 (https://sosm.ag/mqa-format) and if you’d like to know more about pre-ringing and how it can be corrected, I’d suggest reading that. Although clearly an ‘unnatural’ artefact in the purest sense of the term, filter pre-ringing is, I’d argue, inaudible to average listeners in typical listening situations. Digital audio has been mainstream for over 30 years now, on radio and TV, the streaming music player, or in the cinema or nightclub, and I’m willing to bet you’ve never once heard your friends, siblings, parents or grandparents complaining about audible pre-ringing artefacts! Yet, there are some good academic arguments that the ear/brain is sensitive to the timing information of audio signals, and that these pre-ringing artefacts interfere with that timing information in a negative way, degrading some aspects of audio recordings, including stereo imaging precision and the perceived realism of acoustic spaces — things that experienced critical listeners are,

inconvenience, aliasing distortion can be a problem with minimum-phase filters because they’re not as steep as typical linear-phase designs.

Apodisation So, how can pre-ringing be corrected? Using a well-established mathematical technique called ‘apodisation’, which needs to be applied only once to correct all pre-ringing caused by any number of linear-phase filters from A-D and D-A converters in the entire signal path. Again, if you’d like to know more, this is a concept I explained in detail in the 2016 article I mentioned above. Suffice it to say here that the technique is used in a wide range of engineering disciplines besides digital audio, including radio-astronomy, optical photography, digital image processing, mass spectrometry, NMR spectroscopy... and many more besides. In fact, since apodisation is relatively easy to implement I’m at something of a loss as to why so few manufacturers seem willing to implement it... and that’s why I’m so pleased that MQA Labs have built it into their Inspira and Endura plug-ins — the subjects of this review — with simple controls to allow users to

“There are some good academic arguments that the ear/brain is sensitive to the timing information of audio signals, and that pre-ringing artefacts interfere with that timing information in a negative way.” perhaps, more likely to notice. It makes sense, then, that if pre-ringing can be corrected it is probably worth doing so, to remove a potentially deleterious digital-only artefact and make the digital signal path sound more like that of an analogue system. Interestingly, the extent of FIR filter (pre- and post-) ringing is related to the sample rate, with higher sample rates having substantially shorter pre-ringing artefacts. This may be why some critical listeners prefer higher-sample‑rate recordings. Similarly, some prefer the sound of minimum-phase filtering — which has no pre-ringing artefacts but has a non-linear phase response — that many current converters offer as a user-selected option. Unfortunately, very high sample rates generate huge amounts of data to be stored and processed. Along with that

optimise the processing parameters. In essence, Inspira is intended to correct digital audio pre-ringing artefacts during tracking of individual sources, as well as to optimise the low-level resolution and audible noise floor through a bespoke noise-shaped dithering process. Multiple plug-in instances can be used on any number of input sources. Endura does much the same thing but is optimised for use at the end of the mixing or mastering process, sitting across a stereo master output bus when mixing or mastering. (The plug-in does not yet support multi‑channel busses beyond stereo). Both plug-ins are available in VST3, AU and AAX formats, and are compatible with DAWs running on Windows 10 upwards and macOS 10.13 upwards (Intel or Apple Silicon). Authorisation is only via an iLok USB key (iLok 2 or

MQA Labs Inspira & Endura From $99 pros • Successfully removes pre-ringing artefacts. • Unique implementation of apodisation in the plug-in format. • Simple and elegant user controls and interface. • Noise-shaped dithering options. • Endura includes output file validation and independent file storage options.

cons • None.

summary A unique implementation of apodisation to remove pre-ringing artefacts from A-D and D-A conversion in the digital signal path, supplemented with well-designed noise-shaped dithering options.

later). The plug-ins have been tested with Cubase 14, Pro Tools 2024, FL Studio 24, Ableton Live 12, Studio One 7, Reaper 7, Pyramix 14, and Logic Pro 11... and I also used it within Adobe Audition 2026 without issue.

Inspira The Inspira plug-in’s user interface is divided into two primary areas. The left section concerns the apodisation correction with an Align control to adjust the amount of correction being applied. Settings are numbered from -6 to +10, and a graphic above the control indicates the intended amount of pre- and post-ringing correction being applied. A control value of zero applies no correction, while moving clockwise to +4 represents the theoretical ideal correction for most modern A-D converters — removing all pre-ringing without extending the post-ringing. Moving the Align control while music is playing through the plug-in has an audible effect, and the optimal setting is found only by ear. I can best describe finding the optimal setting as the position where the most natural and realistic sound character of the source is found, as if the instrument and its acoustic environment suddenly snap into focus! However, the changes can be very subtle, and I found it easier to hear the effect on solo instruments with a distinct background acoustic than with very dry sources or complex stereo mixes. Turning the Align control to negative values increases the amount

www.soundonsound.com / July 2026

ON TE ST

MQA L ABS INSPIRA & ENDURA

of pre-ringing which, technically, makes pre-ringing worse and tends to make the source sound more unnatural... but that may be a welcome effect in some situations! It’s certainly an interesting thing to hear from an ear-training point of view. The right section of Inspira’s user interface is occupied by a large FFT spectral display showing the input signal’s overall frequency response moment by moment, and it’s switchable between linear and logarithmic frequency scales. Below the signal line is a representation of the dither noise floor, and this can be altered by introducing noise-shaped dithering. There are six preset noise-shaping options selected by virtual buttons below the frequency response plot. The available presets are labelled Flat, Air 1-3, and Psy 1-2. Flat is basic TPDF dither, which sounds a lot like pink noise or tape hiss. The three Air options move progressively more dither noise energy into higher spectral regions, to which our ears and brains are less sensitive, while the two psychoacoustic options are designed to keep noise energy away from the most sensitive regions of the ear’s response using more complex filtering arrangements. A Depth control adjusts the amount of noise-shaped dither applied and, again, is best adjusted by ear to find the optimal setting for the specific material being recorded. The scale ranges from 00.00 up to 15.00, and the user manual explains that the number relates to bits, counting up from the least-significant bit. So, 0 would be dithering at the 24-bit level, and 11 would be dithering at the 13-bit level. In practice, the Air options tend to make

the material sound a little brighter, and the psychoacoustic options a little darker, but it’s quite a subtle effect and very dependent on the spectral content of the source material.

Mastering As I mentioned above, Endura is intended for use in the (stereo) output bus of a DAW’s mixer during mixing or mastering and, like all dithering plug-ins, it should be the very last process in the signal path after the master fader. Again, the apodisation process is adjusted with an Align control covering the same range as Inspira and with the same notional pre‑/post-ringing graphic, spanning -6 to +10 (with zero applying no apodisation). The noise-shaped dithering facilities are similar too, with the same spectral FFT display in the centre of the user interface showing the input signal and dithered noise floor. This time, though, there are seven preset options, comprising Flat, Air 1-2, Psy 1-3, and Auto. So, Endura gains an additional Psychoacoustic option but loses one Air mode, and all five noise-shaping curves are generally a little more assertive compared with the Inspira options. The seventh option is labelled Auto, and relates to an Auto-Shaper function with a learn facility that assesses the audio content as it plays. Learning can be stopped at any time, after which the algorithm’s recommended noise-shaping profile is made available through the Auto button, allowing easy comparison with the stock noise-shaping and Flat options. A Reset button clears the learned dither shape. A third section, on the right of the user interface, provides two stereo pairs of

Endura, configured in its ideal (left) and extended post-ring (right) settings.

July 2026 / www.soundonsound.com

vertical bar-graph meters showing the input and output signal levels, and below these is an additional button labelled ‘Record and Export’. This feature provides options to save the processed audio as a file completely independently of the standard function within the DAW itself. Boxes allow the user to enter a file name and folder destination, and select either WAV or FLAC formats. As you’d expect, the plug-in can be used like any standard plug-in in the master bus, with the DAW saving the resulting audio as a file either in real time, or as an offline render in the usual way. However, the possibility exists, depending on the DAW and how it is set up, that the saved output might not be a bit-perfect copy of the plug-in’s output. Potential reasons include the DAW applying its own native dithering at the output, or the master fader not being at 0dB, or some other plug-in existing in the signal path affecting the audio after the Endura plug-in. MQA Labs have built in a neat facility to test for and, if necessary, resolve such issues, starting with an option box to ‘Write Assurance Metadata’. This is bespoke metadata that’s passed along with the audio data to allow the file integrity to be verified as a bit-perfect transfer... or not. In most cases it will be, of course, but this option, together with the associated Verify button, provides confirmation and peace of mind. If offline rendering or real-time recording through the DAW isn’t bit-perfect, the output from the plug-in can be saved directly using its own Record mode, generating a completely independent audio output file that is guaranteed to be bit-perfect. Using this

function results in two files, of course — one from the DAW, and one from the plug-in. Again, the plug-in’s verification tool can be used to check the accuracy of the files.

Inspira-tional? The efficacy of apodisation in removing pre-ringing artefacts created by linear-phase filtering in any chain of A-D and D-A converters is well established, and the benefits of optimal noise-shaped dithering are equally well proven. So, judged on a purely technical level, these plug-ins are very worthy additions, and could prove particularly interesting in a mastering context. Inspira could be used on a stereo output bus too, if desired. The marketing claim that these plug-ins “restore that analogue sound” is obviously very enticing. I listen to both analogue and digital recordings all the time, on very high-quality audio systems, and I wouldn’t say one format is definitively better than the other in terms of naturalness or spatial accuracy — I’ve heard many digital recordings that have immense depth and realism, and some analogue LPs that have failed to impress, so perhaps there’s more to it than just the presence or absence of pre-ringing? But they do sound different. I’m very pleased to report that these plug-ins are successful in removing the pre-ringing artefacts that just don’t exist in pure-analogue formats. I can definitely hear tonal and spatial differences as I adjust the controls. Moreover, the

Endura’s record and export options.

apodisation and dithering tools provided within both Inspira and Endura are well designed and very easy to use, and I’m not aware of any other apodisation plug-ins being available at the current time. All of this makes these offerings from MQA Labs a unique and valuable technical resource. I should probably mention that, initially, I was mildly hesitant to actively recommend them, purely because I felt that the price when launched was high, and that many users would therefore not perceive a good cost:benefit ratio. But MQA addressed shortly before my review

Venos Stereo Bus Compressor

was published, by reducing the asking price significantly. Regardless of whether or not your ears appreicate that restored analogue depth and naturalness when using these Inspira and Endura, I’m very pleased that MQA Labs have brought apodisation to the party in these wonderfully elegant plug-ins. I definitely recommend checking them out too, since both are available on a free, seven-day trial. £ Inspira $99, Endura $199. E info@mqalabs.com W https://mqalabs.com

NEW!

Adding size, depth and impact to a stereo signal has never been easier. With SPL 120V technology this sound is crowned to king of dynamics.

www.soundonsound.com / July 2026

ON TE ST

Softube Flow Studio Control Surface

Softube bring hands-on control to their Flow mixing and mastering concept. JOHN WALDEN

oftube’s impressive Flow Mixing Suite was the subject of an SOS review in the March 2026 issue. The subscribe-to-own plug-in collection provides the user with a sophisticated collection of high-end plug-ins spanning EQ, dynamics, saturation, ambience, modulation, guitar‑amp modelling and more, capable of handling pretty much any type of mixing task. And, while you can deep-dive into any of the individual plug-ins, a mothership plug-in lets you load preconfigured plug-in signal chains — a Flow preset — and, via an elegant macro control system, quickly adjust the key parameters within the signal-chain without getting distracted by the deeper control options. Softube have now released the Flow Studio hardware controller, designed to give the user hands-on control of these Flows, making what is already a streamlined workflow even more intuitive. The Flow Studio will work with either the Flow Mixing Suite or Flow Mastering Suite plug-in collections but, as it also includes perpetual licences for 25 of the plug-ins from the Mixing Suite, you can make full use of the Flow Studio hardware without needing a subscription. So, if you like to keep moving when mixing and, initially at least, avoid getting lost in

July 2026 / www.soundonsound.com

too many details, might Softube’s Flow Studio keep your mix workflow flowing?

It’s Flow Time Softube’s Flow Mixing Suite and Flow Mastering Suite provide some serious plug-in firepower and the plug-ins within these collections can be used individually within your DAW. However, both suites also share a concept; Flow presets. These are essentially preconfigured signal-chains build by combining individual plug-ins and aimed at a specific mixing or mastering task. These Flow presets are loaded into a dedicated ‘mothership’ plug-in — either the Flow Mixing Suite plug-in or the Flow Mastering Suite plug-in — allowing the user to access all of the individual plug-ins within the Flow’s signal chain from within a single plug-in UI. It should also be noted that, currently at least, the mothership plug-in is a Softube-only environment; any favourite plug-ins from other developers can’t be used within the Flow concept. However, the additional (and, in use, rather impressive) twist is that the mothership plug-in features a set of eight macro controls. Each macro control is preconfigured to target one (or more) key parameters within the loaded signal chain. With just the twist of a macro control or two, you can therefore adjust the most important parameters within the Flow to quickly shape the sound of your source audio. As a means

of getting your key mix elements from zero to hero, this can be a very effective and efficient process.

Multi-level Macros In essence, the Flow Studio’s main purpose is to serve as a dedicated hardware controller for that macro-based Flow concept. Most obviously, it does this by providing the user with hands-on control over the macro elements contained within the Flow plug-in via four high-quality rotary encoders (with the Shift button letting you toggle between macros 1-4 and 5-8). However, it also brings a further twist of its own via the larger rotary control — the Omni knob — as this serves as a master macro control for all the other macro controls. As you rotate this, whatever the processing intention built into a specific Flow preset might be, one control lets you quickly dial in as much, or as little, of it as your audio needs. Under the hood, these two levels of macro control are configured much like any sophisticated parameter modulation system, defining the direction, depth and range of parameter change that the full travel of a Macro might exert on the target parameter(s). And, while the Flow Studio ships with a collection of Flow presets to get you started, a Designer view within the Flow Mixing Suite or Flow Master

Softube Flow Studio £339 pros

• Hardware access to the Flow macro control system for fast and efficient mixing. • Quality hardware and bundled plug-ins provide good value. • Subscription to Flow Mixing Suite not required (but very tempting).

cons

• Softube-only environment; third-party plug-ins not currently supported. • Perhaps not an obvious choice for those that like to deep-dive all the details of their plug-in signal-chain.

summary

The Flow Studio provides hands-on access to Softube’s macro-based track processing concept, delivering sophisticated tonal shaping to any audio source with considerable style and a minimum of fuss.

Suite plug-in lets you build your own presets based around any of the Softube plug-ins you have access to.

What’s In The Box? On first impression, Flow Studio maintains the high standard Softube have set with their Console 1 Fader and Channel controllers. Measuring a compact 144 x 217 x 43mm, it contains an impressive high-resolution screen dominating the upper area, five touch-sensitive rotary knobs — each with a ring of LED lights around it to provide positional feedback for the current control setting — arranged in the centre, and an array of nine buttons, many of which offer dual functions, laid out in the lower part of the surface. It feels well built and the rotary controls are very smooth in use. A high-quality 2m USB-C cable is included for connection to your host computer. Optional mounting stands available for the Console 1 Fader and Channel hardware also fit the Flow Studio if you prefer a somewhat steeper pitch to the control surface. As mentioned earlier, the Flow Studio also includes perpetual licences for a collection of plug-ins and the Flow Mixing Suite ‘mothership’ plug-in. As shown in the ‘Provided Plug-ins’ box, they include compressors, EQs, reverb, delay, tape, saturation, guitar/bass amp/effects and vocal tuning options. No additional plug-in subscription is therefore required. During the review period, I initially uninstalled other Softube plug-ins I normally have on my system so I could spend time with just this selection of plug-ins. I did, however,

subsequently reinstall the full complement of Flow Mixing Suite plug-ins to see how the two products — Flow Studio and Flow Mixing Suite — work together.

Modus Operandi Flow Studio offers three modes of operation; DAW Mode, Flow Mode and Detail Mode, and you can move between these via the three buttons located bottom-right. Each of these brings up mode-specific options on the hardware’s internal display. In DAW Mode, Flow Studio lets you access a compact selection of controls within the currently selected track of your DAW project. These include the track’s fader (with the Omni knob), pan, up to three send levels, and solo and mute buttons. The display provides both VU and stereo LED input/output level meters for the selected track and, usefully, at the bottom of the display, compact text labels indicate what track parameter is linked to each of the rotary knobs. This functionality is DAW‑dependent but, in Cubase, for example, it required no configuration and did not require an instance of the Flow Mixing Suite mothership plug-in to be installed on the selected track; it just worked. Softube have details on their website explaining the exact steps required for a number of the more popular DAWs. OK, so Flow Studio’s DAW Mode is perhaps not going to compete with a dedicated DAW control surface, but if this is your first hardware controller, for the sake of a quick press of the DAW Mode button, it’s a nice additional option to have, especially as all the rotary controls provide such a smooth and accurate response. Once you have inserted an instance of the Flow Mixing Suite mothership plug-in

onto your selected track and loaded a suitable Flow preset into it (or created your own), both Flow Mode and Detail Mode let you control settings used within the Flow preset’s plug-in signal-chain. I’ll get to the macro-based options provided by Flow Mode — which is the Flow Studio’s primary reason for being — in a minute. However, if you also want to use the hardware controls to dig deeper into the parameters offered by a specific plug-in within your signal-chain, then Detail Mode is where it’s at. Tapping the Detail Mode button once gives you control over the first plug-in within the signal-chain. However, tapping for a second time brings up a list of all the plug-ins included within the current Flow preset and you can use the Omni knob to scroll through these. As you do, for each plug-in entry, your see up to eight parameters (shown in two rows of four; the Shift button toggles between the two rows) beneath the plug-in name and, from this display, you can then use the four rotary controls to adjust any of these eight parameters. Tapping the Action button brings up any additional parameter rows for those plug-ins within the list and you can then use the Omni knob to move between these. Finally, tapping the Detail button again brings a focused view for the currently selected plug-in and provides you with additional visual feedback as you use the rotary controls to adjust parameters. There are some additional elements to the Detail Mode workflow such as being able to bypass a specific plug-in or mark specific controls as favourites. Inevitably, it does take some time for the various button navigation options to lock themselves into your muscle memory but, once done, if you do need to get hands-on with finer details of a specific plug-in, Flow Studio does provide

The Flow Mixing Suite plug-in acts as a host within your DAW for the Flow presets containing the individual Softube plug-ins within your signal chain.

www.soundonsound.com / July 2026

ON TE ST

SOF TUBE FLOW STUDIO

the means to do it and the built-in display provides plenty of useful visual feedback.

