David Kosten on Heaven 17 and mixing in Atmos

> Browse blu-rays mixed by David Kosten on the SDE shop

As promised, producer and Atmos Mixer David Kosten (an artist in his own right, with Faultline) has answered SDE readers’ questions about his immersive work on the Heaven 17’s Penthouse and Pavement and The Luxury Gap, and Dolby Atmos Mixing in general.

I hope you find this informative and allows you to appreciate the amount technical and creative work that David puts into these mixes, and indeed all his Atmos work. You can pre-order the Heaven 17 blu-rays until Tuesday 24 Feb.

Sean O : I remember reading long ago that Penthouse… was mainly recorded on an 8 track machine. That can’t be right, can it?

Maxe: I guess, when you start to work, another engineer already had to do some “homework” by digitizing and separating, maybe pre-preparing stems – or has all, or all after transfering, has to be done by you? 

David Kosten: 8 track tape – yes, mainly! Although as is the case with nearly all songs made on limited track tape formats, what this means in practice is that every inch of tape is crammed with multiple sounds to make up for the lack of track numbers. So track one might have a bass synth in the intro, a Linn Drum cabasa in the verse, backing vocals in the choruses and some echo effects in a middle section, and so on… In the mix in 1981 they’ll have worked through that song section by section, sometimes editing a final master stereo tape together (as doing a whole song mix in one pass wasn’t physically possible – not enough hands!) and given every one of these sounds its own level / pan position / set of effects like reverbs, delays, EQ and compression. All of which need to be assessed and hopefully very closely recreated by the Atmos engineer (me).  

The first part of every Atmos rebuild/mix I do is what feels like playing with a giant jigsaw puzzle. Lining up the multitracks to match the song structure of the final much-loved stereo (often what’s on the multitrack will be completely different – usually less tightly edited and with additional verses or choruses the listener never gets to hear) is step one. There’ll always be some varispeed needed here as well as editing, as tape transfer speeds are never consistent, and sometimes a stereo mix was made to run slightly faster too. If there’s three separate 24-track tapes to use for one song, this is complicated by the transfers always running at slightly different speeds to each other, so before editing the sections to match the master, I have to make sure that the speeds of each tape are aligned. This was the case with ‘Temptation’ for instance. A less enjoyable part of the process is realising that the tape speed is actually inconsistent through a song and needs to be adjusted at various points in the process…

When I look at the ‘Temptation’ – screenshot, is there a full run of the original stereo in the top tracks? Probably the ultimate reference…At 2:30 there seems to be a cut – is that the point where the 12″ was extended (or the album version edited)?

DK: No, there’s no edits in the master stereo file. What you’re seeing on that screenshot at 2.30 is actually audio space – after “lead us not into temptation OOHH” – that fantastic moment of pause before the chorus kicks in again. The top audio track is the final master, and the two below are my stereo mixes. They’re way more dynamic than the remastered ‘limited’ version, so the gap in the waveform is visible at that fraction of a second only in my mixes, and looks like a dark line / edit! I’ve edited the multitracks there so you don’t have 50+ tracks of tape hiss playing over it. For me this is classic pop production, akin to that wonderful “AAHH” in Good Vibrations, where you hear the echo float away on its own (possibly my favourite moment in pop music full stop.)

Robbie-O: Hi David, what is easier to work on is it a) a traditional ‘rock’ album or b) a synth-pop/led album? Many thanks

DK: If we’re setting aside all the technical things that might make a job tricky, and just talking about genre v difficulty, then I’d say that a rock / organic led album would often be harder to pull off. Live drums for a start – recreating a live kit sound is often incredibly difficult, especially as it’ll probably have been made originally by a master of their craft, using all kinds of amazing and often unusual gear.  The most difficult part of making The Luxury Gap in terms of mix matching was the drum sound on ‘Lady Ice’. What was on the multi-track didn’t sound remotely like the final mixed version from ’83, and it took me a long time to get the drums to a place I was happy reflected the original closely enough! I think there were 15 or 16 audio tracks of drums – close mics, room mics, multiple overheads etc. The processing by the original mix engineer was extreme and very clever to get that sound. But beyond mix tech, there’s also the question of how to mix a band in surround vs a synth-led piece. I think this is a divisive issue, and has probably prompted its fair share of bland and uninspiring Atmos mixes, where an engineer either chooses to play it so safe you might as well not bother calling it Atmos, or perhaps the original artist hasn’t quite embraced hearing their music utilising the available space. There’s many thousands of choices to make along the way with these projects, and of course a key part is that final placement of audio in the room. It’s also incredibly easy to make a wonderful, coherent piece of classic stereo audio feel like a bunch of disparate sounds, disconnected and maybe fun/impressive at first but ultimately lacking the intensity and emotion of the original stereo. And easier to do this I think with a traditional ‘band’ where everyone is used to a ‘front’ presentation live, and where splitting the music into elements above, behind and all around you needs a very steady hand…

Lord Flashheart: I wonder why some artists are hesitant to have their catalogue meddled with (Talk Talk to think of one) and others that would/could really exploit don’t (Jean Michel Jarre).

