|
|
Steve Marshall
|
|
After a fruitful career as a composer for films and television, multi-instrumentalist Steve Marshall has decided to record music for himself. His approach is highly individual technically and artistically, taking full advantage of binaural recording and surround mixing, and combining sound recordings from places as far apart as East London and South India to create not only his music, but also the space around it. We talked about his sound sculptures, the transaural technique, and his favourite digital audio workstation…
Did you start as a musician first, or as an engineer?
 I’ve been a professional musician for thirty years. I taught myself to play guitar, and adapted what I’d been taught on the piano to playing keyboards in bands. In 1984 I stumbled into advertising music, and bought a four-track studio with the proceeds of my first jingle. I’d always been more interested in the studio than the stage, and just learned as I went along.
The studio keeps growing – and I’m still learning…
What instruments do you play, and what kind of musical training do you have?
I play a bit of everything. I spent 20 years making sample-based music for films, where I had to play it all myself. I’m happiest playing keyboards, but still play some guitar. I’ve got a lovely Hammond A100, which is my favourite instrument, but I’m fond of weirder ones like the Theremin and the Clavioline.
What kind of musical work have you done?
About 200 TV films, mostly for the BBC and Channel Four. Every genre, though I did specialise in wildlife and travel documentaries at one time. My favourite is still “Galahad of Everest”, where Brian Blessed climbed the mountain for the first time. I’ve done lots of kids TV, which is great fun to do. A lot of animation – often tracklaying the effects as well as composing the music. It’s very satisfying doing work that makes people laugh! I think “El Nombre” is on air at the moment – a model animation series set in a Mexican village full of gerbils, who whizz around on motorbikes… For a time I worked as a composer in the BBC Radiophonic Workshop.
What are you current projects?
In early 1993 I met singer Laye Sow in Africa, and took the decision to change career and go full-time into record production. I brought him and an eight-piece band over to the UK from Senegal, and recorded an album which is now on sale in Africa.
I then produced “Djamano”, Laye Sow’s solo acoustic album. Laye and myself played guitars, bass & percussion, and I brought in Richie Caswell to play slide guitar. David Tietz recorded it beautifully, so I was able to just play and produce. The fusion of African music and blues guitar has proved popular, and the album has been released by Orange World, based in Poland. Quite international!
Right now I’m in the middle of making a new album with Richie Caswell as “The Howard Hughes Suite”. It’s Richie’s project. Richie has started all the tracks in his studio, and we’ve transferred them into my Soundscape to develop them. I’m really enjoying experimenting with the engineering. This album is totally different – still using guitars, but with some very unusual found sounds. Lots of synths, loops and samples, but most are re-recorded with amps and mics. We’re using a live drummer too. The working title is “Backwoods” – as a lot of it is actually played backwards… I’ve been using my Hammond and Theremin too.
Also I’ve already started collecting new sounds for the next volume of "Bilocation".
 How did you get interested in surround sound and do you think it is valid for music as well as for sound effects?
I’ve followed surround sound from the sidelines ever since the 70’s, when Ambisonics was developed and then scuttled by the government-run NRDC. An old friend of mine, Ken Giles, now makes the Soundfield mic and it continues to get better. Its potential though is still largely untapped. The fact it can record height information is often ignored, and that’s the part that really fascinates me.
Music and surround sound are natural partners. Innovators like Les Paul and Joe Meek raised studio recording up from being a simple record of a performance, and I see surround sound as a natural progression. But do something new with it! What really irritates me is seeing what is available now in surround – hardly any innovation, just endless remixes of classic rock albums. The same thing happened when CD first came out. The record company suits decide that people who have the playback equipment are of a certain age, and that they’ll buy this stuff. Well I’m that age too, and I bloody well won’t! I’ve bought it all twice already. Surround sound can be really exciting, and I want to be thrilled and surprised by it.
How does surround recording relate to the live experience? After all, the public is generally in front of the performers, with the sound coming from the front. There are reflections from the back wall of the auditorium, but these also happen with a pair of stereo speakers.
It’s quite depressing to see the way people are setting up 5.1 at home now – speakers stuffed behind the sofa etc. It’s okay for films, with explosions behind your head, but not for what I’m trying to do. “Bilocation” is only available in 5.1 and I’ve gone to a lot of trouble to persuade listeners to set up their home system to ITU spec. I’ve put full instructions on the website, and even printed a template on the disc for setting speaker angles. There’s a recommendation to place the speakers far enough from the walls to minimise early reflections. If listeners can be persuaded to this, then surround sound can go to a whole new level: the living room walls will disappear! The fact that all the speakers are equidistant from the listening position means we can use very discrete delays to create an illusion of whatever space we want. But there’s not really that much you can do to enhance a live recording of a band – they have to be at the front, and the crowd have to be at the back… I’m more interested in a kind of “artificial reality” and sonically building spaces that don’t really exist.
With the Howard Hughes Suite album I’m recording now, the 5.1 mix is being planned as we go. I work in an old farm building forty feet long – one big stone room, with just a few acoustic screens. We’ve got ambience mics all over the place. I’ve also been experimenting with using five mics to record, placed at the proper angles for 5.1. We’ve just done a track where I used this for recording a Hammond. The mics were placed around the Leslie, and in 5.1 the results are amazing. I imagined this would produce a single sound that whizzed around the room, but although you can hear the rotation, the effect is much more subtle. The room fills with sound, which is constantly changing. Try it! I also run individual tracks off onto CD, and play them back in interesting acoustic spaces. I’ve got a couple of churches I use for this. Then the results are re-recorded binaurally, and put back into the track. This is very easy with Soundscape. I can then use my transaural technique to add surround ambience to something that was recorded in the studio. A lot of "Bilocation" was recorded this way.