In The Flow (Mode) Of course, the star of the Flow Studio’s show is Flow Mode. It’s here that the macro control concept of the Flow Mixing Suite mothership plug-in meets the hands-on control offered by the Flow Studio hardware. The plug-in needs to be inserted on any track or bus within your DAW project upon which you want to apply some Flow-based processing. Once done, however, everything else — selecting a track, loading a Flow preset, and adjusting that preset to shape your track’s audio — can be done entirely via the Flow Studio hardware. Yes, you can open the plug-in window within your DAW and interact with settings via your mouse. Equally, you can press the OSD button (located on the left side of the hardware) and show the On Screen Display. This provides an additional ‘always-on-top’ floating window on your computer display that mirrors the hardware

Provided Plug-ins At launch, alongside the Flow Mixing Suite mothership plug-in, Flow Studio includes perpetual licences for the following 25 plug-ins from Softube’s catalogue. Guitar/Bass Amps • 30W Top Boost • American Mainstayer 100W • Bass Standard Line V8 • Pacific Dual Tremolo 100W Black • Pacific Dual Tremolo 100W Silver EQ/Compressors/Dynamics • British Class A Equalizer • Equalizers • FET Compressor Mk II • OPTO Compressor • Deesser • Clipper Reverb/Delay • Core Dimensions Reverb • Stereo Delay Other Effects • Bad Speaker • British Class A Drive • Dirty Tape • Kraft • Layers • Lion Head • Model 72 Envelope Filter • Model 84 Chorus • Tinner 6 • Tube • Vocal Tuner • Widener

July 2026 / www.soundonsound.com

display content. You don’t get mouse control within this display, but the visuals can be useful if required. Once in Flow Mode, pressing the Flow button again opens the Flow preset browser within the hardware’s display. The first two rotary knobs let you filter the preset list (for example, by instrument or application type), while the Omni knob lets you scroll up/down through the list. Once you find the preset you want, the Solo button selects and loads it into the track’s instance of the Flow Mixing (or Mastering) Suite plug-in. With a suitable Flow preset loaded, the real fun — and the real potential — of the Flow Studio comes to the fore. Whether you are targeting specific elements of the signal chain by controlling individual macro controls, or using the Omni knob to adjust all the macro controls at the same time to simply scale more (or less) of the Flow’s overall processing flavour, the Flow Studio hardware makes these track-level mix moves super-easy to perform. As you make these adjustments, while your ears will undoubtedly be your primary guide, the Flow Studio’s colourful display gives you plenty of visual support. Twist an individual rotary control, and the display changes to give you specific feedback on the property you are adjusting. Twist the Omni knob and you can then see how all the individual macro controls are being scaled. It’s cool and colourful but, above all, helpful visual feedback. And, if you happen to have the mothership plug-in open on your computer screen, then its display also reflects the moves you are making on the hardware. The various buttons provide access to a number of other options. For example, the Track button allows you to use the Omni knob to scroll through, and then select, a track within your DAW to work on. The Bypass button (Track+Shift) bypasses the Flow’s processing on the selected track for easy A/B auditioning. Tapping the Action button lets you toggle the Omni knob through three modes of operation; Omni Mode (the default where it adjusts all the macros simultaneously), Output Volume (it controls the channel fader of the selected DAW track) or Input Volume (it controls the input gain into the Flow’s signal chain). In short, you can navigate the entire Flow-based processing system entirely from the Flow Studio hardware.

Hands-on Mixing I discussed the workflow merits of Softube’s Flow concept in some detail

The Flow Studio’s screen provides excellent visual feedback. From the top, in DAW Mode, Detail Mode, and a selection of Flow Mode examples starting with use of the Omni knob.

20 YEARS OF ACOUSTIC RESEARCH INTO ONE SOLUTION.

Distributed in the UK by emerginguk.com sales@emerginguk.com / +44 118 402 5090

ON TE ST

SOF TUBE FLOW STUDIO

Don’t get me wrong, though; given the choice, I’d happily pair a full Flow Mixing or Mastering Suite subscription with the Flow Studio. The additional Flow presets and plug-ins it provides would not leave many obvious gaps in your processing arsenal for which you might need to look elsewhere.

Who’s It For?

The Flow Studio measures 144 x 217 x 42mm and weighs 660g.

when reviewing the Flow Mixing Suite itself in the March 2026 issue. I’ll not repeat that same discussion here except to underline that (a) Softube’s plug-ins are excellent in their own right and (b) the macro-based control system built into the Flow Mixing Suite’s mothership plug-in is very elegantly executed; it allows you to make lots of corrective and creative mix moves very rapidly. The Flow Studio hardware adds three obvious elements to an already an impressive workflow. First, whatever the mixing task might be, a little bit of hardware control goes a long way to making the process a more comfortable and more efficient experience compared to using a mouse. Second, once you embrace the macro-based approach to applying signal-chain adjustments, Flow Studio’s rotary controls take what the Flow Mixing Suite mothership plug-in already offers and simply makes it much easier to exploit. And, third, via the Omni knob, it adds an even more streamlined ‘master macro’ approach to the process that means one control might be all you need to start shaping the sound of an individual track. In practice, these elements combine to let you make some very powerful processing moves very quickly. For that first-pass mixing stage where you are simply attempting to get all your tracks and busses into some

July 2026 / www.soundonsound.com

sort of shape, the Flow Studio and some suitable Flow presets can easily get you 80 percent of the way there with considerably less than 20 percent of the effort. You might then need to dig into the details on some tracks, either via the Flow Studio’s Detail Mode or with your mouse, but as a means of getting your mix off to a flying start, the Flow Studio system can put in quite a shift. The assimilation time is minimal, but the mixing time saved can be considerable.

Studio Vs Suite An obvious question for potential purchasers is whether the plug-in bundle included with the Flow Studio hardware — and the associated Flow presets built using just those plug-ins — provide enough options, or whether additional Softube plug-ins (and perhaps a Flow Mixing or Mastering Suite subscription) are required to fully capitalise on the benefits the hardware can bring. I was pleasantly surprised at how much I could get done based upon just the 25 plug-ins supplied within the Flow Studio bundle. The Flow presets built upon just this collection provide plenty of options while also letting you to start building your own custom Flow presets tailored to your specific needs, whether that’s for tonal balance, dynamics, saturation or ambience, or for handling your basic amp and effects requirements with any DI guitar/bass tracks.

So, who will the Flow Studio suit? And, on the flip side, who might it not? There are three obvious tick boxes to consider here. First, are you willing to embrace (or have you already) the Softube plug-in ecosystem for your core mixing tasks? Second, is a more hands-on approach to mixing an attractive prospect? Third, is a macro-focused approach to controlling your core mixing moves something that appeals? I think the last of these is particularly significant and will, essentially, determine the real appeal of the Flow Studio to any potential purchaser. Given the high quality of the hardware, and the broad coverage provided by the included plug-in bundle, there is very little to criticise here in terms of product value. Personally, I see the attraction of the multi-level macro control that Flow Studio can bring to my initial mixing workflow. It’s fast, efficient, very tactile and very effective, and it would absolutely suit the way I like to make my first pass of a mix. But, if you are the kind of mix engineer who prefers to deep-dive into the specific settings of each plug-in within your signal-chain, and particularly if you like to do that with plug-ins drawn from a number of different developers, then that’s a perfectly valid position and, if so, Flow Studio is perhaps not such a compelling choice. Softube’s documentation describes ‘flow’ as a ‘mental state of immersion’ and I think that concept underpins what the Flow Studio is about. It’s an impressive combination of hardware and software that attempts to get you into the creative zone with your mix and let you move swiftly — and without distraction — through the bulk of the work. The three tick boxes above all duly noted, I think the Flow Studio achieves that underlying aim with some considerable style and finesse. This is perhaps not a product that will suit everyone but, if you like to get immersed in the flow of your mix and not get distracted by the technicalities of every processing choice, the Flow Studio delivers an elegant means to facilitate exactly that approach. £ £339 including VAT. W www.softube.com

"Wow, yes, honestly, great sounding DI boxes. Everything I put through them sounds fantastic. These are now my go-to DIs on every session.”

Brian Scheuble

(Elton John, Stevie Nicks, John Mayer, Tom Petty, Ringo Starr....)

The sound of yesterday, built for tomorrow.

Learn more

INTER VIE W

PE T E R T Ä G TGR E N : PRODUCING M E TA L

you want to go all the way to a demo sound. You approach things together with whoever is in the band. They have a vision of things, and you make it come true for them. So, sometimes, it’s not about the sound. It’s about the feeling of the whole production... You take the band that comes in, and you create for what they are and nothing else.”

A drum kit set up for recording in the Abyss live room. Note the pair of Coles 4038 ribbon mics close to the floor in front of the kit (left).

Immortal Story

Photo: Irina Panova

band right away: ‘Oh, this maybe is not so good. Let’s try to do it like this.’ Or: ‘Oh, you played it four times. Let’s cut it down to two times instead.’ So, it all depends.” Many of Tägtgren’s productions have been polished and radio-friendly, but he’s not afraid to embrace the harsher and more lo-fi side of metal too. A good example is the 2019 album Revelations Of Oblivion by death metal stalwarts Possessed: “When I did the Possessed album, I wanted to keep the sound very old-school. I wanted to keep, let’s say, not sloppiness, but I wanted to avoid tightening everything up too much

because I don’t think that would sound good with their style — to have a real laptop production, where everything is cut up into pieces and tightened up like a drum machine type of thing. “With black metal, I don’t think you should do too much editing. It should just flow with you, I would say. I would probably try to get as good a sound as possible right away. And then, in the mix, you can make it a little bit more dirty. That way, you have the opportunity to change things. But if you record it really shitty, then you’re fucked. You can’t really do anything with it unless

Tägtgren’s black metal credits include producing and mixing five albums by Immortal, beginning with 1999’s classic At The Heart Of Winter (1999). “When they came in, and we were going to do At The Heart Of Winter, we didn’t think about what we were going to make, I think. The focus was mainly just to get a really good sound. I just tried to form the sound together with the band and tried to get this ambience, a feeling of coldness. The songs were a little bit different. They were a little bit slower than the previous stuff. We’re all big Bathory fans, and Bathory was using a lot of reverb and stuff. So, we added a lot of reverb to get the old ’80s feeling of black metal. And that was totally the opposite of Immortal’s previous album. That was really dry. I guess they didn’t want to stay with what they did in the past. They wanted to move on. So, how we did it came naturally. We were just working until we got what we thought was really cool and unique.” Immortal’s Northern Chaos Gods (2018), on which Tägtgren played bass as well as producing and mixing, came nine years

Abyss is a hybrid studio, with an impressive selection of outboard equipment.

Photo: Irina Panova

July 2026 / www.soundonsound.com

INTER VIE W

PE T E R T Ä G TGR E N : PRODUCING M E TA L

The New Generation Starting with Hypocrisy’s End Of Disclosure (2013), Tägtgren has often enlisted the creative support of his son Sebastian. Sebastian not only played drums on Pain’s I Am and Coming Home, for instance, but he co-produced the album and entirely composed and performed two of the tracks. “He started writing music himself. Me and Till bought him a really nice Powerbook for Christmas so he could sit and fuck around because he did ‘Matematik’. So, he’s been sitting with that thing and making some songs and things like that. And I don’t see that he has any problem with it, to be honest. It sounds really good. He came up with the sounds for ‘Don’t Wake The Dead’ and also ‘Revolution’, and I didn’t really have to change anything. I just had to mix everything together.” The songs on I Am span many years of writing and production. ‘My Angel’ which features France’s Cécile Siméone, was released as a single in 2011. The second oldest number, ‘Party In My Head’, became a hit single in 2021, whereafter the rest of the album was written in between tours and other commitments. “It’s always hard to mix an album that you recorded the songs for in different stages, because it’s never the same drum sound,

July 2026 / www.soundonsound.com

Photo: Irina Panova

after the group’s previous full-length, All Shall Fall (2009), during which period the line-up shrank from four to two, and co-founding member Demonaz stepped into the role of frontman formerly occupied by Abbath. Did Tägtgren have to alter his approach? “Well, both yes and no. You try to do new things all the time. The album sounded really cool, and the songs were great... We had different sounds and so on. It was kind of hard, but you have to try everything you got. But in the end, it’s the band who decides, because it’s their money. Since I’m a musician myself and work with labels, I know how important it is. You can’t piss away the band’s money, because they will have to recoup it anyhow. It’s not the label’s money. It’s the band’s. So, you always keep in mind what the band wants, or at least I do. I know a lot of producers that just go with their own feeling, but you have to be careful so that you don’t mess up a whole production. I’ve heard a lot of bad things about that. I always try to play it a little bit safe when I work with other bands. You shouldn’t go totally nuts trying things.”

Additional outboard includes a fully loaded SSL XLogic system and two Neve 4081 preamps (right).

it’s never the same overhead, it’s never the same bass, it’s never the same of this and that. Every song that you do is like starting from the beginning. There’s no map that you can follow or anything like that. So, it was kind of challenging to get all 11 songs to sound like they were on the same album. But when you record an album all the way through with bands, it’s so much easier. You set a good sound on one song, and it usually works on the rest. You have to go a little up and down on snare and guitars or whatever. But in general, you can keep it similar and then go plus or minus 1dB or 2dB for some instruments. But when you have totally different sounds on every song, it’s a nightmare. “When you’re going to start recording an album, it’s good to have a goal that you keep on following while you’re recording it. I think, to make the process easier, you should definitely know what you want before you record, because then you can adapt to that all the time that you’re recording. And then, at the end, it should be close to what you had in mind. I think the best thing to do when you’re done recording everything is to just leave it alone for a while and then come back and start mixing it, because when you do all these things at the same time, you start focusing on the wrong thing that really doesn’t matter in the long run, so to speak. So, I think it’s good to have a little break and then go crazy. “Sometimes, when you’re mixing a song, you go: ‘Oh, it would be cool if you put this right here.’ So, you try that,

and then you listen to it a couple of times to see if it works or doesn’t work. You just try a bunch of ideas when you get the feeling that something should be here. And sometimes, you can’t really find what the something is. So, you really pull your hair out. And in the end, hopefully, you find what you have in mind, because what you have in mind and what works can sometimes be two totally separate things. I like challenges. That’s why I’ve been doing this crazy shit myself for 30 years — you know, trying to produce myself, record and mix.”

It’s Who You Know No matter how dark and intense the music, Peter Tägtgren feels that the process of recording it should, above all, be enjoyable. And one of the things he values most about the metal scene is the attitude of the people he gets to work with. “All of these bands are like your kids, you can’t choose which band you love best. But I have great memories with Celtic Frost, Dimmu Borgir, Marduk, and all the bands I recorded. I was very lucky to record bands that don’t have big egos or anything like that. So, 99.9 percent of the time, I was walking into the studio with a smile every morning. There was never really anything like: ‘Oh my God, I can’t work with these guys for another three weeks. I’m going to kill myself.’ I never really had that feeling. Which is good!”

Experience a new level of authenticity.

iamReverb’s distance controllers put you right where you wanna be. Experience a new definition of naturally authentic reverb and become one with the music.

Starting from $ 99,iamreverb.com

The size of sound is the size of space: Enter rooms, halls and cathedrals – or create the room of your dreams.

iamReverb

ON TE ST

ASM Leviasynth Polyphonic Synthesizer

The name Leviasynth suggests an enormous beast of tremendous depth... which isn’t too wide of the mark. GORDON REID

shun Sound Machines (ASM) are a brand of the Chinese manufacturers Medeli. This isn’t an unusual arrangement; many companies create brands to address different markets, often to minimise the impact of affordable products on existing, higher-priced ones. But on this occasion, the concept has been flipped on its head. Known (or rather, not very well known in the West) for their budget arranger keyboards and digital pianos, it’s unlikely that a top-end synthesizer would have made a significant splash had it carried the Medeli name. I don’t know whether this was the only reason that the ASM

July 2026 / www.soundonsound.com

brand was established but, either way, it proved to be an astute move because, almost as soon as it was released, the ASM Hydrasynth became a respected and successful instrument. Now, seven years later, I have in front of me the evocatively named ASM Leviasynth. Is this a replacement for the earlier model, or something altogether new? Without wanting to give the game away, it’s the latter, and it’s rather exciting.