DK: Now we’re delving into why I decided to get into this kind of work in the first place. Imagine being a classic artist whose work was made primarily in a pre-computer age. And then you are played new versions of your songs which sound like thin, poorly and quickly mixed versions of the songs you probably spent weeks mixing, working with the most elite talents and studios at the time. You’d probably be hesitant too. I was listening to some of my favourite music ever getting butchered by engineers who presumably didn’t have the time or inclination to really put the necessary graft into their Atmos rebuilds and mixes. And felt I could make a positive difference. I’m not saying the artists you mention view or viewed things this way with regard to immersive audio – but in general I can understand why any artist would need convincing of the format and opportunity. As time goes on and they hopefully hear work that’s also truly reflective of their passion, focus and energy, I’m hopeful I’ll be able to sit with new potential customers and convince them to let me have a go!

CDLP – One of criticisms of the two Fleetwood Mac Dolby Atmos mixes released on Blu-ray is the lack of bass in the mix. It’s a very clean and crisp Dolby Atmos mix but compared the the original 5.1 mixes, there isn’t any low end. I’ve read that a lot of Dolby Atmos mixes are created without the need of a sub-woofer because the low-end is coming through the speakers, is that true for your work?

DK: So here I think we’re getting into an issue with individual listener playback formats, audio connections and misconceptions about Atmos itself. I’m not saying that people aren’t hearing thin mixes, or feel on some mixes that 5.1 might sound better than Atmos. But it’s not the Atmos format, as such, causing the issue. Mixing in Atmos, especially with classic music which primarily lived on vinyl or early CD masters, I always allow the additional low end present on the multitrack through, so there’s usually more bass than there ever was on the original stereo, and yes – anything that flies below a certain frequency threshold will absolutely go direct to a subwoofer. That goes for whether the mix is played in Atmos, stereo or 5.1. Subwoofers are there to handle the ultra low frequencies and take the load off mid range drivers and this follows whether the mix is in Atmos or not. Assuming whoever was mixing in Atmos printed enough bass, the problem seems to arise when different listeners have different bass management systems, AV playback decoding and sometimes even direct audio connections from a blu ray player for instance. There’s also a channel in Atmos mixing called the LFE and often this gets mistaken for a subwoofer channel, when it isn’t – Dolby describe that channel as an effect channel – rarely used and mostly there for extra FX impact in film mixing. But perhaps some Atmos engineers have huge sounding bass rigs and misjudge the amount of bass needed in a mix, and mix thin – and the original 5.1 mixer simply did a better job – but also likely I think is somewhere along the line from individual blu ray disc to speaker there’ll be a compatibility / decoding issue, as a well made Atmos mix ought to sound great regardless. 

Grum: Tubular Bells is one of my favorite surround mixes, I was astonished to discover it was your first Atmos mix. I’d always thought the album was ok, but the surround mix gave me goosebumps in places and I return to it regularly. Now you have more experience, is there anything you’d do differently? 

A less nerdy question. Have you ever just gone crazy with a mix? vocals swirling, every guitar note coming at you from a different place? There seems to be rules about placement that everyone follows and while I want an album to sound similar to the original, sometimes I feel I’d also like to hear a ‘mad’ mix of a few tracks I love, Something throwing familarity out the window and giving me everything from an unexpected direction. Basically a mix that will give me a headache.I imagine even if you wanted to, getting an artist to agree to it would be quite difficult.

DK: Tubular Bells was truly one of the most difficult moments in my career. I was under extreme time pressure, as the label decided to go the Atmos ‘full rebuild from the multi-track’ route at the last possible moment, and it was right up there on the complexity scale too. Plus I knew that an album as beloved and iconic as this would generate a lot of interest and scrutiny. Although I don’t think I’d re-do all my mix choices now, I think with three years of added experience I’d be calmer tackling it, more able with some of the difficult technical elements and curve-balls, and I also have some better tools (my headphones now are a pair of very fancy Audezes which have dramatically improved workflow and my ability to peer into a stereo mix when trying to figure out what effects and balances to use). 

The going crazy thing… Yes – few artists would wish for their carefully honed, beautifully judged stereo classic to be followed by the headache version in immersive – however fun that might be for a one-off demo. I’m really careful to try and make mixes which will allow for long term multiple listens and don’t tend to go in for the Atmos equivalent of early stereo mixes where sounds flew from ear to ear but often became irritating…HOWEVER – pretty much every song I mix has automated elements floating about, in motion, catching the ear. With Heaven 17 it is often vocal echoes jumping about the room, or synth effects split up into its various dry / reverb / echo elements and all of them moving around in different places etc.  The giant stone sound in ‘Crushed’ always sweeps from front speakers to back. I love all that stuff, where it feels appropriate. A hi-hat floating in a circle around the room for the duration of a song – less interested in that. Something that feels like a gimmick no – an element that demands some special attention yes. 

HenkG: Currently, there are Dolby Atmos mixes on the market that will never reach the standard demo quality. Have you noticed this? Is it because the label or band doesn’t have a decent sufficient budget for it? I was recently listening to Big Big Train’s new album, as an example. Both multi-channel tracks largely lack the centre speaker, and the height speakers are only occasionally present. The aforementioned Cure and the latest Human League albums also don’t excel in Atmos. The albums you’ve done so far have established a certain benchmark and are certainly an asset to anyone’s music library, with Claudia Brücken (Night Mirror) being my musical highlight so far. 

DK: I feel your pain. Much as I want to be supportive of anyone making a living in music, it just kills me when I hear some of my favourite records mixed badly, thinly, carelessly or timidly in Atmos. It’s why I decided to get involved in the process in the first place. Why fucking bother if you’re not going to make proper use of the format?! There’s two elements in play here of course. First is how insanely difficult it is to rebuild a mix accurately from the original multitracks. I cannot stress enough how hard and time-consuming it is to do this part of the process properly. Some mixers might approach things from a ‘this is my interpretation of the song and myself and the artist aren’t trying to recreate the original mix’ – but that is not how I’ve been approaching things at all. I spend weeks working on this until I cannot realistically get any closer to the ‘enhanced’ original I’m striving for – the enhancement usually being a more dynamic and deeper sounding version.

And then the part that pisses everyone off when it’s not done right – making use of all the space and speakers in a way that sounds thrilling, fun, moving, emotionally appropriate and exciting. It’s hard to get right – and I get that some mixers just want the project done with and out the door as quickly as possible because there’s really not much money in this rebuild / mix game in comparison to mixing new music from existing pre-made stems, where the hourly rate is an order of magnitude better! But the choice to not use the speakers in a way that sounds more than ‘stereo plus’ isn’t time, money or effort related. It’s a creative choice that I don’t get or agree with. 

KJDD: David, Thomas Dolby once said many years ago that he had no interest in surround sound releases of his albums. His reasoning was at the time, the consumer set up is likely different from person to person, thus the (sound imaging) reproduction could suffer and not resemble what the mixer hears in the studio. Assuming his opinion hasn’t changed in the intervening years, do you agree or disagree with him? And if you agree with him, how to do reconcile this “problem” when you do mixes? If you disagree, how would you convince him that going multichannel, whether 5.1 or atmos, is worth those shortcomings?

DK: Thomas – please come visit my studio for a cuppa! I totally disagree with the reason for not bothering. But of course I agree that everyone’s system sounds different and the range of differences in immersive will probably be greater than with stereo. But we’re in the business of JOY here. And lord knows there ain’t much of that around these days – we need all we can get! I see or hear on a daily basis what a well made mix does to a listener.  It’s an astonishing experience to visit a favorite song in this way. Some fans will have listened to an album every week for 50 years or more, and they or the artist can be floored and brought to tears by a piece of work made with care and love. It needs both of these to get over the threshold though. And also a really good Dolby-approved mixing set up, where choices are well made and as compatible with other systems as they possibly could be. I know for sure when someone posts a video saying my mixes are awful (personal favourite “Kosten has the reverse Midas touch and everything he mixes sounds like shit, and also my teenage daughter hated it”) that at least it wasn’t because of a speaker compatibility error at my end haha. 

Karin: Do you feel some kind of pressure working on historical albums such as Tubular Bells?