The Software Astute they may be, but the people at ASM/Medeli still have the ability to confuse. When I first read about their eight-oscillator, 16-voice polysynth, I wondered how they had implemented

half an oscillator per voice. Of course, they meant eight oscillators per voice. Well... sort of. The Leviasynth is, at heart, an extension of an FM synthesizer, so they really meant eight operators per voice. Mind you, all of the operators in an FM synth can act as carriers (ie. sound generators) if the appropriate algorithm is available, which it is. With each operator offering 323 single-cycle waveforms and a six-stage DAHDSR amplitude contour generator, you can construct sounds of great complexity on the Leviasynth without invoking any additional facilities. Of course, this would restrict the Leviasynth to a minuscule fraction of its potential because it offers 144 factory algorithms, plus an editor that allows you to create any other feasible 8-op algorithms. What’s more, the Leviasynth allows you to morph between up to eight of these as a sound develops, sliding smoothly between them (well, 100 steps per transition, but that’s pretty smooth) or even jumping backward and forward if you apply a suitable modulator. You can

even save the sounds created between discrete algorithms! The possibilities are mind-boggling. The ridiculous flexibility of the initial sound generation doesn’t stop there because each operator offers seven modes. Yes... individually! In addition to FM, there’s phase modulation synthesis, three models of phase distortion synthesis, a form of width modulation that you can apply to all of the waveforms, plus something called HTE Sync, which is an extension of hard sync. Once generated, each voice passes through two filter sections. The first comprises a digital filter with 18 modes including two SEM-inspired, morphing state-variable filters; a 48dB/octave low-pass filter; and a morphing formant filter. The second is an analogue 24dB/octave low-pass filter bank. Both filter sections offer overdrive, are resonant, and will self-oscillate. After the audio amplifiers (which offer all of the usual sound-shaping facilities) the signal reaches four effects sections in series: pre-FX, Delay, Reverb and post-FX. The pre- and post- sections offer the same set of nine effects types, but this can possibly be justified because (for example) the sound of a signal passing through a rotary speaker effect before reverberation is quite different from what you obtain if you apply the reverb first. The effects algorithms are quite suitable for sound design and live work, but I might still be tempted to use external units or software in the studio. Outside of the signal path, you’ll find five additional six-stage, loopable, sync’able contour generators, with maximum stage times of 15 minutes! In addition, there are five sync’able LFOs that will oscillate as slowly as 100 seconds per cycle, and can be smooth or quantised to as few as three amplitude steps, as you choose. And, in addition to the preset destinations on the panel, there’s a 32-slot modulation matrix within

which you can apply these, the operator contours, and all manner of internal and external sources to a huge list of possible destinations. The other main elements in the architecture are a multi-mode arpeggiator with 64 factory and 64 user-programmable phrases, and a three-track sequencer that offers two Note Tracks and a Macro Track with eight lanes for recording your twiddling. These do everything that you might expect, but one facility deserves special mention. This is called Entropy, and it takes the fashionable concept of random playback variations and then force-feeds it steroids. Currently, this is only available within the arpeggiator but, by the time you read this, it may also be incorporated into the sequencer. I suspect that there are many owners who are going to love it. Oh yes... and, in addition to the Single mode, there’s also a Multi mode. This is misnamed; it’s a Duo mode that, in large part, turns the 16-voice Leviasynth into two eight-voice Leviasynths. You can layer these, split and crossfade them across the keyboard, and transition between them using velocity splits and fades. Numerous parameters then control how things such as the ribbon, arpeggiator and macros work on the two sounds.

The Hardware With its solid construction, 4.3-inch touchscreen, numerous buttons and friendly knobs, the Leviasynth isn’t as hard to use as the description above might lead you to fear. Mind you, given the number of parameters and the amount of visual feedback, it would benefit from the 15-inch screen found on my Korg OASYS. But maybe I’m just being greedy. On the top panel, you’ll also find two 3.5mm analogue CV inputs that you can assign as sources in the modulation matrix.

Round the back we find quarter-inch sockets for audio output and expression and sustain pedal input; in, out and thru MIDI ports; a USB‑B port; and a socket for the external 12V power supply.

ASM Leviasynth £2799 pros • A synthesizer of immense flexibility and potential. • It’s easier to navigate than you might imagine. • It’s nice to have a long ribbon controller, polyphonic aftertouch and MPE. • It feels reassuringly solid.

cons • It may be too complex for some users. • The rear‑panel I/O is awkward, particularly on the desktop module. • It only offers a single stereo pair of audio outputs. • Wall-warts. Bloody wall-warts.

summary This is a hugely deep synthesizer. The name suggests that it may be potent enough to eat the souls of the damned, and the imagery is not inappropriate. If you’re in the market for a serious polysynth, you cannot ignore it.

Next to these lie pitch, gate, clock and dual CV outputs that you can also assign in the matrix, allowing you to drive external synths and effects. You’ll also find quarter-inch and 3.5mm headphone sockets together with a dedicated volume control on the front of the synth, which is good. In contrast, the rear panel offers just one stereo pair of quarter-inch audio outputs, two pedal control inputs (sustain and expression), MIDI in/out/thru sockects and a USB‑B port. The Leviasynth would benefit from having two stereo pairs for the Multi mode, so this is a missed opportunity. What’s more, all of the rear-panel sockets are recessed and, in true Douglas Adams fashion, their legends are written in black on a black background. That’s annoying! On the Leviasynth desktop module, the sockets are so far recessed as to be inaccessible unless you pick the thing up and turn it over. Perhaps this could be to accommodate the cables if you mount the unit in a rack using the supplied rack ears. Oh yes... and while I’m on the subject of connectors, both the Leviasynth keyboard and its desktop sibling are powered by wall-warts. Oh, come on!

www.soundonsound.com / July 2026

ON TE ST

Native Instruments Absynth 6

years is a long time in software, but NI’s Absynth Software Synthesizer Twenty‑five is still with us, still green and still a force to be reckoned with.

Native Instruments Absynth 6 £179 pros • The browser. • The sound. • The large variety of oscillators on offer. • The startling capability of the envelopes.

cons • No undo function in the wave editor. • User presets don’t appear in the visual browser yet.

summary Absynth 6 is a worthy successor to Absynth 5 and a logical evolution to what was already a very powerful and great‑sounding synth. I would highly recommend trying the demo at least. I suspect you won’t be disappointed.

July 2026 / www.soundonsound.com

The patch page is where the towers of power all come together. JEM GODFREY

’ve had a version of Absynth lurking in my plug-in folder since about 2001, when software synths were still largely a novelty act (anyone remember Plugsounds 1-6? Samplecell?). However, I found it intimidating and too spiky sonically to use in the squeaky-clean pop productions I was making at the time. Also, like the lethal drink it was wittily named after, it was rather offputting visually. As a result, we didn’t hit it off and so there it sat, unloved, at the top of the pull‑down list for years until it eventually got moved to the ‘unused plug-ins’ folder where it languished miserably ever after. The end.

Twenty‑five years later, however, unlike my pop career, Absynth lives on, and any bit of software that can claim a near uninterrupted lifecycle of over a quarter of a century has to have something about it. So when news of Absynth 6 started doing the rounds, my interest was piqued anew. Redesigned from the ground up by a team lead by original creator Brian Clevinger, Absynth 6 takes all that was undeniably great about its ancestral lineage and smacks you round the chops with a thoroughly modern reboot of what is now regarded by many as a soft synth classic. The basic principles of what made Absynth so sonically versatile are still there — three oscillators,

or source modules in Absynth-speak, arranged into three channels with a top-to-bottom semi-modular signal flow modulated by two further insert modules per channel. These modules process the sound coming from the oscillator module above it and can variously be filters, modulators or waveshapers. These channels are then mixed into the Main Channel at the bottom of the UI, which has space for three further modules, two of which give you the ability to add more filters, modulators or waveshapers, with the third module reserved for effects and to which you can add the six heritage Absynth classics: the Aetherizer granular processor, a multi‑comb filter, a three-voice independently programmable resonator, a pipe effect (which is an exciter that enables you to position modulatable virtual pickups along the ‘length’ of the input to create flanging, pitch-shifting or rotary type effects), a multi-tap delay and a three-voice delay. As if that wasn’t enough there’s also a surround section with a multi‑channel panner for mixing in spatial audio. It’s also worth pointing out that Absynth has had surround capability since its third iteration way back in 2005. Future-facing stuff indeed in this increasingly binaural world.

(Not) Lost In Space When the synth fires up it presents you with a very visually appealing browser page, which shows what appears to be a pointillist version of an Anglo-Saxon battle helmet viewed from above. Each dot in the ‘galaxy’ is one of the 2000+ presets that now come as standard, and they’re grouped together in colour‑coded areas of related sonic genre. As you hover over a dot, the name of the preset and a description of its sonic character are helpfully displayed above it. Clicking on the dot previews the sound and a double‑click will load it. There are also search filters to the left of the browser display that thin out the dots as you home in on what you’re looking for, including a nice set of sliders that let you slide and select from choices like dark to bright, clean to noisy and so on. All combined, it’s a brilliantly creative way of quickly finding sounds in what would otherwise be a life‑sapping scroll down a seemingly never-ending list of presets. At the bottom of the browser page resides the Mutator, which is a user‑definable randomiser. There’s also a Mutation History list that populates itself as you descend into mutantry. It’s another one of those bits

of common-sense genius to save your blushes if you’ve been a bit click‑happy.

Oscillators Let’s return to the oscillators, which is where Absynth really starts to bare its teeth. As was the case with Absynth 5, you can choose from a wide variety of signal generators starting with a single oscillator, loaded with (but not limited to) all the usual suspects. In fact, there are 43 preset waveforms to choose from, of varying complexity. As with previous versions of Absynth, you can also draw your own waveforms and hear the effects rendered in real time, which is very useful for better understanding what your doodles are doing to the sound. A spectral mode is also available, which shows 512 bars of the waveform harmonics in amplitude and phase axes. One thing I found myself doing was hitting Command+Z on my QWERTY trying to undo my edits when things went awry. As far as I can tell there is no undo function on Absynth currently. A workaround I developed was to select Copy from the Edit pull-down menu in the wave editor, and if I went down a rabbit hole I didn’t like, I’d select Paste to restore the waveform to where it was before I ruined it. But it seems odd that a born sound-design machine like this doesn’t give you an undo option, even if it was just a single undo. Especially when the Mutator provides a super‑handy history list of your mutations.

There are also 56 morphing waveforms to play with. With the morphing waveforms, aside from the in‑built wavetable-like morphs that you can control by clicking and dragging the waveform in the patch editor page (or by LFO or envelope), you can also create your own by sending your chosen waveforms over to the wave editor page where they appear in a list. You can then set up to four anchor points across the two individual waveforms, around which the morph takes place, controlled by the morph transition bar. The resultant sound is more than just a simple crossfade and can produce some very interesting wavetable-like tones as it changes from one to the other. I did notice that sometimes when changing quickly between waveforms the note would hang and the only thing that would stop it was changing to a different patch and back again, which, if you’re halfway through editing the patch of your life, might be a bit of a thing if you hadn’t saved anything at that point. Oscillator pitch can be set over a rather extravagant near-16‑octave range, 128 semitones down and 64 up. Additionally, you can set the resolution of the transposition to be greater or lesser than one semitone; so entering in a ratio of 1.500, for example, will increase the tracking to a tone plus a half tone across each semitone in an octave. The resolution can be fine-tuned all the way down to increments of a thousandth of

The preset explorer in the browser page is inspired and fun to use.

www.soundonsound.com / July 2026

TALKBACK

Jamie Stewart WILLIAM STOKES

ith Xiu Xiu, Berlin-based American artist and producer Jamie Stewart has for over two decades been at the vanguard of experimental-leaning indie rock. Their latest release is a wildly eclectic covers album, Xiu Mutha Fuckin’ Xiu: Vol. 1. At the moment I can’t stop listening to Clyde McPhatter’s Greatest Hits. My bandmate Angela [Seo] and I are massive fans of Black American ’60s soul. I’ve always been particularly blown away with records from that era. It was a group of people in a room, all live, and yet it sounds so extraordinarily well put together, so rich and so thoughtful. I mean, clearly they put an extraordinary amount of time into the mic placement, and they must have done a ton of run-throughs to get the balance as good as they did. And it’s clearly just a two-track, like, all the instruments are on one side, the vocals are on one side, with a little bit of bleed. At that time, not an extreme choice. But on a record today, that would be a totally extreme choice. So it’s interesting to listen to something where they were just dealing with the technical limitations of the time and not having a super clear — I don’t want to say understanding — but a clear conception of what the possibilities of stereo were. A lot of the very first stereo records, people were just going, “OK, well, we’ve only done mono. Let’s just try this fucking two-track thing. Everybody hates it. We all think it sounds like shit, but the record company is demanding we do it at stereo, because they can feel like they could sell more records that way, or something.” So, you know, you’ve got the horns and the bass and the drums and the guitar and the organ and the piano like all cranked on one ear, and then the vocal blasting on the other side. It always sounds pretty incredible.

100

July 2026 / www.soundonsound.com

The artist I’d most like to collaborate with Since I was a kid, Prince has been my number one. When I was a little kid, I was listening to songs like ‘Darling Nikki’. That’s not a song for a nine-year-old! But as an adult, I realised how much his music shaped me, in that it was really totally insanely filthy at times, but also done with an extraordinary amount of emotionality and care and musical thoughtfulness. And he was, you know, technically one of the best musicians there ever was. I’m technically not a particularly good musician. I mean, I work real hard and I’m willing to do 100 takes to get something to where it needs to go. But it would be fascinating to me to work with somebody who is both largely responsible for me having such a dirty musical mind, but also who’s 9000 times the player that I’ll ever be! The first thing I look for in a studio What the room sounds like. I’ve had a chance to work in a couple of really excellent studios, but most of the time it’s been at home studios that we have, and with the studio we have in Berlin, it’s the first time we’ve had an actual ‘regular studio’, that has a control room and a live room. Not that I’m a genius at mic placement or anything like that, but largely, when I walk into any interesting room I’ll just be very curious as to what the reflection is — and not that it would necessarily determine any choices that I would make! But just that very pure sound, of the reflection. I think it’s the case for a lot of people. You walk into a space and it just orients you, it hits you right. It can be a great sound, or an unusual, just an unusual physical sensation. The person I would consider my mentor My dad [Michael Stewart] was a very successful musician and record producer. He was a workaholic and had a lot of drug problems when I was growing up,

so we didn’t really get a chance to spend all that much time together, and didn’t talk about music all that often. But the couple of times that we did... my life would not be the same without having those couple of conversations. I remain deeply, deeply indebted to his wisdom, for lack of a better word. He told me three things. One: I asked him how to play rhythm guitar, and he just said, “Mute the strings, put on a metronome, go like this [mimes strumming a guitar], and just get your right hand going. Because that’s what it is!” The second thing was when I started getting a little more serious about music and he asked me why. I might have been 19 or something, and I didn’t know why. It didn’t even occur to me. I just stood there looking at him blankly. And he said, “The reason you get into music is to try to touch somebody. That’s the goal.” And we fail at that all the time. But the goal is to have what we’re working on give something to somebody else. It put the act of making music into a very different perspective. And it put the act of listening to music in a very different perspective, too. All of these records that I had been obsessed with and listened to constantly, it was like, “OK, the reason that this is resonating with me is because very clearly, the person who made it was trying to give something to the listener.” It was a really good thing to hear, particularly as a very young person. And the third thing: he worked on a lot of pop and R&B records. And they were great, really beautifully done, but he always said he felt like he didn’t take things far enough. He said, “Always take things too far!” So, between the rhythm guitar thing, thinking about what a listener might get out of what we’re doing, and always trying to take it too far, that was really hugely important to us. My go-to reference track or album I have to admit that is a that is a habit that I very foolishly do not partake in

Photo: Eva Louise Hoppe

Xiu Xiu are Angela Seo and Jamie Stewart.

very often. And I really fucking should. I’m always impressed whenever we’re soundchecking, you know, if the engineer plays like a hip-hop song, and then a ’90s country song, then they’ll play white noise, then they’ll play Joy Division. I’m like, “OK, this person is listening to what is happening in the room. Yeah, this night is going to work out.” My secret weapon in the studio is It’s really boring. It’s a really old plug-in that has been coming stock with Pro Tools since the year 2000. The kind of shitty pitch-shifter that it has. We use it all the time. You take the Time Correction off, and it just completely changes it. We have a pretty good collection of late ’70s and early ’80s preset drum machines, just organ-style drum machines, and they sound incredibly good. But there’s not a ton of low end and the grooves are kind of rickety and weird. They’ll float in and out of time a little bit, because they’re 50 years old! But if you record it at the original tempo and then shift it down an octave without time correction, it becomes twice as long. It’s still in time with itself but creates a nuts polyrhythm. The kick drum will be, like, every other bar, there’ll be a really low one and a high

one, you know, and then the snare will hit in different places... the hi-hat will be twice as slow against itself, twice as low and twice as long. That’s one of my ultra-favourite things to do with it. If they ever remove that plug-in from Pro Tools, we’ll just have to quit. The studio session I wish I’d witnessed Probably Roy Orbison, ‘In Dreams’. Just to see how that happened, in a single moment. I’m sure they did a million takes, but the one that we all know and love: that was not something that was pored over, over several months. It happened in two minutes and 45 seconds of just extraordinary playing, and extraordinary singing, and then that was it. They were done. The producer I’d most like to work with Luckily, the producer I’d most like to work with is one we regularly get to work with, John Congleton. We’re old friends, and because of that, I trust him implicitly. Even though he has worked on a ton of very popular mainstream records, he has his own sort of world view and his aesthetic is super far-out. He loves taking risks. He loves taking things to their extremes. He loves experimenting. He

loves doing things that in a billion years I would never think of, but at the same time, is completely in line with something that I would want to hear. He’s also really funny and just a great hang. Incredibly nice and super encouraging and really sweet — and extraordinarily technically skilled. Oh, and a great songwriter too. Not just his understanding of traditional Western song structure, but also his understanding of a lot of experimental music and a lot of far-out classical music, and then having the chops to make things 90 times weirder than we ever would have thought it would be, but at the same time, super engaging and listenable. On many occasions, I have sent him something to mix and to produce and just said, “Do whatever you want.” And it comes back, and it’s perfect. The studio experience that taught me the most Cory McCulloch was responsible for producing the first four records that we did, and I really miss working with him. He was walking me through an early vocal take, and I was having some trouble with it. And he just said, “Stop thinking.” And it immediately clicked. This was not about trying or about being analytical. This was about being in what is happening emotionally, what’s happening musically, in this moment. The attempt to just be in what is happening in that moment musically, to remove any of my conscious mind from it, is always something that we strive for. And it’s physically impossible to do, unless you’re like a Tibetan Buddhist master monk! But it was exactly what I needed to hear at the time. I’m not saying all music is about that, but the music that we’re trying to do is an attempt to ‘get out of the way’ of conscious thought. I would not exist musically without that exact moment. The advice I’d give myself of 10 years ago I’ll say it for 12 years ago, because we were at a point where it seemed incredibly possible that the band was not going to be able to continue. And it was extraordinarily stressful and just sad. I felt like I was getting divorced or something. It hurt my heart, the possibility that this was not going to work any more. So my advice would be: just take a deep breath. The muse will smile upon you again. It’s going to take a couple years, but don’t worry. It’s going to be OK.

www.soundonsound.com / July 2026

101

RECORDING & MUSIC TECHNOLOGY SHOW

Saturday 27th June 2026

TIC ON SKETS ALE

NOW

Over 100 Brands Exhibiting At The Show Come and get hands-on with Microphones, Monitors, Recorders, Synths, Audio Interfaces, Mixing Consoles, Outboard, Software and much more. Brands Exhibiting Include: 1010 Audio AKG Amphion AMS Neve Antonus API Audio Artel Arturia ASM Synth Atomic Analog Audeze Audient Audio Merge Audioscape Austrian Audio Avantone Barefoot Sound Black Box Analog Designs Buchla Buso Audio Chord Electronics Coles Electroacoustics Cranborne Audio Crumar DPA Microphones Dr Alien Smith

Drawmer Dreadbox Ear Trumpet Labs EFNote Empirical Labs EVE Audio Eventide Expressive E Extinct Audio Focal Franklin Audio Funky Junk Gainlab Audio Gemini Genelec GRP Synthesizer GS Music Harrison HEDD Audio Highland Dynamics HUM Audio ICON IK Multimedia Imersiv Interphase Audio IsoTek Systems JBL Bandbox

JBL Pro Kali Audio Kernom Kerr Acoustic Lynx Studio Technology Mackie Majella Make Noise Pro Audio ME-Geithain Merging Technologies Microtech Gefell Milab Modal Electronics MOTU MXL Microphones Myburgh Nektar Neumann NORD Nordic Audio Labs One Louder Studios Overstayer Pearl Microphones Pope Audio Present Day Production PSI Audio Pulse Techniques (Pultec)

PWM Recording The Masters Reverbophone RODE Rodec Rupert Neve Designs Schoeps Sennheiser Sessiondesk Slate Mics Solomon Mics Sound On Sound Soundcraft Source Elements Soyuz Microphones SSL Tantrum Audio Teaching Machines Teenage Engineering Telegrapher Loudspeakers Trinnov Vanguard Audio Labs Vegatrem Vicoustic Voyage Audio Wolff Audio Zoom

TALKS & DISCUSSIONS AI and emerging technologies in the music studio With Natalia Milanesi an audio technologist, sound engineer, music producer and educator. Natalia is a Grammy Award-winner, having worked with artists such as Sam Smith, Flyte, Jaz Coleman, Martin 'Youth' Glover, Bernard Butler, and companies such as BBC, Disney, Netflix and ITV. Hosted by: Penny Eau

MixBus Podcast Live Panel talk with top producer / engineers Marco Pasquariello, Neil Pickles and Julian Kindred. Talking about how you get started in the recording industry. Hosted by: Kevin Paul

Field Recording: A Practical Guide Author Marcus Leadley talks about his new book and the growing popularity of Field Recording. Whether you're interested in wildlife or urban soundscapes, you'll find his advice on choosing the right digital recorder and microphone equipment invaluable.

University of Westminster Harrow Campus, London. 20 minutes out of central London between Wembley & Harrow on the Metropolitan tube line.

www.gearexpouk.com A

EVENT

ON TE ST

PHIL WARD

he intention behind Barefoot Sound’s Footprint active monitor range, from which I have previously reviewed the 01 and 03 models, is to integrate many of the electro-acoustic philosophies and principles found in the company’s high-end models within more compact and cost‑effective offerings. It’s a strategy that has taken Barefoot from being a relatively niche aspirational brand to a prominent position in the more ‘mass market’ monitoring sector. Where once Barefoot were up against the likes of PMC and ATC, for example, the Footprint range competes head-on with models from ADAM Audio, Neumann, Dynaudio, HEDD and Focal. The recently launched Footprint 02 Gen2 is an improved and upgraded version of the original Footprint 02. The list of improvements and upgrades is so extensive, though, that I’d be very tempted to put it in the ‘new product’ category. The Footprint 02 Gen2 comprises a closed-box, active three-way nearfield system, carrying horizontally opposed 6.5‑inch (165mm) bass drivers in the side panels. The softened edge and gently radiused front panel is home to a nominally five‑inch (127mm) midrange driver and one‑inch (25mm) tweeter.

Motor Skills The bass and midrange drivers incorporate pressed‑aluminium diaphragms, while the tweeter diaphragm is a fabric component configured in a ‘ring radiator’ format (like a dome but with its apex fixed). All three drivers incorporate elements within their motor systems aimed at enhancing linearity and reducing distortion. These elements generally take the form of shorting rings and copper caps on the pole pieces to reduce the flux and inductance-modulation phenomena that are inherent to a moving‑coil driver architecture. Such things are, I think, fast becoming obligatory in any speaker that aspires to genuinely high performance. The cabinet feels extremely rigid and is far heavier than its compact dimensions would suggest. I was not a fan of lifting the Footprint 02 Gen2 onto the monitor shelf. Barefoot’s specs claim -3dB bandwidth limits of 38Hz and 40kHz for the lows and highs, respectively; response linearity of ±1dB between 42Hz and 30kHz; and maximum sound pressure level (SPL) of ‘≥111dB’. The SPL spec doesn’t define

104

July 2026 / www.soundonsound.com

Barefoot Footprint 02 Gen2 Three-way Monitors

Barefoot’s latest compact three-way represents a significant step up from its predecessor. a measuring distance or space (half or quarter), so it’s somewhat meaningless. In contrast, the low‑frequency bandwidth spec includes some interesting further information; the system Q is 0.56, which means the low‑frequency resonance is slightly overdamped (as the bass driver’s voice coil warms in use the Q will increase and the system will become less damped), and the high-pass slope is 18dB/octave. Now, the inherent high-pass slope of a closed‑box system is 12dB/octave, so clearly the Footprint 02 Gen2 includes a 6dB/octave high-pass filter in its signal chain. This is almost certainly included to offer some reduction of the bass

drivers’ diaphragm excursion at very low frequencies. It will increase low‑frequency group delay somewhat, but probably not very significantly.

Ins & Outs Around the back of the Footprint 02 can be found a mains input socket and switch, a balanced analogue input socket, a stepped input sensitivity knob, and 3.5mm jack sockets for Barefoot Sound’s proprietary MEME (Multi Emphasis Monitor Emulation) technology. Downstream of the input socket and digital signal processing (more on that later), Footprint 02 amplification comprises low‑, mid‑ and

high‑frequency Class‑D amplifiers rated respectively at 200, 150 and 60 Watts. The LF/MF and MF/HF crossover frequencies are 180Hz and 3700Hz respectively. One of the consequences of employing opposed side‑mounted bass drivers is that doing so demands a relatively low LF/MF crossover frequency, in order that midrange output isn’t radiated sideways where it could potentially cause all sorts of havoc with dispersion and integration with studio acoustics. The 180Hz crossover frequency of the Footprint 02 is low, but I might have expected it to be even lower, say nearer 150Hz. The trade-off here, however, is that the lower the crossover frequency, the harder the midrange driver has to work, and the more its engineering must take on some of the characteristics of a bass driver (larger excursion, longer voice coil, more generous surround...). And if there is one advantage of a three-way scheme over the two-way, it’s that midrange drivers can be optimised for that role rather than also having to play bass. Now, while the opposed side‑mounted LF driver principle does introduce a crossover frequency conundrum, it also, as Barefoot make much of in their marketing, offers the advantage of driving minimal low‑frequency

mechanical energy into the structure the monitor is standing on or attached too.

Knowing MEME Barefoot Sound’s MEME technology provides switched options (from a compact box with a rotary switch) that alter the bandwidth and frequency response of the Footprint 02 in order to ape the response characteristics of three alternative monitor types. The options are christened Hi-Fi, OldScl and Cube and are intended to represent an imagined generic ‘hi-fi’ speaker and, presumably, a Yamaha NS10 and an Auratone 5C. However, in appearing not to emulate the dispersion, distortion, dynamic compression or delayed resonance characteristics of the alternative monitors, MEME (and other manufacturers’ similar technologies) can’t really hope to do anything but hint at what a mix might sound like when played on them. For this reason — and the more philosophical one that, to my mind, we engineers ought to understand our monitoring well enough to know how it translates without relying on approximate emulations — I’m not convinced of the genuine value of technologies such as MEME. Your mileage may vary, of course, so feel free to ignore my stick‑in‑the‑mud tendencies. During my time with the Footprint 02 Gen2 I made some acoustic measurements to analyse how each of the MEME options changes the monitor frequency response. You’ll find the resulting response curves further down the page.

Knowing New Now to cover the revisions in this new Gen2 model. Firstly, Barefoot Sound introduced Alongside the balanced analogue input and sensitivity control are a pair of mini-jacks for connecting the MEME controller.

Barefoot Footprint 02 Gen2 £3495 pros

• Fabulous extended bass. • Great detail resolving power and imaging. • Plays loud.

cons

• Tonal balance quirks.

summary

The Footprint 02 Gen2 displays some tonal quirks, but beneath them is a well engineered monitor of genuine performance and quality.

a signal routing technique christened SPOC (Spectrally Optimised Conversion) in their Footprint03 model, and the same feature has now been incorporated in the Footprint 02 Gen2. SPOC is a patented technology that rejigs the usual order of things in terms of signal flow and crossovers. In DSP‑based active monitors the crossover functions are usually all implemented in the digital domain; however, this means that the typical spectral balance of music (with high frequencies usually at a lower level than bass and mid frequencies) results in lost digital resolution in the high end because the full 24‑bit conversion range must be reserved for low and mid frequencies. With their SPOC technology, Barefoot implement the high‑frequency high‑pass crossover filter in the analogue domain before the digital conversion, so that the full 24‑bit conversion range is available. The result is said to be an up to 15dB improvement in high‑frequency signal‑to‑noise ratio. It’s a simple but neat idea, I think. I’ve already introduced the drivers, but they’ve all taken a step up compared to those of the original Footprint 02. The bass driver’s aluminium diaphragm replaces the original model’s paper item, the midrange driver has grown in diameter by 25mm (with a corresponding increase in voice-coil diameter) and gained a new high-flux magnet system, and the tweeter has gained an aluminium phase plug and, like the midrange driver, a high-flux magnet system. Perhaps the most significant change associated with the Gen2 drivers, though, is that the Footprint 02 Gen2 has become a fully active system, where the original model was an active/passive hybrid in which the midrange driver and tweeter were driven

www.soundonsound.com / July 2026

105

ON TE ST

BAREFOOT FOOTPRINT 02 GEN2

Diagram 1: On-axis measurement of the Footprint 02 Gen2. Second and third harmonic distortion is shown in green and blue, respectively.

Diagram 2: Comparing the default on-axis response (red trace) with the MEME system set to Hi‑Fi (purple), OldScl (blue) and Cube (green).

by a common amplifier and integrated via a passive filter network. It’s not only the drivers and amplification scheme that have changed. The Gen2’s cabinet has gained a compound radius on the front panel (subtle but very pleasing to my eye), and its aluminium trim panel is now thicker and has become a braced structural element of the cabinet construction. The rear‑panel mounting bracket attachment points are new also.

The Plots Thicken I measured the Footprint 02 Gen2 in my usual large measuring space using a ground‑plane technique to maximise the bandwidth over which the data is effectively anechoic. So, Diagram 1 shows the axial frequency response and harmonic distortion. SPL was 90dB at 1m (2π). I was somewhat surprised by the emphasis centred on 100Hz and the slightly depressed low‑midrange axial response. Based on this measurement, the Footprint 02 Gen2 doesn’t meet its claimed ±1dB frequency response specification. I’d expect the tonal character this response implies to be subjectively apparent — although, as I’ve

106

July 2026 / www.soundonsound.com

written previously, you can’t judge the overall quality of a speaker based on one frequency response curve. Above the low midrange, the remainder of the response curve is pretty linear but considering the slightly idiosyncratic character below, say, 500Hz it’s harder than usual to make any judgment on how the subjective tonal balance will play out. The harmonic distortion shown in Diagram 1 is very nicely

my (very long) experience, speakers designed for domestic hi-fi applications display a huge range of tonal balance characteristics, some arrived at by canny and capable speaker engineers who know what will sound good in a domestic environment (and sell well), and some arrived at somewhat more accidentally (perhaps by those more concentrated on incorporating the latest snake oil than by hitting a frequency response target). The MEME OldScl and Cube response curves display, as expected, increasing degrees of low- and high‑frequency bandwidth restriction, along with a degree of midrange emphasis. These curves might approximate the frequency responses of the modelled monitors, but to my mind, and for the reasons I’ve already explained, they won’t really make the Footprint 02 Gen2 sound much like them in any other respect. And perhaps to reinforce my point about a hi-fi balance not always being characterised by a ‘smile’ response, the monitor that most likely inspired the OldScl MEME option was originally launched as a hi-fi speaker. Diagram 3 illustrates a the Footprint 02 Gen2’s vertical dispersion, and it’s very competently managed with no big discontinuities. As a result, the Footprint 02 Gen2 ought to be relatively insensitive to vertical listening position. I also checked the horizontal off-axis response and it showed equally well managed results.

What’s The Gen? Listening to the Footprint 02 Gen2 revealed, as expected, a subjective tonal balance that to my ears was

“The Footprint 02 Gen2 is a hugely capable and impressive monitor. It has fabulous power to reveal the contents and quality of an input signal...” controlled — especially low‑midrange third harmonic at around -60dB. That’s a very good result. The higher level of second harmonic in the same band is relatively innocuous, I think. Diagram 2 shows the frequency responses with the various MEME modes engaged. The Hi-Fi option increases energy at low and high frequencies to give the response more of a ‘smile’ character. I’m not sure if this is characteristically ‘hi-fi’, however. In

a little generous in the bass and a little undercooked in the lower midrange. But listening also revealed immediately that beneath the subjective tonal balance quirks, the Footprint 02 Gen2 is a very capable monitor. The low‑frequency band is extended in terms of bandwidth yet dynamically precise and fast with great pitch resolution: classic closed-box bass, really. The levels of midrange and high‑frequency clarity, detail and stereo

image focus are very high, and that translates in a mix context to a monitor that reveals very explicitly what’s going on with every element of a piece of music. The very low levels of odd‑order harmonic distortion in the critical band from 100Hz to 1kHz would appear to be significant here. Odd‑order distortion isn’t far away from being uncorrelated noise, and the less noise a monitor adds to the signal, the more detail is potentially audible. The Footprint 02 Gen2 is also able to play subjectively very loud without losing its grip on the input signal. Its revealing and precise nature remains intact even when your instinct is to take a step back or reach for the volume knob.

to see if I could adjust the response with a little EQ. So, Diagram 3 shows a comparison between my usual nearfield The Footprint 02 Gen2 inhabits a crowded slot in the nearfield monitor monitoring and the Footprint 02 Gen2 market. It joins such models as the mounted in the same position. I’m going Genelec 8331A, the Dynaudio Core to remain quiet on the identity of my 47, the PSI A17M, the Focal Twin ST6 usual monitors so as not to introduce any and the Neumann KH310A, all of which element of ‘shoot-out’ to this review. And would be worth considering. I disabled the room alignment software of that system for this measurement. Below 100Hz, Diagram 3 shows that room I couldn’t get away from feeling that modes at 40Hz and 90Hz somewhat the axial anechoic response I measured dominate but also that the Footprint was reflected in what I was hearing. So, as 02 Gen2 reaches further down into the illustrated in Diagrams 3 and 4, I did some low‑frequency depths (partly because in-room measurements to investigate it has a slower LF roll-off). Between, whether the anechoic measurement say, 150Hz and 500Hz, however, the reflects the reality in the room, and then Footprint 02 Gen2 clearly lacks a few dB of level. This ties in with my anechoic measurement. Diagram 4 shows my attempt to modify the Footprint 02 Gen2’s balance using the parametric EQ of my Audient ORIA interface. Not only did I lift the 150Hz to 500Hz band; I also did a little work on the low‑frequency room modes too. The result of my EQ on the Footprint 02 Gen2 was to bring it far closer in balance terms — very close in fact — to my usual monitors with their room alignment enabled. To put that in context, Diagram 3: On-axis response (red), versus the response measured 15 degrees above and below axis my usual monitors are some of the best (blue and green traces, respectively). nearfields I have ever heard. With that EQ engaged, the Footprint 02 Gen2 is a hugely capable and impressive monitor. It has fabulous power to reveal the contents and quality of an input signal and it can do so to pretty low frequencies and at noticeably impressive volume levels. The only question I’m left with is to consider why the Footprint 02 Gen2 doesn’t really offer a flat frequency response and neutral balance out of the box. I wonder if it’s been ‘voiced’ to impress with a little bass lift and midrange duck — which, somewhat Diagram 4: The in-room response of the Footprint 02 Gen2 (red), and an unnamed alternative monitor (blue). ironically, is what often happens in the world of hi-fi speakers... To my ears, then, it takes just a few dB around 250Hz for the Footprint 02 Gen2 to really sing. If you would be happy to consider that, or perhaps combine the Footprint 02 Gen2 with a monitor alignment system such as Sonarworks, I can heartily recommend it.