DK: Absolutely yes. On all fronts. There’s the making of it – time pressure, tech difficulty etc. But you’re talking about the weight of responsibility I think, and I always feel that very intensely across all my work both as a producer and mixer. And for an iconic, culturally important piece of work, that is of course a weight one cannot and should not ignore. There’s also the proximity to fans / listeners, especially in this age of access and socials, where people are more than comfortable being rude or abusive if they don’t like what you’ve done. Passions clearly run high. I’ve chosen to kind of engage with all of this, in the hope that I learn more about what audiences want, and make better work, and accept the inevitable downside when it arrives (hopefully not literally) at my door. I hope as time goes on and people start to hear a body of work in this format, that they’ll know I really put the effort in and do every job to the best of my ability. I guess if that’s not enough (see Reverse Midas guy above) then so be it! 

MNPT: How many multitrack channels you have to work with for both albums? Could you work from the original multi-track ? Did you have, like the Propaganda Atmos mix, more loose tracks available than the original multitrack ? And if the answer is positive, did you used them ? What strikes me on more Atmos mixes is that drums are so narrow and thin mixed and moved more to the background (read: mainly audible in the stereo image) while it’s a very important and wide instrument, is that a conscious choice. Same for the bass (player)

DK: Yes, after a bit of research we managed to find pretty much all the original multitracks to mix from. The only major anomaly was with ‘Crushed…’ on TLG where I only had a mid-work-in-progress multitrack, but I was able to fashion a mix from that which competes well with all the others. There’s usually a sound or part that’s missing (sometimes things get added late in the day on a studio slave reel and this doesn’t get archived at the time) – but again here we were mostly OK. I’d hate for instruments as important as drums and bass to sound like they’re relegated to the background in my mixes! I hope they’re loud and clear. But you’re right – they usually sit at the front and I’m guessing this tends to be historically because left, centre and right speakers have more weight and power than ‘satellite’ or smaller rear / height speakers. Things are changing though, and I’m sure this will begin to be reflected in mixes. In fact in TLG and P&P I’ve put drums and percussion all around the listener – nothing quite like a giant Linn Drum tom fill flying across you!

Yves: As a quadraphonic amateur, I often wonder why 5.1 mixes don’t have more discrete channel separation (like the quad mixes of Deep Purple’s Machine Head, ZZ Top’s Tres Hombres or any Doobie Brothers’ Quadio). Is it a matter of taste (of artists who allow surround mixes / of the mixer / of the majority of the listenners)? Even in stereo, I know that some people don’t like the mixes of songs like Buffalo Springfield’s For What It’s Worth. I do. Listening to this kind of song with headphones is a very pleasant and amazing experience.

DK: For what it’s worth – I absolutely love extreme use of speakers and panning. You’ll find that all over my productions. It’s not usually viable though to ask the producer or artist of a classic album to disregard all their balancing work and allow the Atmos mixer to go nuts with it. Maybe that’ll happen at some point though – I’m still new to this!

Quante: 1. What’s the point in artists allowing non-surround (sub-par) mixes out of their albums coming out? 2. Why are there amateurs (or is it inexperience?), who mix the likes of The Cure’s recent album, putting out woeful surround sound mixes? It devalues the concept of spending good money on something that doesn’t deliver an enjoyable spatial experience. 3. Should there be a training concept for spatial mixing? Not everyone is a natural. 4. Alan Parsons described fold downs to 5.1 from Atmos as cheating – is it? 5. Do mixes sound good in the studio and then this doesn’t translate through to the multitude of different equipment people are listening to their blu rays through? Do you listen to a mix through different domestic sound set ups, including through sound bars when reviewing a mix? Does a soundbar require a different mix to a 11.4 separate av / hi-fi set up? I’m thinking of how records were reviewed for how they played through a transistor radio as well as a hi-fi. 6. What do you think about the Steven Wilson digital download website? If surround sound went purely digital and dropped blu ray physical releases I’d never hear another new release.

DK: What is the point indeed?! I don’t have an answer to this. But if you’ve read some of my other answers you’ll know how frustrating I’ve found it when this kind of work is done without care, or the mixer and/or artist actually doesn’t even really believe in the format. I’ve been told certain mixers active in this area think it’s gimmicky, and that’s why they stick to a slightly enhanced version of stereo, with a couple of reverbs at the back and just enough audio outside the front speakers to avoid being hit by the trades description act! Training is a great idea – not sure how many of these sub-par mixers you speak of would think they need any though! 5.1 fold downs are a key part of the Dolby format’s flexibility, and I’ve had nothing but praise for both what I’ve heard and also what listeners have said about playing my mixes in 5.1 from the Atmos files too. I know some mixers don’t like using it and that’s cool too.