ALTERNATIVES

Diagram 5: The Footprint 02 Gen2 before and after corrective EQ (red and blue traces, respectively).

£ £3495 per pair including VAT. T KMR Audio +44 (0)20 8445 2446 E sales@kmraudio.com W www.kmraudio.com W www.barefootsound.com

www.soundonsound.com / July 2026

107

ON TE ST

The new Scene Rebalance tool, which does for film sound what previous (and now improved) stem-splitting tools did for music.

iZotope RX12 Advanced Audio Restoration & Repair Software

Since RX11 came out, there’ve been huge leaps in AI-based audio repair and separation tools. iZotope’s boffin’s have now delivered their reply — can RX12 reclaim its crown? HUGH ROBJOHNS

Zotope don’t rest on their considerable laurels — their boffins beaver away constantly, finding clever new ways to improve their products, both in scope and in depth. RX, the company’s well known and highly regarded ‘toolkit for audio repair’, exemplifies this approach perfectly: an updated and enhanced version seems to be released roughly every two years. For example, the last two versions, RX10 and RX11 came out in late 2022 and mid-2024, respectively. Now, after another two years of complex R&D, RX12 is here — and it’s a timely release, since significant progress in the neural net field over the last few years has enabled competitors to leapfrog some of RX11’s capabilities. So while RX12’s brand-new features may catch the headlines, it’s just as

108

July 2026 / www.soundonsound.com

important that several tools have been completely rebuilt and significantly improved, the majority benefiting primarily from iZotope’s expanding expertise in machine-learning and neural-net technologies. Below, I’ll take you through exactly what RX12 brings to the party, and consider whether it still represents the benchmark in audio repair and restoration.

Versions & Tools As usual, RX12 is available in three forms called Elements, Standard and Advanced. The flagship version, Advanced, contains all of RX’s 90-plus tools, plug-ins and standalone processing facilities, while the first two editions contain selected subsets at commensurately lower prices. The Standard version retains a very comprehensive set of over 70 tools and plug-ins, essentially lacking the newest

additions and most of the modules focused specifically on professional film and TV applications, although it does retain the Dialogue Isolate tool. In contrast, RX12 Elements contains a very small subset of six relatively basic — but still capable — noise-removal tools. Both the Elements and Standard editions can, of course, be used as starting points, with upgrade paths available to more capable RX versions should that become desirable. I was invited to review the Advanced edition, so that I could evaluate all the developments now on offer, but I’m conscious that not everyone will need this ‘full fat’ version, so I’ve included a table in the online version of this review (https:// sosm.ag/izotope-rx12) to make it easy for you to see what tools and facilities are or aren’t included in each version. All three editions of RX12 benefit from a GUI refresh that makes them look a little more modern and, I think, attractive. But anyone familiar with previous editions should be pleased that it retains the same overall configuration and workflows that we’re used to. For instance, iZotope’s release notes flag as a major update a larger spectrogram screen at the heart

iZotope RX12 Advanced £1299 pros • Some very powerful, effective new tools. • Cutting-edge AI updates to previous processes. • All previous tools retained. • Improved GUI. • Significant workflow improvements.

cons

System Requirements RX12 is compatible with macOS from v14.7 (Sonoma) and newer, and either Intel or Apple Silicon (native and Rosetta) processors. It’s also compatible with Windows 10 (22H2) and newer. Many of RX12’s tools can be used as plug-ins and are generally available in 64-bit AU, AAX, AAX Audiosuite, VST3, AU ARA and VST3 ARA formats, although a few support only a subset of all these formats. For example, the Repair Assistant plug-in is only available as AAX Audiosuite, AU and VST3, while Ambience Match and De-rustle are AAX Audiosuite only. The Spectral Editor ARA support is currently limited to Studio Pro 8 (Studio One 7) and Logic Pro. When using RX12’s plug-ins, the following DAWs are officially supported: Logic Pro 12, Pro Tools 2025, Ableton Live 12, Cubase 15, Nuendo 15, Fender Studio Pro 8 (Studio One 7), Reaper 7, FL Studio 25, Adobe Audition 2026, Adobe Premiere Pro 2026, DaVinci Resolve 19 and Reason 13.

• Stem splitting limited to four channels. • Advanced doesn’t come cheap!

summary In most respects, iZotope’s flagship tool has caught up with the AI-based competition, to deliver perhaps the most comprehensive audio restoration and repair tool on the market.

of the user interface. It’s certainly there, and to be welcomed, but if I’m perfectly honest it’s probably not something I’d have noticed without the prompt! One thing I did notice very quickly, though, is a new monitor volume slider control in the top right-hand corner: this makes it easy to adjust the volume when critical listening, and is a very handy addition, especially if working on a laptop without an external interface. Many of the improved and rebuilt tools seem to work faster and more precisely, too, without sacrificing any audio quality. Another welcome addition to a lot of the

individual tool windows is a new button with an ‘ear’ symbol located alongside the Preview/Bypass/Compare buttons. When previewing the tool’s processing, selecting this ear button switches the monitoring to reveal what is being removed — the ‘delta’ or ‘difference signal’, if you like — rather than the processed result. In most cases, this makes it far easier to hear when the processing is optimised, because it’s easier to check that the algorithm isn’t throwing away anything important, and while some tools had such a facility built in, it’s great that to have a common approach on the different tools. So that’s a very welcome addition. A major update, in the form of another new button alongside the preview buttons, affects RX12’s three stem-splitting tools. Labelled ‘Stem

Split’, this button activates a new ‘Stems View’ mode within the spectrogram display, where the split output stems are now time-aligned above each other in the same spectrogram. It’s a very big workflow improvement compared with the previous setup, in which the split elements appeared in separate tabs. More on that side of things below. A couple of other worthy but perhaps less significant updates include a new search facility to find specific processing modules — useful, given that there are now over 50 in RX12 Advanced — and that the history panel is now resizeable, which means it’s easier to keep track of complex processing chains.

New Tools One headline new feature of RX12 is a brand-new ‘Trim Silence’ tool. Back in

The music stem splitting tool can now display all stems on a single page of the GUI, rather than in separate tabs — a move that facilitates considerable workflow improvements.

www.soundonsound.com / July 2026

109

ON TE ST

I Z O T O P E R X 12

my broadcasting days I used a similar function in an early version of the SADiE DAW to strip out long pauses in speech or effects recordings, and for creating those impossibly fast ‘terms and conditions’ gobbledegook speeches often heard at the ends of radio adverts. RX12’s Trim Silence tool works in much the same way, with a threshold level control determining what the algorithm identifies as ‘silence’. Three additional ‘time’ controls set the Post-Roll time (a retained period after the start of a detected silence to help preserve reverb tails); the maximum length of silence retained in the edit (between words); and the crossfade time at the end of the silent period, ramping back into the incoming dialogue. iZotope suggest that Trim Silence is useful for editing podcasts, for example, as well as for film dialogue editors stripping out large pauses between takes… although it will always be the ‘terms and conditions’ editing tool to me! Regardless of the intended application, this new tool works very well and it’s easy to set up — optimising the threshold control is really the key, of course — and it’s extremely helpful for quickly sorting the wheat from the chaff in long, unmonitored recordings.

The new Trim Silence tool, which can strip out pauses between words or sections of audio — potentially useful for podcast and dialogue editing (and speeding up T&Cs on adverts!), though it’s worth noting that unlike strip silence in a DAW, this a destructive process.

an update that brings improved speed, higher accuracy and easier control, which are all welcome improvements. Similarly, the Breath Control tool has been completely rebuilt, again with more sophisticated machine-learning technology, and it delivers faster and more precise results than previously when removing or reducing breath sounds from

“RX12 Advanced has regained its premium position as a most impressively comprehensive and capable audio repair toolset.” However, note that this is a destructive tool: once the file is rendered its internal edits can’t be individually picked apart and reworked, as they might in the non-destructive ‘strip silence’ process of many DAWs. Another minor limitation is that this tool always processes the entire audio file that’s loaded into RX12; you can’t strip the silence just from a selected portion within the file. Perhaps that’s more of a reviewer’s problem than a real-life practical issue, though, since most applications would probably want silences stripped from the entirety of the recording. A couple of RX’s other dialogue-focused tools have been improved significantly. The Dialogue Isolate tool, which splits speech recordings into separate stems containing voice, reverb and noise, so that these individual components can be rebalanced, has been updated to use brand new neural nets. It’s

110

July 2026 / www.soundonsound.com

speech and singing. As is often the case when iZotope rebuild a ‘legacy tool’ in a radical way, though, the previous version is retained, and made available as a ‘classic’ algorithm option. These rebuilds may not look like big-step changes, but the incremental improvements they deliver are welcome, and not only do they benefit the sound quality, but they’re also quicker and easier to set up and optimise. I continue to be astounded at how these remarkable processing tools approach perfection in what, when I started in this industry, was considered a technical impossibility! iZotope’s latest machine-learning techniques have also been deployed in a complete rebuild of the De-bleed tool (and its new, real-time plug-in equivalent), to further enhance the reduction of unwanted spill and cross-bleed in recordings. Once again, RX11’s legacy De-bleed tool is retained as a Classic

Mode option within the new De-bleed control window, and in that mode a reference track can be selected and aligned with the audio to be processed, so that the tool’s algorithm can learn what to remove. The new version of De-bleed no longer requires a reference track, though! Instead, the type of unwanted bleed signal is selected from a list of options (vocal, snare, kick, toms, cymbals, drums, guitar or piano) and the algorithm’s neural-net ‘brain’ identifies and removes just that, with the amount of bleed attenuation adjustable via a De-bleed level control. There are also frequency-range limit controls, to help the algorithm focus on a specific problem area. This new approach to De-bleed processing is remarkably accurate and fast. I suspect, though, that the classic tool might still prove better in situations where the bleed signal is more complex, so its retention is welcome.

Feeling Rebalanced It seems that intelligent ‘stem rebalancing’ tools are all the rage these days, and it’s now a highly competitive sector. Alternative stem-splitting systems like SpectraLayers, Gaudio Studio, Lalal, LANDR, Fadr, Moises, and even Logic Pro’s own native tools, have all advanced significantly over the last year or two, and increasingly made RX11’s facilities appear outdated by comparison. RX12 addresses this: it has largely caught up again in terms of quality and accuracy, thanks to far more advanced and critically trained neural-net

Little AlterBoy More Choices for Your Voices

Unlock a new realm of audio exploration and discover a palette of effects for expressive tonal control that go way beyond pitch-shifting. Little AlterBoy also includes a powerful combination of formant-shifting and tube saturation (and more!) to dramatically alter the texture and character of your vocals and instruments.

“If anything is just too pure and not exciting enough, reach for Little AlterBoy ASAP.” – Andrew Sarlo (Big Thief, Nick Hakim, Bon Iver)

Free 30-Day Trial at www.soundtoys.com

ON TE ST

I Z O T O P E R X 12

The improved Breath Control module — note the ‘ear’ button, which now appears on all modules and allows you to preview what a process is removing from the audio.

processing. This can be seen most clearly in the Music Rebalance tool, which has been carried over from RX11 but with substantial improvements in accuracy, cleanliness and speed, all thanks to radically re-engineered neural nets. In RX12 it is also now usable as a standalone real-time plug-in. Some users may be frustrated that RX12’s Music Rebalance tools still only provide four basic separate stems — it splits out only vocals, drums, bass and ‘other’. In contrast, the cutting-edge of the competition can identify and isolate a wider range of instrumentation. That said, in many cases the four core stems provided by Music Rebalance will be sufficient for typical mix rebalancing requirements and, on the upside, iZotope’s four stems are of noticeably higher quality and precision than RX11 was able to achieve. Obviously, if

more comprehensive reworking is being undertaken, some will find the four stems offered here a frustrating limitation. Looking forward, I’m told that iZotope’s boffins are working hard to develop more individual instrument stem separation capabilities for future editions, but we’ll have to wait and see when the fruits of those labours become available. Supplementing the updated Music Rebalance facility, iZotope have added in RX12 a brand new Scene Rebalance tool. This is designed to apply the same kind of stem-splitting capability to film soundtracks, splitting out dialogue, music and effects tracks, to allow creative rebalancing when required. In my experiments with this, I was very impressed — with both the quality and accuracy with which it achieved those separations. Generally it does a very creditable job in isolating both foreground and background effects from the music and the dialogue. As I mentioned earlier, both Music Rebalance and Scene Rebalance, as well as Dialogue Isolate, all benefit massively from the new Stems View mode, which time-aligns the multiple stem waveforms/ spectrums one above the other within the same spectrogram display, rather than across separate tabs as we had in RX11. This new arrangement makes it so much easier to see what’s going on, and to mute or solo individual stems in a very straightforward and intuitive way to check what sound elements are located where. I really like this new display format. Making time or frequency selections on one stem applies to all by default, but if you need to process an individual stem it can be selected and displayed on its own in the usual way, for independent processing applications. Overall, then, while it might be fair to say that there are even more sophisticated stem-splitting tools available at the real cutting-edge of this technology, few are as well integrated or as easy to work with as the latest RX12 versions, which are definitely more accurate and capable than anything we’ve seen previously in RX.

Impressions

The improved, neural-net‑based De-bleed tool no longer requires a reference track to reduce spill (the older version remains available, should you need it).

112

July 2026 / www.soundonsound.com

We’re used to each new version of RX bringing some ‘Wow!’ aspects, and in this regard RX12 doesn’t disappoint. Once again, iZotope have enhanced the platform’s premium tools in genuinely

Pricing There are three versions of RX, whose tools and functions are listed in the separate Version Comparison table. There are also various upgrade paths between versions and from previous editions such as RX11. Below, I’ve listed the pricing of each version, and some of the main upgrade paths. (Prices include VAT.) Full Price RX 12 Elements: £95 RX 12 Standard: £389 RX12 Advanced: £1299 Updates RX 11 Advanced to RX12 Advanced: £259 RX 11 Standard to RX12 Standard: £119 Versions of RX are available in some of iZotope’s larger bundles, including the Everything Bundle (£2899) and RX Post Production Suite 9 (£1699). There are also loyalty discount upgrade prices for owners of all previous versions of RX and other iZotope products.

beneficial ways, with significant steps forward in many areas. Ignoring, for a moment, the limited number of stems identified in the Music Rebalance tool, I’d say that RX12 Advanced has regained its premium position as a most impressively comprehensive and capable audio repair toolset. Of course, there’s always a ‘but’, and in this case, it’s that the flagship RX12 Advanced edition is a seriously expensive investment, and a huge step up in price from the Standard edition. Even at the full price, though, it’s still good value for money if you have need of many of its 90-plus skilfully designed, highly effective and impressively usable tools and plug-ins. Broadcast and film professionals, then, will no doubt justify the cost quite easily. But music producers, content creators and other occasional users may well struggle to fund a purchase outright, and I suspect that many existing users may choose to skip a generation between upgrades, to help manage the costs, while ensuring they receiving more substantial updates. For long-standing RX customers, though, the upgrade cost from RX11 Advanced is much more affordable, and there are often attractive upgrade offers for Standard Edition owners too. Price aside, this is a very capable package. £ See pricing box. W www.izotope.com

Discover SOS For Artists : A Complete Platform for Modern Music Creators

Your end-to-end music production suite Designed to support independent artists, producers, and collaborators at every stage of the music-making journey, SOS For Artists brings together all the tools needed to take your music from idea to release.