With the jobs I do I simply don’t get the time or budget to even consider working on a whole separate mix format. I think Steven’s and IAA’s download sites are great things to have, and if it means more people are able to listen to immersive music and make it more popular then of course that’s a fantastic thing. It’s still a niche market compared to stereo, so any expansion of listenership is brilliant. I’m with you of course too though – I grew up with vinyl and will never not want to have music on a physical artefact. 

Alvin: Why the conservative restraint when you can really let things take off in a format that just screams to take those speakers to the max? It’s like, in really a lot of cases the format just isn’t used to do what it was designed for in the first place: immersive audio. A shame really. 

DK: Yep – there seems to be a theme developing in these questions… And I’m totally in agreement. There’s two key things here of course. The time, effort and skill needed to rebuild and recreate a song from a raw, original multitrack containing absolutely no level, processing or panning information. And once you move into the immersive part of the process, the incredibly annoying and frankly bizarre reluctance of so many mixers to use the available speakers and space. Oh another thing in play here too: it’s also really, really hard to make a convincing Atmos mix when presented with perhaps 40+ separate instrument stems. Or in the case of Tubular Bells or Propaganda – getting close to 100 of them. It’s very easy for a song to sound like a collection of disparate parts, with none of the cohesion or emotion present in the original stereo. Is it possible playing safe with the immersive mix is a way for some mixers to avoid this problem entirely?! I dunno. Kills me though when I’ve heard music I would have loved to mix get a boring Atmos or poorly rebuilt mix. 

Mr Kinski: if you could pick any three albums to work your magic on, what albums would you choose? 

DK: Three isn’t fair! And sadly, Dare and The Lamb Lies Down On Broadway were two of my original top 3, when I started work in this format! If we’re talking multi-track rebuilds from scratch i.e. albums predating say 2000-ish digital workstations and stems, then Out Of The Blue by ELO; Kid A by Radiohead; Oxygene by Jean Michel Jarre; Metamatic by John Foxx. The Man Machine or anything by Kraftwerk. Anything by Talk Talk and The Cure. Reproduction and Travelogue by The Human League. The first two Blue Nile albums. Anything from Gabriel-era Genesis and beyond, or his early solo work especially PG3 or 4. The Dreaming or Hounds Of Love or anything by Kate Bush. Sons and Fascination by Simple Minds. Siouxsie, ABBA, Japan, Depeche Mode etc etc 

Doadie: What does David think about the value of music videos showing artists in their prime at the time the music was recorded? 

DK: I’m 100 percent on board, though for remastered video versions if doable, and making use of all the available space on a blu-ray. What’s so funny looking at videos made by some of the artists I’ve mixed in Atmos is how awful their promo videos were in comparison to the elite music they made – and how spectacularly they’ve dated too! But it’s wonderful seeing them, especially when upscaled or transferred at 4K. 

Althan: Hi David, I really would like to know how long it takes to mix an album in Atmos. Also I wonder how an artist like you gets paid for his/her work, since a lot of musicians get almost nothing for streaming. Would you or did you consider mixing totally different styles of music (classical, jazz) in Atmos? 

DK: Well it really is all about how complex and layered the music is, and how much what’s on the original multitrack deviates from the sound of the final mixes I’m trying to recreate. In some cases the workload split between rebuilding an original mix and the actual Atmos immersive mixing part can be as radical as 95 percent v 5 percent. Propaganda for instance.  Fine Young Cannibals was more like 60/40 and took me two solid weeks in total – by far the quickest album to make so far. I almost didn’t leave my studio for over a month working on Tubular Bells. The hourly rate must have been less than minimum wage! (you get paid a fee for mixing an Atmos album. Discussions are starting to happen about physical sales royalties.)  At this point, I’m definitely not working on these mixes to get rich though – but of course there needs to be some kind of balance and nobody wants to go bankrupt because they worked a month trying to recreate a cowbell reverb on A Secret Wish. I’d love to work on any great music, whatever the genre. Hopefully as time passes I’ll get the opportunity to play with all kinds of artists and get lots of surprises with who picks up the phone. 

Namkyr: Dear David, when mixing, how much do you consider the fact that each speaker layout can produce a different sound image, e.g. the reproduction of high channels via ceiling speakers compared to up-firing speakers I use?

DK: It’s a problem no different really to “how do you deal with people playing stereo mixes on phones, in cars, tinny earbuds and audiophile systems?”. For Atmos you do need a system set up that is true, not flattering, that’s neutral and officially installed by Dolby themselves. And then the only answer is to be a good mixer that delivers a product well balanced enough to withstand most of the anomalies and differences that’ll be thrown at it. It’s no coincidence that a brilliantly mixed song tends to still sound great on both a shitty phone speaker and an amazing hifi. I guess an immersive system probably triples the chance of there being issues though!

Bertielego: 1. Back in the days of DVD-Audio and SACD, we saw the emergence of 5.1 versions where surround remixers take liberties, slightly modifying or even extending certain tracks. Think, for example, of some tracks on the 2005 reissue of Simple Minds’ New Gold Dream (81-82-83-84), or on the 2005 reissue of Mike Oldfield’s Crises. Would you consider doing the same? Or do you see it as disrespectful to the original? 2/ Is it a deliberate choice to remain strictly faithful to the original album track listing? I feel that some of the reissues you’ve worked on could have benefited from remixing a few key non-album tracks in Atmos/5.1/stereo, as Steven Wilson did on the recent FGTH reissue. I imagine everyone would have been thrilled to have ‘Frozen Faces’, ‘Jewel’, ‘Femme Fatale’ and even ‘Disziplin/Thought’ as bonus tracks on A Secret Wish; or ‘I’m Your Money’, ‘Are Everything’ and ‘BEF Indent’ as bonus tracks on Penthouse And Pavement.

DK: Yes I’d absolutely love to have the time, the budget and the label and artist permission to do more wild bonus / extended mixes and non album tracks. If you knew what goes into the pre-mix licensing part of putting these blu ray collections together I think you’d understand better why it’s such a frustrating rarity though. Also if you could see some of the music that’s on some of the multitracks which show up – never previously heard songs, demos, b-sides etc. Wow. But I can’t just go do a mix of them without a lot of other people’s OK…

Glen: If mixing, say, The Beatles, would the remixer preserve the original left/right balancing of the instruments, or would he/she take the attitude of “George Martin panned the guitar to the left, but I’m going to put it anywhere I want”? 

DK: So far, I’ve been borderline obsessive about matching the panning when working on my rebuilds! Even going as far as following mad original mix panning around when it happens – Tubular Bells had a ton of this and I carefully matched all of it! You’ve hit the reason why though. My goal isn’t to argue with George Martin. So far it’s been to honour absolutely the dedication and genius of whoever made the music I’m recreating, and then to bring that work out into the immersive space. I like to be able to hear the mixes as if they’ve just been made, with every original sound and reverb tail available for me to use in the Atmos mix. Frankly though – this part of the process is incredibly hard – near impossible to get perfect. Which for a perfectionist like me is really painful. I do my best though to make the artist really happy and to bring a fan true joy at hearing all the familiar balances and sounds in a brand new way. Seeing the artist becoming emotional at a playback has become both a wonderful regular event and also the benchmark for success too. If they don’t cry I’ve not done a good enough job!

Chris Busby. Do you think a particular style of music or combination of instruments works best for Atmos? Also given the comment about Temptation, does it work best with a huge number of multi track layers or less?

DK: Good question. Synth based music is a gift to the format. Maybe because there’s no instinctive feeling about how a performance of it ought to ‘look’. A live band recording spread around the listener provokes a load of questions, and needs very careful mixing to retain its power. Can still sound fantastic though. So I don’t think there’s a best. But there’s probably an ‘easiest’. A lot of artists and mixers think you need loads of layered audio tracks for a song to sound good in immersive. Not so. Less can be much more. And of course a solo instrument or voice with reverbs or echo can still fill a space in a spectacular way on its own. A chorister in a cathedral doesn’t need layers of extra sounds and parts to move you. And an 8-track recording can sound every bit as good as a 72 – especially as there’s much more space to hear all the little details in a sound which bring so much added pleasure. You have to work hard for a giant layered mix to sound good in both stereo and Atmos. 

Kauwgompie: Are you working in any other 80’s Atmos projects right now or in the near future?

DK: I wonder if at some point it’ll make sense for the work I do to be documented as it happens, from first receipt of the multi-tracks onwards. As a fan I’d love it if that happened. So far though – I’m sworn to secrecy until an announcement is made! (Thanks for the great support btw, it doesn’t go unnoticed!)

Jurgen: When you start mixing in surround, do you go for a standard placing of the instruments or do you try different placings? For instance … the drums in most of all the surround mixes are always in the front. Is this because it sounds the best or is this what most people want/like? But is this the best place for every song? A rhythm guitar can be placed left or right, front or back … do you know in your head where to put it or is it by experimenting that you know the right place?