Create 1. Sample Selection Made Simple Over 3 million royalty-free samples growing monthly. 2. Instruments & Synths Includes a curated collection of high-quality plug-in instruments from brands like Spitfire, UJAM, Arturia, GForce, Sonuscore, and more. 3. Stem Extraction with AudioShake An integrated stem-splitting plug-in allows you to isolate vocals, drums, bass, or instrumental parts. 4. Starter DAWs Included For newcomers, the subscription includes lite versions of Steinberg Cubase and Ableton Live.

Produce 1. Mix & Effects Tools The effects collection includes EQs, Compressors, Reverbs, Delays and Saturators from trusted names. Including Eventide, Arturia, Baby Audio, IK Multimedia, and Synchro Arts. 2. Mastering — Fast, Intelligent SOS For Artists includes access to the advanced AI mastering engine from LANDR.

Distribute 1. Global Digital Distribution Distribute your music to over 150 platforms including Spotify, Apple Music, Tidal, Amazon, and Deezer. Or, if you use these platforms transfer all tracks to SOS For Artists without losing data. 2. Monetization & Royalty Splits Retain 100% of your royalties or use the Royalty Split feature and divide the earnings with your collaborators.

FREE PLAN

ESSENTIALS

ign up and explore S the features Kickstart your creativity with 7,800 samples Preview the LANDR Mastering tool Preview Distribution

or £12.99** per year per month £ 1,000 worth of Tools inc 18 Plug-ins 1200 Sample Credits U nlimited MP3 + up to 36 WAV masters U nlimited Distribution

ONLY

99*

* Prices exclude VAT. Taxes will be applied at the UK / EU Country rate at checkout. ** Monthly price does not offer same number of Plug-ins or Sample Credits *** Also available in most countries worldwide in your local currency.

3. YouTube Integration Link your official YouTube artist channel to keep distributed content and subscriber base in one place. 4. Promotion & Reporting Promotion tools include Linktree-style landing pages and content creation tips to help boost visibility. Detailed royalty reporting shows trends, top countries, and top-performing platforms.

PRO

199*

or £19.99** per year per month £2,000 worth of Tools inc 64 Plug-ins 2400 Sample Credits Unlimited MP3, WAV + HD WAV masters Unlimited Distribution

ONLY

www.sosforartists.com

ON TE ST

Akai MPC Sample

Sampling Workstation Akai recapture the magic of MPC with a powerful, portable groovebox. SIMON SHERBOURNE

hen I first saw the MPC Sample I laughed out loud... in a good way. Rumours of a ‘mini MPC’ were going around, which I’d assumed would be like an MPC One but even smaller. The new MPC has that rare design property of showing you exactly what it is at a glance. My laugh of delight was realising Akai had channelled the spirit of the original MPC60 at an adorable scale, and that it was going to be very much its own thing rather than a new incarnation of the modern MPC OS. The question was: would it be a novelty collectors’ item, or a useful, entry-level beat machine?

114

July 2026 / www.soundonsound.com

The Sample is much smaller than it looks on YouTube. While it is, in fact, solid and quite hefty, it has an ’80s handheld game vibe, like a massive old MPC shrunk to dimensions somewhere between a Teenage Engineering KO2 and Roland SP‑404 — references we’ll come back to later. The whole pad grid would fit comfortably inside a 3x3 section of a regular MPC’s grid. The deep pads feel and respond much like the previous generation of MPC’s pads before the new MPCe design, which is to say fantastically. Another clue that the Sample means business is the screen. While only about three fingers wide, it’s high-res and multicoloured, supporting detailed views

with small fonts and waveforms. It’s far superior to what the machines we already mentioned offer, especially the KO2, and it’s even better than the display on Roland’s TR-1000.

Sample-centric The MPC Sample doesn’t have a track-based structure like its bigger siblings, and it’s a purely sample-based instrument. More akin to the SP-404, you have one sample per pad to work with, which you sequence via multiple patterns. There are eight banks of 16 pads. In the default Sample mode you can play the pads and record patterns in real time. Settings, controls and waveforms

Akai MPC Sample £339 pros • Straightforward. • Massive sampling time for a portable. • Synchronised sampling. • Fun and effective performance effects options. • Project interchange with other MPCs and MPC Desktop.

cons • Limited melodic sequencing. • Locked-out while sampling or playing a Song. • Limited step sequencing.

summary A super-mobile MPC with simple and direct access to classic sample and chop beat‑making.

for the most recently tapped pad are displayed on the screen. Three buttons above the screen move through various control sets, such as sample trimming, tuning and filter controls. Editing is done with the three knobs below the screen. As you move around, the knobs are rarely in sync with the parameters they control. Spec’ing knobs instead of encoders may have been a budgetary consideration, but it also feels like a nod to the 404. In the initial release, constantly having to twist the knobs to pick up the control position was my least favourite thing about the Sample. Thankfully the 1.3 update added other options, with the default being a scaled takeover. I’ve got mine set to instant takeover. The general layout is really easy to grasp. A mode section jumps you between the regular sample triggering mode, sequence launching and the two different styles of effects. The blue buttons provide pad modes that we’ll dive into. Sample Select opens the sound browser. Then there’s a transport section with separate Record buttons for audio

sampling and sequence recording. All other operations and menus are accessed via the pads with Shift.

Getting Stuff In A new project presents a blank slate of empty pads to populate. A whole set of ‘kits’ come pre‑loaded, which are themselves simply pre-configured projects, as a project is essentially a single drum program in the MPC way of thinking. To load a sample to a pad, you simply tap the Sample Select button and browse the internal drive or an external SD card (up to 1TB) using the push encoder. There’s a load of good stuff already on board — I didn’t add any of my own drums, just some loops and melodic samples. However, the Sample really is built to sample. Press Sample Record and the screen tells you to tap a pad to record, and provides options for source, monitoring, length and trigger threshold. Sources are the built-in mic, the rear quarter-inch inputs (individually or stereo), USB (direct iOS device connection is supported), or resampling of the mix bus. In Free mode you can record up to 20 minutes thanks to disk streaming! Tapping different pads redirects recording on the fly. Whether loaded or sampled, once a sound gets into a pad you can adjust start, end and loop points, a simple attack/ decay envelope and velocity sensitivity. You can shift pitch, and even time‑stretch samples from half to double speed, or match the tempo to your sequence. The multi‑mode filter has a really nice, sticky resonance character. Again the envelope is attack/decay, with a depth control. The stage lengths on both envelopes are relative to the length of the sample and anchored to the start and end points, so they interact strongly with the trim controls. A massive plus is Seq sampling mode, which synchronises sampling to playback to capture a perfect loop matching the

current sequence’s length/tempo. This is so often missing in the release feature set of samplers. What is on the wishlist, however, is a way to have a captured loop autoplay after record. You can do the Octatrack hack of placing a trigger on the first step ahead of time, but this currently doesn’t trigger on the first pass after record. (The TR-1000 had a similar issue). If this could be automated, the Sample would become a decent looper and quick layering machine. Lastly, the Sample has a really cool always-on resampling buffer that keeps a rolling record of the last 25 seconds of sound coming from the machine. At any time you can hit Recall and it will drop this onto an empty pad.

Pad Modes Of the four pad modes two are straightforward: Mute turns the pads into a mute grid; Loop toggles looping on/off for the selected pad. Chop and 16 Levels deserve a deeper look. Chop puts you into sample‑slicing mode for the selected pad, and will be familiar to users of other MPCs and most other devices that have adopted the same conventions. Chopping schemes available are Threshold (transient detection), four, eight or 16 even regions, or Manual. Manual mode is the classic ‘lazy chop’: pad 1 triggers playback, and subsequent taps in the empty pad area instantly create new slices by ear. Editing is easy after the fact, and you can zoom in on the waveform to neaten things up. And that’s it, no further steps are required unlike the other MPCs. You can access your chops any time just by selecting the parent pad and hitting Chop. I expected the compromise would be that all the chops would share a voice but nope, pads can be set to Poly mode and act like a kit. Time‑stretching is possible post-chopping — great for

On the back panel are 3.5mm jacks for headphones, sync out (TRS CV and gate) and MIDI I/O, quarter-inch jacks for audio I/O, a record gain knob and a USB‑C port.

www.soundonsound.com / July 2026

115

ON TE ST

MINI REVIEWS

of the Big Muff Pi, Ram’s Head and Russian variants, so it’s really three pedals in one. The circuity varies between each type, and even within the same type at different times. To my ears, the so-called Ram’s Head model has the most focused sound, while the other two sound fatter. With straightforward Volume, Tone and Sustain knobs it’s easy to dial in. A switch at the lower right of the GUI is used to switch between versions. The Small Clone chorus pedal used charge-coupled delay lines, sometimes known as bucket brigade chips, along with LFO modulation of the delay time and a fixed wet/dry mix. This was essentially a simplified version of the Clone Theory pedal, which I used as part of my own setup back in the day. In fact, I rather wish MixWave had recreated the Clone Theory instead of the Small Clone, as it has more adjustment! Here, you get a single Rate control and a switch for two different Depth settings. Nevertheless, the Small Clone produces a lush chorus effect, and the original’s analogue nature introduced subtle variations — modelled here — that contributed towards its musicality. The Deluxe Memory Man delay pedal also used charge‑coupled delay lines, and that helps the delay repeats to become progressively less distinct, in a natural-sounding way. EHX also added modulation capability so the pedal can

The Struggling Musician’s Guide To Earning Money by Adonis Aletras Book Making a living as a musician has never really been a secure or stable career choice. For all the creative or emotional satisfaction it can bring, it’s not always a guarantee that bills are going to get paid, the roof stays over your head, or that there will be food on the table. In a compact 83 pages, author Adonis Aletras’ book aims to get you thinking about squaring that particular circle, distilling some core ideas from his many years of experience of doing just that in his own musical journey. Part 1 starts with a pretty stark but undeniably accurate discussion of why, in isolation, streaming numbers don’t make a good business model. With that reality acknowledged, Part 2 succinctly encourages an exploration of how your

120

July 2026 / www.soundonsound.com

combine echo with chorus, vibrato or subtle pitch drifts. There are the expected controls for Delay time, Feedback, Blend (wet/dry mix) and Level, with a further knob controlling the Chorus/Vibrato effect, as selected by a switch at the top of the case. The bypass LED is augmented by a second LED that indicates Overload, and an Ext Delay switch in the GUI allows for up to 3000ms of delay time instead of the basic 550ms maximum. At the shortest delay times the modulation can be used to create

skills and areas of expertise drawn from your musicianship can be transferable, and therefore used to develop alternative or multiple income streams. These may well have music at their heart but, just as importantly, they can offer sources of income that then provide the freedom and security to pursue your personal music ambitions. Amongst other things, these include things like sync licensing, remote work for hire as a musician, providing technical audio services, or teaching either in person or online. Part 3 focuses on basic strategies for staying afloat financially given the many uncertainties of working within the music industry, while Part 4 outlines some initial strategies that help you take the required action and move towards a more stable pattern of income without losing sight of your musical ambitions. Here, as elsewhere, there is common-sense advice and encouragement to think strategically about

chorus and vibrato-style effects without obvious echoes. I recall the Electric Mistress flanger with great affection, probably because, after distortion and a wah‑wah pedal, it was probably the first effect pedal I bought. Its flanging sounds still stacks up with the best of today’s pedals, though its analogue circuitry, again based on a charge-coupled delay line, was notoriously noisy. The pedal was used by some to create chorus-like effects, but turn up the feedback knob (Colour) and it could deliver a screaming jet flange. At faster modulation rates and lower Colour and Range settings it could emulate a rotary speaker, making it another very flexible pedal. In summary, I applaud the sonic accuracy of these virtual pedals. MixWave have managed to recreate the original effects very convincingly while, I’m happy to say, choosing not to replicate the excessive background noise of those early pedals! The three flavours of Big Muff all sound suitably analogue, and if you like your distortion as pristine as possible there’s an option for upsampling too. Available individually or as part of a bundle, these plug-ins are also sensibly priced. Paul White £ $149 (discounted to 109 when going to

press). Individual plug-ins also available separately from $69 ($39) each. W https://mixwave.com

how you build a sustainable path; there’s no waffle, just clear, simple and sensible advice written in a succinct fashion. If your idea of a musical career is built on the possibility of a hit single opening all the subsequent doors, then good luck and fingers crossed for you. However, that’s not really the route to success you’ll learn about here. Nor is this a comprehensive textbook on the more realistic approaches it introduces. What it is, however, is an efficient reality check for anyone — young or old — who is just beginning to work out how they make their musical dream into a pragmatic, sustainable pattern of work. This is simple advice clearly articulated and intended to put some sensible strategy into your musician lifestyle. You could easily consume the entire book in a single extended sitting and, if you are just starting to build a career for yourself centred on your skills as a musician, that would be a couple of hours very well spent. John Walden £ Paperback £10.38, Kindle £3.70. W www.amazon.co.uk

Capture with Clarity Capture sound with precision and artistry. The AKG C-Series condenser microphones deliver studio-grade clarity for speech, music, and ambient environments—engineered from decades of acoustic expertise to elevate every recording. Discover the difference at akg.com.

INTRODUCING

MODERN ACOUSTICS

LOW NOISE FLOOR

WIDE DYNAMIC RANGE

ECO-FRIENDLY DESIGN

AKG Professional is exclusively distributed in the UK and ROI by Sound Technology. For more information please call 020 8962 5080 or visit soundtech.co.uk

C-Series

C104 | C114 | C151

SPOTLIGHT

Focusrite ISA C8X

Expandable Eight-channel Interfaces ADAT-equipped audio interfaces are a blessing if you need more channels, but the choice can be overwhelming. We round up some of the best eight-channel offerings. LUKE WOOD

hoosing the right audio interface is one of the most crucial decisions for any studio, and the amount of options that are available mean it can be an overwhelming task. Deciding how many inputs you require is an important step, and many people recommend opting for slightly more than you think you need! Studios tend to grow over the years, and you may find yourself needing to record larger ensembles, add some line-level connections for synths and drum machines, or have some additional preamp ‘flavours’ ready for action without any patching in. Thankfully, most interfaces now include digital I/O options that make it simple to add additional inputs as your needs increase, and by far the most common is the humble ADAT format. Originally introduced by Alesis to carry digital audio into their multitrack tape recorders, the protocol has long outlived

the machines it was designed for and remains a common fixture in the pro-audio world. A key consideration is the sample rate at which you want to work — a single ADAT connection is capable of transferring eight channels at 44.1/48 kHz, four channels at 88.2/96 kHz and two channels at 176.4/192 kHz. Many devices support the S/MUX protocol, which allows multiple ADAT ports to be combined to maintain the channel count at higher sample rates. If you only ever want to work at 44.1/48 kHz, then you’ll always have eight channels available via ADAT; if you need to run at 96kHz, you’ll need two ADAT ports per eight channels rather than one, and running at 192kHz without sacrificing any channels would require four connections. A common starting point for many setups is an eight-channel interface, so in this month’s Spotlight, we’ll take a look at devices that come loaded with eight of their own preamps, and support an additional eight or 16 channels via ADAT.

Antelope Audio Discrete 8 Oryx

122

July 2026 / www.soundonsound.com

Antelope Audio Discrete 8 Oryx Antelope Audio’s Discrete 8 Oryx supports 26 inputs and 38 outputs, and features eight built-in mic preamps that offer up to 75dB of gain — all eight will accept line-level sources too, and instruments can be connected to the first two via front-panel TS sockets. As for local outputs, dedicated monitor and re-amp outputs are provided alongside eight analogue line-level outputs (they’re DC-coupled, and so can be used to control modular synth gear) and four independent headphone mixes that offer adjustable crossfeed. Antelope are renowned for their clocking expertise, and they’ve equipped the Discrete 8 Oryx with three word clock outputs that can be used to sync other digital gear. The interface includes onboard DSP that hosts Antelope’s range of analogue-modelling effects, and comes with companion software that provides control over routing, low-latency monitoring and the DSP plug-ins. In terms of expansion, there are two ADAT inputs and outputs, meaning the unit will accept up to another 16 channels of I/O at base sample rates. £ £1030 including VAT. W www.antelopeaudio.com

Audient EVO 16

Fender Quantum HD 8

Audient iD48

Audient EVO 16/iD48

Designed to make recording as quick and simple as possible, Audient’s EVO range introduced a new Smartgain system that automatically sets preamp gain based on the incoming signal level. It’s available simultaneously across all inputs, and promises to get users up and running in less than 20 seconds. The eight onboard preamps offer 58dB of gain, and will happily accept line-level signals too — the first two are also equipped with JFET instrument inputs. Eight line-level outputs are provided on rear-panel TRS sockets while front-panel TRS sockets offer two independent headphone outputs, and the unit also includes a word clock output. Helpfully, the EVO 16 offers some built-in monitoring control, including speaker switching, mono summing and talkback, all of which can be controlled directly from the front panel. Two sets of ADAT I/O allow users to add up to 16 additional channels in both directions, and Audient have some potential companions in their own range: the ASP 880 and 800 each deliver eight preamps based on the company’s console circuitry, while the EVO SP8 houses another eight Smartgain-capable preamps. There’s also the flagship model of the iD range, the iD48, which is built around eight of Audient’s praised console preamps. It too offers 16 additional channels via ADAT, and one of its headline features is the inclusion of per-channel switchable insert points that can be used to easily integrate outboard gear during both tracking and mixing. £ EVO 16 £399, iD48 £749. Prices include VAT.