DK: I sort of dealt with some of this thought process in another answer – but I’d add that a kind of instinct develops about what might both feel good and still retain mix coherence. And of course it’s easy to try, reject and improve as you go. I mixed all of Tubular Bells – spent ages doing it. And then scrapped the entire thing – as although it was fun and had a lot of sounds in interesting locations, it didn’t feel right or present itself to my ears as a ‘piece’. I didn’t believe it. Learned a lot from that moment. It’s coloured every single mix decision I make in Atmos now. Find the joy, tell the story, make it emotionally authentic, don’t go for gimmickry, (make the artist cry). 

WhoRU: Seeing that you’ve invested in your Dolby Atmos studio, is this a sign that there’s a growing interest in record companies looking to release more Dolby Atmos mixes in the coming years. Can we expect a few David Kosten Dolby Atmos releases this year? Even though SDE gives everyone a two week window to purchase your mixes, there still seems to be folks who have missed out whether it’s financially at the time or were late to the party to Dolby Atmos. Do you see your mixes ending up on ‘Headphone Dust’ as an alternative way of purchasing your material.

DK: It’s almost four years now since I had my system installed, and nearly three since Tubular Bells was released as my first commercial Atmos mix. I’m so happy I’ve managed to make some listeners enjoy music they already adored in a new way, and hopefully I haven’t made too many people furious at my mix choices along the way. Who knows with Headphone Dust? It’s a great idea of course for music that’s simply not available in a physical format.

What’s your goal when creating an Atmos Mix for a band such as Heaven 17?

DK: It’s a really simple goal. Bring some joy to the listener. How I try to get there is by honouring the original recording, and being as skilled as I can manage at interpreting their intentions in this 3D space. Sounds easy to do but it really is very difficult to get right.

How have you used the centre channel on these albums?

DK: I think all the H17 lead vocals use the centre speaker, sometimes with mono effects added. I experimented a lot with separating ‘wide’ processing – reverbs, delays, harmonizer effects etc – from dry vocals, and there’s a lot of fun with placing effects in the space, away from the front. I also put some synth sounds and fx through the centre too. Works really well as a way to go musically from black and white to colour – mono at the front expanding into a fully immersive track.

Why do you think some Atmos Mixers virtually ignore the centre channel?

DK: When I started my Atmos journey three or four years ago, I was quite cautious about this too – partly as I wanted to try and ensure lead vocal levels were heard exactly as I wanted them, and avoid any issues with badly set up systems. But I quickly learned that this approach infuriates immersive music listeners, who love all their speakers to be used (seems fair to me!) and also began to relax about it too once the responses to my work were so positive. I still take care with vocals through the centre speaker as format compatibility issues can arise. But I can understand why some mixers don’t bother. It adds to the complexity and potential incompatibility of a mix. And undeniably sounds ‘different’. 

How ‘respectful’ to the original mixing decisions do you try to be, within the confines of the 3D space of Atmos, when working on records like P&P and TLG?

DK: I am respectful to the detriment of my sanity when it comes to the recreation of an original mix! The only thing that makes it almost OK in terms of my workload is when I get positive feedback, or a genius like Martyn Ware who has almost every single artistic decision he ever made still there locked in his memory banks looks at me and I know that he knows what I have gone through to make his music sound ‘right’ haha! 

Would you ‘fix’ an obvious issue (glitch etc.) if you came across one, even if it was on the original?

DK: My not particularly deeply thought out rule is ‘if they would have wanted to get rid of something but couldn’t at the time then it’s probably ok for me to think about ditching it’. Like tape hiss for instance. With these rebuilds there’s an opportunity to lose it almost completely, and even though that’s sometimes quite a shock, if you’re used to hearing tape noise on a classic song, I love hearing the sounds appear like magic out of thin (hiss-free) air. I’ve fixed clicks and pops that clearly weren’t meant to be on an original recording. I wouldn’t correct pitchy vocals though. 

How collaborative is Atmos Mixing? Would you sit with another person and work on a project together, or is it a solo endeavour?

DK: Something along these lines might be happening soon…watch this space! So far though it’s been by far the most solitary endeavour of my career. Makes the artist and fan playbacks all the more important.

As a producer and an artist in your own right do you think that gives you an advantage over Atmos Mixers who may not have produced for others or created their own music?

DK: This is truly the key to all my work! Rebuilding and mixing in Atmos wouldn’t be happening for me without 20+ years of producing other artists and working on Faultline. I should get a clicker to count the number of problems needing to be solved when working on a Luxury Gap type album rebuild – I reckon it’s literally in the tens of thousands. And 99 percent of the answers will have come from a prior experience as a producer or mixer, or from working with an amazing engineer. Impossible to do this work at the level I want without every minute of my past work being called upon. 

Can you explain how what you hear on the raw multi-track tapes of Penthouse and Pavement and The Luxury Gap differs from what we know as the finished albums.

DK: One thing I think a lot of listeners won’t realise is how careful the original engineers were to print ‘conservatively’ to analogue tape. Both in terms of volume levels and also EQ. What this means in practise is that they often didn’t print sounds with much top end as it can distort tape quite easily. And the raw sounds out of drum machines were often recorded very thin and quite shockingly unlike what the final versions sound like. A TON of mix processing is needed to get some of those kind of sounds close to the final versions. But other sounds could already be very close. Glenn’s vocals for instance. Barely needed any processing at all. And the orchestral stuff – tons of balancing and panning and horrific amounts of time spent working out the complexity of the mixing of all the orchestra tracking on The Best Kept Secret. But not much actual work was needed on the sounds themselves. They were beautifully recorded in the huge room at Air and just worked straight from tape. 

What has to happen with those raw transfers before the immersive mixing begins?

DK: Depending on the song, the bare minimum is using varispeed and often lots of editing to match the song length with the definitive stereo master. ‘Crushed By The Wheels…’ for instance needed a ton of edits to match the final song structure. And ‘Temptation’ was spread across three different 24-track tapes, all of which had been transferred at slightly different speeds – synchronizing these and editing them all to match the final stereo master’s song structure took me about a day! Then I like to make sure that every sound has its own audio track in Pro Tools (my digital workstation). That’s because inevitably, each sound will needs its own effects, tone controls, and level control. Sometimes a single tape track could contain 5 or 6 different sounds – I put each one of these on its own audio track. So an initial 24 tracks could suddenly expand to 100 or more audio channels. From that point the rebuild begins – a forensic recreation of all the decisions made in an original mix. That’s an article on its own. Definitely the most time consuming and difficult part of the process to get right. 

What are the challenges in this rebuild process? 

DK: You need to be able to hear into a mix and work out exactly what they did at the time, within probably 0.1db accuracy – multiplied across perhaps a thousand sounds on an album! What reverb is the snare using, Is there compression on the vocal. What’s that EQ doing on the guitar and why does it change from verse to chorus? How is that mono synth sound so huge and wide in stereo on the final mix. All these questions need answering before a track starts to even vaguely resemble the original version. I return to a song maybe 10 or more times during the process. I get to a point that feels like progress, and move on and try to get another of the songs to the same level. This definitely isn’t “push the faders up, get pretty close and have done with it”. 

How important is it to know the album you are working on? If you love it are you likely to go the extra mile to make it sound incredible?

DK: I’m lucky in that so many of the records I’ve rebuilt and mixed have been favourites, but I think my main requirement is that they’re great, for me to find the energy and motivation. It’s SO much work to do this at a level I can feel proud of and want to present – life’s much too short to give that kind of attention to something that’s not brilliant! It really did help though that I already knew every millisecond of A Secret Wish. That one was a technical nightmare, and knowing the basics of what needed to go where, without constantly having to A/B against the original tracks probably saved me weeks of work. 

How much do you keep an eye on your peers – someone like Steven Wilson – and check out their work?

DK: Well Steven is certainly a key reason I’m at this point. He and I go back a long long way, and of course we produced and mixed his The Future Bites album together. So we’re in touch often – asking questions of eachother about mix or rebuild problems we’re encountering etc. He’s been an absolutely incredible source of support and help. Talk about lucking out with him for a mentor. He sends me lots of his Atmos work and while of course sometimes I feel like I’d have done something differently, I listen very carefully and make a mental note of anything surprising and fresh which I’m jealous of! 

Is it important to you that your mixes get a physical release on blu-ray?

DK: Yes, it’s truly the best thing isn’t it. Such an incredible format for this type of work too. So much room for so many mixes and options. I also love the notes that the engineers or artists write for the booklets. I’m basically a fan, and I love physical product. 

Thank you to the SDE readers who contributed questions and many thanks to David Kosten for taking the time to answer these so diligently!

The two Heaven 17 SDE-exclusive blu-rays which feature David’s Atmos, 5.1, stereo and instrumental mixes are available until Tuesday 24 February 2026 and are released on 24 April 2026, via SDE Records. You can pre-order by heading to the SDE shop or using the buttons below.

Sign Up For the SDE newsletter

Submit