Avid Pro Tools Carbon

Described by Avid as “the closest thing to zero latency”, the Pro Tools Carbon aims to bring the performance of the company’s flagship systems to project and home-studio users. It’s based on the same technology as HDX systems, and employs a Hybrid Engine architecture that makes it possible to switch Pro Tools tracks over to DSP Mode, where their signal path is transferred to the Carbon’s DSP and can be monitored with just 1ms of latency — plug-ins that are available as both AAX DSP and AAX Native will be moved seamlessly between the Carbon and Pro Tools’ own audio engine. Along with its eight mic preamps, the unit features eight line inputs that bypass the preamps, and offers four independent headphone outputs along with a dedicated stereo monitor output. The first two input channels will also accept instrument signals. Four Toslink sockets then provide 16 channels of additional inputs and outputs over ADAT. £ £3399 including VAT. W www.avid.com

Recently rebranded from PreSonus to Fender, the Quantum HD 8 has been designed to provide users with a convenient solution for full-band recording sessions. Its eight built-in mic pres are kitted out with an automatic gain function, deliver 75dB of gain and will also accept line-level signals, while a set of 10 line-level outputs include a dedicated main monitor path — those working with CV-equipped gear will be pleased to know that the outputs are DC-coupled, and in an increasingly rare move, the unit’s audio connectivity is complemented by MIDI I/O. Around on the front panel you’ll find a pair of instrument inputs, each of which is joined by a dedicated re-amp output, as well as a pair of headphone outputs. Along with the local controls, the Quantum HD 8’s preamp gain, phantom power and output routing can be controlled using the company’s free Universal Control app or Fender Studio Pro — the interface is supplied with a perpetual licence for the latter. Two pairs of ADAT I/O are present, meaning the Quantum HD 8 can be expanded with an additional 16 inputs and outputs. £ £879 including VAT. W www.fender.com

Avid Pro Tools Carbon

Fender Quantum HD 8

W www.audient.com

www.soundonsound.com / July 2026

123

SPOTLIGHT

E X PA N D A BL E E IGH T- C H A N N E L IN T E R FA C E S

Focusrite Scarlett 18i20

MOTU 10pre

Focusrite Scarlett 18i20 / ISA C8X Sitting atop Focusrite’s hugely popular Scarlett range is the 18i20, which employs the same converters that power the flagship RedNet series. Its eight preamps include analogue Air Presence and DSP-based Harmonic Drive modes inspired by vintage studio consoles, and are joined by 10 line-level outputs (four of which can be used as Main and Alt monitor outputs for built-in A/B switching between two playback systems) and a pair of independent headphone outputs. There’s an Auto Gain function that makes quick work of setting input levels: hit the button, play for 10 seconds and the interface will adjust the gain accordingly across all eight preamps; the signal level is continually monitored, and a Clip Safe function can be engaged to automatically back the level off during takes to avoid distortion. All of the inputs also accept line-level sources, and the first two are equipped with Hi-Z instrument inputs. Up to 16 extra inputs and outputs can be added via ADAT. The recently released ISA C8X also features eight built-in mic pres, two of which are Focusrite’s praised ISA design. Equipped with Lundahl LL1538 transformers, they deliver the classic tone for which the ISA range is renowned while offering some additional character thanks to new, all-analogue Console and 430 Air modes. The C8X offers insert points for the two ISA channels as well as 12 line-level outputs, and will accept up to 16 additional inputs and outputs via ADAT. Both interfaces can be controlled

124

July 2026 / www.soundonsound.com

remotely using the Focusrite Control 2 app, which handles everything from routing configuration and monitor mixes to preset saving and recall. £ Scarlett 18i20 £649.99, ISA C8X £1899.99.

ADAT, and also includes 128 channels of AVB-based network I/O. £ £1649 including VAT. W www.motu.com

W www.focusrite.com

While best known for their large, modular converter systems, Prism Sound also offer a range of high-end audio interfaces. The Atlas employs the latest generation of the company’s A-D and D-A converters, combining them with a set of eight mic preamps that each feature the company’s Overkiller soft-clip limiter circuitry and eight line-level analogue outputs. Two instrument inputs are then provided on the front panel along with a pair of independent headphone outputs. Local control is provided for the Atlas’ most commonly used features, with more in-depth control over the likes of routing and low-latency monitoring carried out via a companion desktop software application. As well as functioning as a USB interface, the Atlas can be fitted with Prism’s Dante or MDIO-HDX expansion cards, allowing it to be integrated with Dante networks or Avid’s Pro Tools HDX and HD Native rigs. As for I/O expansion, the unit offers an additional eight inputs and outputs via ADAT. £ £3994.80 including VAT. W www.prismsound.io

Prices include VAT.

MOTU 10pre

Seeing as the 8pre is no longer in production, it seems only fair to include the 10pre, which bends the rules here slightly with 10 onboard mic pres. MOTU opted for placing eight channels on the rear panel so that they’re easily accessible for rack-based setups, with two more on the front panel for ‘on the fly’ access during sessions — all 10 inputs also accept line or instrument signals. The first two channels feature insert points for integrating external gear into their signal path, and the unit offers eight DC-coupled line outputs as well as two dedicated headphone outputs. An onboard 64-channel DSP mixer handles routing duties, with all input and output busses boasting four-band EQ and compression, while input channels are treated to a high-pass filter and gate. Helpfully, a set of front-panel switches offer switching between up to three pairs of monitors along with mute, mono and talkback functions. The 10pre can be augmented with 16 additional channels of I/O via

Prism Sound Atlas

SSL 18

SSL are famous for their large-format consoles, but they’ve steadily introduced a whole host of new studio gear in recent years, including a family of audio interfaces. The largest, the 18, includes eight preamps that will accept mic and line-level sources — the preamp circuitry is completely bypassed when selecting a line input — while the first two can be switched to Hi-Z instrument mode. The preamps are designed to be clean-sounding by default, but those who want some ‘grit’ can take advantage of a 4K mode that introduces some analogue saturation inspired by SSL’s 4000‑series consoles. Thoughtfully, there’s also a dedicated talkback mic input that avoids the need to sacrifice any of the main inputs. There are 10 line-level outputs, four of which can be switched between +9 and +24 dBu alignment levels, and outputs 9‑10 can be repurposed as insert sends for the first two input channels. All of the line-level outputs are DC-coupled, and the 18 also includes MIDI and word clock I/O. Two headphone outputs are provided on the front panel, and as well as offering three modes that optimise them for different headphone sensitivities, they can also be configured as additional line-level outputs. A front-panel monitor level control is joined by Cut, Alt and Talk buttons, with more in-depth control and routing taken care of by the SSL 360 software application. As for expansion, the 18 is equipped with two ADAT inputs and outputs, allowing users to add up to 16 additional channels in each direction. £ £899.99 including VAT. W www.solidstatelogic.com

SSL 18

Universal Audio Apollo x8p / Volt 876

Universal Audio have a couple of options in their range. The first is the Apollo x8p, which packs in eight of their Unison preamps — an innovative design that modifies the input circuitry and is capable of emulating preamps and channel strips from the likes of Neve, Manley, API and others. Onboard DSP makes it possible to track through and monitor the company’s UAD analogue-emulation plug-ins in real time, and the unit can be purchased alongside various plug-in bundles. All eight channels accept mic or line-level inputs, and the first two feature front-panel Hi-Z instrument inputs; dedicated main monitor outputs are joined by eight line-level outputs,

and the unit offers a pair of independent headphone outputs. Dual ADAT I/O then allows for up to 16 additional channels. A newer addition to the UA range is the Volt 876, which offers a broadly similar I/O count, but without Unison preamps or onboard DSP. Its eight preamps feature a Vintage mode that’s inspired by UA’s 610 valve console, and are each joined by a compressor that’s based on the 1176. Once again, up to 16 additional inputs and outputs can be hooked up via ADAT. Routing, configuration and low-latency monitoring for both interfaces can be carried out using the company’s UAD Console software. £ Apollo x8p £3499, Volt 876 £859. Prices include VAT.

W www.uaudio.com

Universal Audio Apollo x8p

Universal Audio Volt 876

www.soundonsound.com / July 2026

125

Studio Pro

TECHNIQUE

There’s more than one way to comp the perfect take in Studio Pro... ROBIN VINCENT

omping is probably the reason why God created the DAW. She was so fed up with all that tedious tape splicing just to nail a vocal within the allotted studio time that the great creation refineries were stoked up and put to work crafting the binaries to enable digital audio editing. In these enlightened times, comping is so embedded in the modern workflow that it’s hard to believe anyone could get frustrated with it. So, what is comping and how does Fender Studio Pro handle it? Comping, or compiling, is the process of stitching together a fantastic performance from the best parts of several other performances. It’s building a single perfect take out of a number of takes, to create the illusion that the artist performed the entire part perfectly. It can be used lightly to correct a bum note, or wildly to dissect every take into syllables and rebuild the whole track piece by piece. What’s for sure is that it’s made finding that perfect performance so much easier, and you can worry about being authentic on the live tour. It’s most commonly used with vocals, where you might loop a section of a song and record multiple takes from which you can choose the best, or choose parts of several takes to form the final take. It can also be used for guitar, piano, or any instrument that you’re recording as audio, and Studio Pro makes it dead easy to work with. Studio Pro always has more than one way to accomplish something, and when it comes to comping there are two systems in play: Takes and Layers. Comping always

Screen 1: By default, if you record multiple takes on the same track, Studio Pro stores them as Takes and lets you choose which one to play.

happens in the Layers, but recording Takes is usually how you end up in Layers in the first place. You can skip straight to Layers if you prefer to do your Takes in Layers rather than taking Takes into Layers later. Confused? Let me explain.

Takes Recording Takes in Studio Pro is a natural consequence of how recording works in this DAW. Left to your own devices, you’d probably do this without thinking and may not even know it’s happening. Whenever you record a segment of audio in Studio Pro, it is called a Take. However, that’s never important until you record another one over the top. Recording on a track over the top of audio that’s already there will generate another Take. While you’re doing it, it looks like the audio underneath is being overwritten. And if that’s your intention, then you can consider it so. However, each time you overwrite audio on a track, Studio Pro is saving the original audio as a Take and calling the new recording Take 2. Do it again and you get Take 3, Take 4, and so on. Nothing is overwritten, nothing is destroyed, it’s all kept as Takes. The great thing is that you

can go back and swap the earlier Takes back in place of the latest one. This is like the first level of comping, where you can choose the take you want to use in a section of your song. You are not bound to the final performance; you have options. The easiest way to record multiple Takes is to set up a loop and let Studio Pro continue recording as it loops. You can drag in the timeline to set loop points, and press the / (forward slash) key to enable the loop. Ideally, give yourself a bit of space before you start playing, so set the loop a bar or two before you actually want to start recording. Turn on the click if you need it. Hit Record and your song will loop between the markers, and you can keep recording until you feel you’ve done what you need. Your track will look like it’s just recorded the last pass, but if you click on one of the tiniest buttons you can find in the bottom left of the recorded clip, you will be able to select which take you want in the track (Screen 1). You can, theoretically, cobble together a comped Take from all the other Takes, but you’d have to copy tracks, switch Takes, slice out parts, mute parts, and put it all back together. We have tools that make this a lot easier. But first, let’s look at the more common way to do comping.

Layers

Screen 2: A more sophisticated alternative to the Takes system is to record each take to a new Layer.

128

July 2026 / www.soundonsound.com

Studio Pro does its comping through the power of Layers. They differ from Takes in that they are somewhat more intentional. You do get the feeling that Layers and Takes will get rolled together into the same feature set at some point, but at present they are separate things. I’ll show you how they can work together in a moment, but first, let’s do comping properly with Layers. This time, we’re not setting up a loop; we’re just focusing on the section of our

Screen 3: The Expand Layers button will display all of your Layers, allowing you to build the best take from all of them. Highlighting a section in any layer promotes it to the main track for playback.

song where we want to put together a performance. The way this works is in a more record/rewind/re-record kind of way. So, set up your track for recording and record your first take. I appreciate it can be confusing to call it a take considering we’re not doing a Take as such, but that is also what it’s called — bear with me. Your first recording was fine, but you think you could have done better. You could do that by creating a new track and recording into that; however, Layers are much more interesting and are the key to comping in Studio Pro. So, in the track header, click on the Layers drop-down menu and select Add Layer (Screen 2). The Layers menu can be found beneath the Record Input and Group settings. A box pops up for you to name your Layer if you wish. Click OK and your recorded clip vanishes. Don’t worry, it’s hidden in another Layer. For the moment, we’ll do another take or two, and each time repeat the process of creating a new Layer, and then we’ll go and find them. Now, there’s another very similarly tiny button you can find in the bottom left corner of the track header. The bottom one expands automation lanes, but the one above it expands the Layers. Under the track, you should now be able to see

the other takes you recorded on individual Layer tracks (Screen 3). This is where we can do the comping. Only clips in the main track play back. To place clips in the main track, we need to highlight them in the Layers. As you highlight bits of your takes, those sections are automatically mirrored up on the main track. You can swap and try different bits out just by highlighting them. You can also swap between Layer parts using the little button in the bottom‑left corner of the clip, and it will replace it with a same-sized piece of another clip. Studio Pro automatically crossfades the clips together and lets you adjust the transitions to ensure it’s all as smooth as possible. You can audition a Layer without promoting it by using the S (solo) button on its header. Or you can place the entire Layer up into the main track by clicking the little up arrow. You keep testing and trying various

aspects and parts of each performance until you have the perfect combination to fit what you need for the track. Hide the Layers away again, and the track is there with the perfect take, right in your project.

Takes To Layers So, wouldn’t it make perfect sense for looped takes to be generating Layers rather than more Takes that you have to manually switch between? Yes, it would, and this is already possible in two ways. The first way is a button hidden away in Studio Pro that turns on a feature called Takes to Layers. And it does exactly what we’re talking about. To find that button, you need to look down at the transport bar and find a little panel that has some sync options and pre‑roll settings, and you’ll find a little up arrow that pops up with the words Record Panel. Click the arrow, and a row of buttons that you’ve never seen before appears above the transport. One of those buttons, under Record Mode, is called Takes to Layers (Screen 4). Enable it, and now every time you record Takes with a loop record, every new take creates and occupies a new Layer. Open that up, and you have a history of every performance you gave, ready for you to choose the best bits from. Why this is not enabled by default is anyone’s guess. The second way to do this, and this is particularly useful if you can’t find the Takes to Layers button, is to right-click the current Take and choose ‘Unpack Takes — Unpack Takes to New Layers’. It will instantly throw the Takes across Layer tracks under your main track, ready for comping.

Finishing off Once you have your perfect comped take, you might want to turn it into a single clip to make it easier to deal with. This is easily done by selecting all the clips in the main track and pressing Ctrl/Cmd+B to bounce the clips into a single clip. It’s also available in the right-click Event menu. But here’s a tip: before you do that, duplicate the Layer and then hide them all away, so if you change your mind, you have the original comp to come back to.

Screen 4: The Takes to Layers button in the transport bar changes Studio Pro’s default behaviour, so that new Layers are automatically created for each take.

www.soundonsound.com / July 2026

129

Pro Tools

TECHNIQUE

Getting the key, tempo and structure of your songs right can make a world of difference. JULIAN RODGERS

Set the Elastic Audio mode of audio tracks to Polyphonic, and they’ll change pitch alongside MIDI and instrument tracks when you use the MIDI Operations Transpose function.

Anyone who has spent time playing in bands will know that while instrumentalists may have preferences about which key a song should be played in, ultimately the correct key is the one that best suits the singer’s voice. Very often, songs are written before a vocalist has fully lived with the arrangement, and it’s extremely common to discover that the song simply sits better a tone lower or a semitone higher. Finding the right key can completely change the confidence, tone and emotional impact of a performance.

If a song has been developed entirely using MIDI, changing key has never really been a problem. One of MIDI’s biggest advantages is that pitch can be changed freely and without any loss in quality. Audio tracks can be pitch-shifted in real time using Elastic Audio, though pitch‑shifting always affects audio quality so I’d always recommend re-recording in the new key if possible. Fortunately, there is a unified way to transpose an entire session containing both MIDI and audio tracks. Before transposing audio tracks, Elastic Audio must first be enabled on every track you want to pitch-shift. In most sessions, that will mean excluding drums and percussion. Elastic Audio is enabled from the Edit window track header using the Elastic Audio / ARA selector located directly beneath the automation mode selector. To enable it across multiple tracks at once, select all required tracks and Option+Shift-click (Alt+Shift-click on Windows) the desired Elastic Audio mode. Not all Elastic Audio modes work with this workflow. Varispeed mode won’t function using this technique. Monophonic mode also doesn’t work, despite seeming like the obvious option for sources such as lead vocals or bass guitar. In practice, Polyphonic mode is the safest choice when auditioning alternate keys across an arrangement. With Elastic Audio enabled where required, the next step is to select everything you want to transpose. The simplest way to do this is often to enable the All group in the Edit Groups list, place the insertion point after the final clip in the session, then use Shift+Return to extend the selection back to the start of the timeline. Now go to Event / MIDI Operations / Transpose, or use the shortcut Option+T on Mac or Alt+T on Windows. This opens the MIDI Operations window with the Transpose section displayed. At first glance, this might seem like the wrong place to look when working with audio tracks, but despite the name, this process can transpose MIDI and Elastic Audio-enabled audio simultaneously in a single operation. Enter the number of semitones you want to shift the arrangement by and click Apply. If you’re still searching for the best key, simply move up or down

ro Tools is often thought of primarily as a production environment, but it can also be an extremely powerful songwriting and arrangement tool. Many songwriting problems are actually arrangement problems, and one of the most valuable things a DAW can offer is the ability to experiment freely without damaging the version you already have. That might mean trying the song in a different key to better suit a vocalist, experimenting with alternate section orders to improve pacing, or changing the tempo to alter the energy of a performance. Possibly the single most important of these changes, particularly in vocal music, is finding the best key for the singer.

Get Shifty

130

July 2026 / www.soundonsound.com

...MIDI Operations window.

a semitone at a time until the song settles naturally into the vocalist’s range. The ‘MIDI Operations’ name comes from a Pro Tools 2025.6 UI update. Previously called Event Operations, the rename can make this audio-capable workflow easy to miss.

Home On The Arrange One of the biggest advantages DAW-based writing has over traditional linear recording media is the ease with which arrangements can be radically restructured. There are several ways to approach this in Pro Tools, such as using Pro Tools Sketch, but here I’ll concentrate on rearranging sections within the Edit window. Before making major structural changes, it’s important to create a way back to the original arrangement. Not every experiment improves the song. A good habit is to first select all tracks and duplicate the main playlist before beginning any major arrangement edits. Note that playlists of alternative takes won’t follow your rearrangement. With your safety copy in place, Shuffle mode becomes extremely powerful for arrangement experimentation. In the upper-left corner of the Edit window are Pro Tools’ four edit modes: Shuffle, Spot, Slip and Grid. Most music production

work happens in either Slip or Grid mode, but Shuffle mode is uniquely useful for restructuring arrangements because clips ‘shuffle’ past each other as material is moved around the timeline. That behaviour makes it ideal for quickly testing alternate song structures. It’s always worth properly marking out the structure of the song using markers. Markers can be created in the Markers ruler or be dropped in in real time during playback, which is often the quickest and most intuitive way to sketch out a song structure before refining their positions afterwards. With the insertion point focused in the Markers ruler, pressing Tab moves forward through markers while Option+Tab on Mac or Alt+Tab on Windows moves backwards. When it comes to separating sections ready for rearrangement, it’s usually easiest to work section by section. Double-clicking between consecutive markers automatically selects the entire section between them. You can then extend the selection vertically across all tracks using Extend Edit Down, found under Edit / Selection. The shortcut is Control+Shift+; on Mac or Start+Shift+; on Windows. With the section selected, press Command+E on Mac or Control+E on Windows to separate clips across the current timeline selection. Repeating this process section by section quickly turns the arrangement into a set of discrete structural blocks that can easily be moved around. Before actually rearranging those sections, though, I find it much easier to turn them into Clip Groups. This makes it

far harder to accidentally leave something behind when moving large sections around the timeline. Additionally you can name the Clip Groups (Verse 1, Chorus 2 and so on), which will help as your markers will no longer reflect the song structure. To create a Clip Group, select all clips within the section and choose Clip / Group, or use Option+Command+G on Mac or Alt+Control+G on Windows. Once grouped, the entire section behaves like a single clip, making it easy to drag around in Shuffle mode. Because Clip Groups can be deactivated instantly, there’s very little down side to using them for arrangement work. To ungroup, use Option+Command+U on Mac or Alt+Control+U on Windows. At that point, experimenting with alternate arrangements can become extremely quick. Entire verses can be moved, choruses repeated, bridges shortened or removed, and sections reordered in only a few seconds. Don’t forget that duplication is equally useful here. If you want to hear what happens with a double chorus before the riff, you can try that idea almost instantly without committing permanently.

Tempo Tantrums The final area I want to look at is tempo. Finding exactly the right tempo for a song is absolutely crucial to how that song feels. Small tempo changes often have a much bigger emotional impact than expected. Experimenting with tempo in Pro Tools is straightforward, but remember that any form of time‑stretching introduces some degree of audio degradation.

Mark out your song sections in the Markers ruler, use those to separate your clips, create Clip Groups for each section and set Pro Tools’ Edit mode to Shuffle, and you can easily experiment with your song’s structure by dragging entire sections around en bloc.

MIDI material, of course, has no such limitations and can follow tempo changes perfectly without any loss in quality. The key to making tempo changes work across an entire arrangement is understanding track timebases. In each track header, directly next to the Elastic Audio / ARA selector, you’ll see either a small blue clock icon or a green metronome symbol. These determine whether the track uses sample-based timing or tick-based timing. Only tick-based tracks follow tempo changes. MIDI and instrument tracks are tick-based by default. Audio tracks, meanwhile, need both Elastic Audio enabled and tick-based timing selected. Once Elastic Audio is active, changing the session tempo causes the audio to be stretched or compressed to match tempo automatically. By default, Pro Tools sessions operate in Manual Tempo mode, where a single global tempo controls the entire session. If you open the Transport window, you’ll see the current tempo displayed there, and you can simply type directly into the tempo field to try alternate values. For more advanced arrangements where the tempo changes dynamically throughout the song, Pro Tools also offers the Conductor track. For songwriting purposes, though, Manual Tempo mode is often all that’s required. The goal at this stage is usually not detailed tempo mapping but simply finding the point where the song feels right. One particularly useful trick when working with musicians who think in feel rather than bpm values is Tap Tempo. Simply click on the tempo display in the Transport window and tap the T key repeatedly in time with the pulse you’re imagining. Pro Tools will calculate and update the session tempo automatically based on your tapping. This can be a surprisingly intuitive way to find the natural pace of a song without getting distracted by arbitrary bpm numbers. There are many, many other ways Pro Tools can assist in the writing stage of a project but key, tempo and song structure cover much of the big picture. One final note of caution: as sessions become more polished and complex, major arrangement changes can introduce unforeseen problems, so it’s worth getting these fundamentals right early.

www.soundonsound.com / July 2026

131

Does ORTF have a high‑frequency roll‑off toward the centre?

The 110‑degree spread of the ORTF array aligns the cardioid patterns so that each microphone is -3dB at the centre (55 degrees off‑axis), which avoids the centre being louder than the rest of the stereo spread as the microphone patterns overlap. A problem I perceive here is that no microphone is perfect, and therefore there is significant high-frequency roll‑off as you get off‑axis. Even mics like the Neumann KM184 at 55 degrees off‑axis are down almost 10dB at 16kHz. The Shoeps MK4 is down about 5dB at the same angle and frequency, which might be hard to notice (just 2dB below target), but still there. Am I missing something? SOS Forum post Hugh Robjohns, Technical Editor The answer is yes, potentially. But it’s extremely dependent on the specific mics in use. Large‑diaphragm mics will be much worse in this respect than small diaphragm mics, of course, just because their polar patterns are inherently much less linear with frequency. But, as you have noted,

Our free Stereo Miking guide goes into detail about the pros and cons of ORTF and other spaced, coincident and hybrid stereo mic arrays.

there can be some variation with small-diaphragm capacitor mics (SDCs) as well, particularly in the extreme high frequencies (HF). The real question, though, is whether this issue is perceptively detrimental to the stereo soundstage! That’s obviously a subjective decision and I have to say that it’s not something that has usually bothered me. Yet, if the middle of an ensemble lacks HF due to being off-axis, it’s easy enough to reduce the mutual angle of the capsules and increase the capsule spacing to compensate for the consequent loss in image width. After all, ORTF is not the only defined near-spaced mic array. DIN and NOS both have smaller mutual angles than ORTF, for example, and therefore more HF for the centre of the ensemble. Or you could consider a three-mic array like the Decca Tree. If you’d like to learn more about how ORTF and other stereo arrays can be adapted to suit the source, your mics, and the location in which you’re recording, then I covered this topic in some depth in our recent Stereo Miking eBook, which you can access for free at https://sosm.ag/stereo-miking-guide.

Am I hearing two stages of conversion when listening to monitors with built-in DSP?

I’ve been looking at the ADAM A8H active studio monitors, and it seems the onboard DSP uses an integrated 24-bit/96kHz A-D converter. How does this work if I’m already using higher sample rates in the DAW? Are we no longer hearing the interface conversion? Apologies if this is a stupid question! SOS Forum post Hugh Robjohns If you are feeding the speakers from an analogue output on the interface, then yes, you’re hearing both the interface’s and the speaker’s conversion. The (reconstructed) analogue signal going into the speaker is re-digitised through its own A-D at 24/96, processed in its DSP, and then converted back to analogue for the power amps and their respective drive units.

136

July 2026 / www.soundonsound.com

While it’s true that built-in Monitor speakers such as Adam’s A8H, that have onboard DSP and analogue inputs, do add another stage of A-D/D-A conversion to the D-A of your interface, it’s not really a problem!

If you’re working at DAW sample rates above 96kHz, with the huge analogue bandwidth that permits, the speaker will be throwing everything above 48kHz away — however, since the speaker’s tweeter couldn’t reproduce those frequencies anyway, it really doesn’t matter! If you’re using sample rates above 96kHz to minimise pre-ringing artefacts, then I’m

“If the middle of an ensemble lacks HF due to being off-axis, it’s easy enough to reduce the mutual angle of the capsules and increase the capsule spacing.” sorry to say that the speaker’s converters will have introduced new ones. Elsewhere in this issue, I review the MQA Inspira and Endura apodisation plug-ins, and you can look to those and any similar ones that might emerge in the future to crack that particular nut if it bothers you. If the speakers have a digital input, though, and you feed them from a digital output on the interface, you circumvent both the D-A in the interface and the A-D in the speaker, as well as removing about 0.7ms of latency. Instead, the digital signal will pass through an ASRC (asynchronous sample rate converter), to be converted from whatever rate the interface is working at, to 96kHz for the speaker’s DSP. Inside the speaker, the DSP output is converted to analogue once again for the amps and drive units. I hope that clarifies things for you!

“Since the speaker’s tweeter couldn’t reproduce those frequencies anyway, it really doesn’t matter!”

NEXT MONTH IN

E-mu Systems SP-1200 E-mu’s sampling drum machine defined the golden age of hip-hop. In the next issue of Sound On Sound we look at its history and the remarkable story of the company behind it.

August issue on sale Thursday 23rd July. Available at TG Jones, WH Smith Travel and all good newsagents.

Subscribe at www.soundonsound.com/subscribe

Digital version (no WH Smith reference)

The Advertisers Index

To Advertise in Sound On Sound please contact: chris.cope@soundonsound.com Ableton

www.ableton.com

Oeksound

www.oeksound.com

AMS Neve

✆ 01282 457011

RAD Distribution

www.raddist.com

Arturia Software & Hardware

✆ 0033 438 020 555

OBC

Rupert Neve Designs

www.rupertneve.com

Audient

✆ 01256 381944

Slate Digital

www.slatedigital.com

DPA Microphones

www.dpamicrophones.com

SOS For Artists

www.sosforartists.com

113

Erica Synths

www.ericasynths.lv

Sound Particles

www.soundparticles.com

Fabfilter Software Instruments

www.fabfilter.com

Sound Service UK

www.soundservice.uk

Freqport

www.freqport.com

Soundtoys

www.soundtoys.com

111

Gear Expo UK

www.gearexpouk.com

Source Distribution (AKG)

✆ 020 8962 5080

121

NEXT MONTH IN

102-103

Genelec

www.genelec.com

Source Distribution (Rodec)

✆ 020 8962 5080

GIK Acoustics Europe

www.gikacoustics.com

SPL

www.spl.audio

Harrison Audio

www.harrisonaudio.com

Synthax Audio UK

www.synthax.co.uk

Heritage Audio

www.heritageaudio.com

Telegrapher

www.telegrapher.com

iamReverb

www.iamreverb.com

Toontrack

www.toontrack.com

IBC

Josephson Engineering

www.josephson.com

Trinnov UK

www.trinnov.com

Kenton Electronics

✆ 020 8544 9200

Warm Audio

www.warmaudio.com

IFC

E-mu Systems SP-1200

E-mu’s sampling drum machine defined the w w w . s o u n d o n s o u n d . c o m / November 2016 137 golden age of hip-hop. In the next issue of www.soundonsound.com August issue on sale Thursday 23rd July. Sound On Sound we look at its history and

SHOEGAZE SAM MORTLEY

t was 2015. I was just 21 years old. I was in my university dorm room, delving deep into new and old music, as a typical student does. Being a guitar‑centric dude who was initially into 2000s bands such as Kasabian, Kings Of Leon, Editors, Muse and Feeder, I found the 2010s tough going. Ultra-polished, sexual‑innuendo‑obsessed pop and stripped‑back bland singer-songwriters swept the airwaves. Fortunately, my generation had the Internet at our fingertips. And looking back into history led me to a musical genre that had indie-rock familiarity, whilst opening up a world of blissed‑out ethereal soundscapes that I never knew were possible! Enter shoegaze and dream-pop. My first listens included Ride’s Nowhere, whereupon Internet recommendations led me to Whirlpool by Chapterhouse.

Both of these albums are at the commercial end of the shoegaze spectrum, and were therefore ideal starting points to draw me in. Then came the more challenging and genre‑defining Souvlaki by Slowdive and the legendary Loveless by My Bloody Valentine. Both of these took a little while to absorb and digest, but once they clicked, they really clicked! I then discovered Kitchens Of Distinction, Medicine, Curve, Seefeel and many others. I started to realise that these albums had the common theme of dreamy and otherworldly guitar sounds, and that the vocals were treated differently to how they are in most mainstream music. Mixed low and heavily effected, the singing was not always fully defined or distinguishable. This, in turn, forces the listener to use their imagination and/or to simply feel the vibe of the sounds blended and mashed together. The sonics of the guitars were

“It makes perfect sense that the genre continues to grow and morph in the current day, as the sonic possibilities are now endless.” 138

July 2026 / www.soundonsound.com

being pushed to the point where the resulting noise didn’t sound like a guitar. It was like impressionism and surrealism combined. And Alan Moulder instantly became my production hero. These factors single‑handedly got me interested in the production not just of this style of music, but of all forms of music, and led me to embarking on a Music Production & Sound Engineering BA Hons degree with dBs Institute in Plymouth, where I graduated in 2025. I bought myself a Fender Jaguar and Jazzmaster, and grew my hair out into ’90s curtains whilst sporting a baggy Breton T-shirt to make myself feel the part. My pedalboard filled up with stompboxes, and I even bought myself some old rack gear such as the Roland GP-16 with its FC100 foot controller, an Alesis Quadraverb and a Yamaha FX500, which I discovered were common for the original shoegeezers to record and perform with.

The 2010s was a great time for the revival of the shoegaze genre because, having encountered the classic early ’90s artists, I then discovered bands of my generation such as Airiel, Ringo Deathstarr, Pinkshinyultrablast, Blankenberge, Miniatures, Deserta, Flyying Colours, Malory, Exit Calm and countless others. I was amazed to see how international the sound had become, from Pia Fraus in Estonia all the way to Resplandor in Peru! It makes perfect sense that the genre continues to grow and morph in the current day, as the sonic possibilities are now endless. A style that thrives on new and extreme effects embraces plug-ins, virtual pedalboards, IR processors for reverb and amp sims. And now, with the advent of immersive audio, the ethereal, atmospheric nature of this genre has the potential to be fully realised. Keep on gazing at yer shoes!

Tell Us What You Love! What makes your audio world tick? Tell us, in around 500 words, and we’ll consider it for publication. Please send your submissions for a ‘Why I Love’ column to whyilove@soundonsound.com.

IMAGINE IT. BUILD IT. MIX IT. Superior Drummer 3 gives you the sounds and tools to create the drum mixes you’ve always imagined. Never settle for less.

WWW.TOONTRACK.COM

_AstroLab Avant-garde stage keyboard AstroLab 61 defines a new era for the stage keyboard combining essential performance instruments, true-to-life acoustic sounds, timeless synths and crisp digital landscapes with intuitive controls and an innovative ecosystem - now available in 37, 61, and 88-key formats for keyboardists, pianists, DJs and seamless studio to stage expression.

FREE eBook RECORDING TECHNOLOGY

BASICS &BEYOND SECOND EDITION

Get your FREE digital publication from Sound On Sound This FREE illustration-rich eBook is aimed primarily at newcomers to the subject, but will prove equally valuable to anyone struggling with the complexities of today’s sophisticated recording technology. The purpose of this 170-page guide is to introduce readers to the essential components of a modern recording system and to explain the recording process in an easy-to-follow way, demystifying the inevitable jargon, both as it crops up, and with a comprehensive glossary. Written in the accessible, no-nonsense style of the Sound On Sound team of authors and editors, Recording Technology: Basics & Beyond covers: ■ What to buy ■ Studio setup ■ Computers for audio ■ Audio interfaces ■ Monitoring ■ Acoustic treatment ■ Mic techniques ■ How digital audio works ■ Understanding your DAW software ■ Upgrading your system

■ Software instruments ■ Wiring your studio ■ Plug-ins ■ Recording audio ■ Understanding MIDI ■ Recording vocals ■ Mixing ■ Compressors ■ Equalisation ■ Mastering ■ A.I. - in Music

Don’t miss out! Sign up and share the link with friends and colleagues on social media.

https://sosm.ag/recording-ebook

SOUND ON SOUND UK July 2026