Do you have any particular surround mixing tricks to share with other Soundscape users?
 For “Bilocation” I developed my own method for producing a totally 3-D soundfield using only stereo recordings. I record binaurally, using omni mics, and then use a transaural process to create “virtual headphones” using the 5.1 speakers. It’s basically a phase-cancelling technique: each of the stereo channels is mixed with the other, out of phase and delayed. This gives all the binaural spatial effects associated with headphones, but on speakers. In “Bilocation” a helicopter flies very clearly overhead. Quite weird in your living room!
The transaural technique is a very powerful tool, and I think it could be very useful in film post. It seems so far that any stereo recording made with spaced microphones can be decoded into surround – either made with omni or cardioid mics. I’ve not tried it yet with orchestral recordings (they often put the mics twenty feet apart) but I’ve tried everything from half an inch to three feet apart – and I’ve always ended up with a solid soundfield. It seems that recording with spaced mics makes a kind of “audio hologram” that can be re-created in surround. It’s easy and quick to do.
Just put your stereo original onto a pair of tracks, then copy it downwards so the original and the copy are in sync. Swap the left and right channels on the copy, and reverse their phase. The copy also needs to be delayed by however long it took for the sound to pass from one mic to the next. If you know how far apart the mics were, you can work this out at one millisecond per foot. If not, then adjust the delay until it sounds right. I use the sample delay module on the Soundscape mixer. It’s handy because it has a phase-reversal switch too. For recordings made with a dummy head, the delay should be about twenty-two samples (six inches apart at 44.1 kHz). The very highest frequencies can confuse things, so I just roll off the top frequencies in the out-of-phase copy. Then play it back through your surround speakers, feeding the front and rear speakers of each side with equal amounts of the same signal. This produces a “sweet spot” that can be a few metres across. Do this with something like a binaural recording of the Dawn Chorus, and you’ll have birds singing all around you, and above your head!
It’s important to have the listener facing forward though. I don’t mind this, as my own vision of surround sound is essentially cinematic – films without a picture for seated listeners in the optimum listening position. I’m trying to make “sound sculpture” to be enjoyed at home. Mixing for clubs is totally different, as the listeners are moving about. So really, for clubs you’re virtually doing a quad mix, with no real soundfield.
What does “Bilocation” mean, and how did the concept come about?
“Bilocation” is to be in two places at the same time, and I’ve made a montage of recordings made in widely different places: on the back of a galloping horse, inside Indian temples, on fairground rides etc. There’s a lot of musical elements too, but they’re also recorded in different acoustics. For some parts, I whirled the playback speakers around. Because I’ve then nested several different acoustic spaces, the effect is very strange and absorbing. One of the places I recorded was a massive dome in South India, with a long repeat echo. There are children shouting into the echo, and swallows calling as they fly through the building. This gives a huge impression of space, but superimposed on this is a recording made in a small stone church. The human brain use sound for mapping our surroundings far more than we give it credit for, and so this kind of conflicting information has a very odd, dreamlike effect.
I lived in East London for 10 years, and recorded everything around me – the traffic, helicopters etc. Then I spent a month travelling all the way around India on trains, researching for a series I was working on. I then moved to Wiltshire and continued recording. I didn’t really know why – it became an obsession! I made a few attempts at doing something with all this material in stereo, but could never make it work. Then when I discovered surround sound and the transaural technique everything clicked. In the end I put the album together in only three weeks.
For Bilocation, why did you choose DTS over Dolby surround?
I did try Dolby, using an obsolete piece of software, and it sounded awful. I’ve been told it is possible to get good results with some tweaking, but I never pursued it. I did phone Dolby, who are based a few miles away from where I live, but they weren’t at all helpful. DTS on the other hand were great. They gave me help and support, and were genuinely interested in what I was trying to do.
 How was the DTS encoding made?
The software is incredible – Minnetonka make it. I don’t actually have it myself, and a friend coded it for me. I played around mixing, with the outputs of Soundscape driving a surround amp and speakers, until I was happy. Then I mixed the six channels down and made six WAV’s of them. This filled up two CDR’s for a forty minute piece. The WAV’s are fed into the software, and at the press of a button out comes the DTS encoded version, which fits easily onto one CD. I think it uses data compression of 10:1 but you’d never know: the coded version was exactly as it sounded when I mixed it.
Talking about equipment now, what experience do you have with Soundscape and what is your current setup?
I’ve used Soundscape for about ten years, and couldn’t imagine working any other way. I started with an SSHDR1, then upgraded to a “plus”. Last year when I decided to commit to production I bought a R.Ed 32. It was ex-demo, and had all the plugins. Then I bought a Mixpander, and it’s transformed things for me – I just can’t run out of DSP! The new VST support is great too. I’ve had some VST instruments for ages but the latency made them unplayable before. I use two SS8I0-3s so I’ve got sixteen analogue ins and outs, and that’s plenty for me. I’ve still got loads of analogue gear, and I’m keeping it. I like valve front-ends, and I still keep my old Studiomaster mixer for its mic amps. A lot of my material goes into Soundscape digitally, and I have a little 1U patchbay with a phono socket for every digital input and output in the studio. I just use short phono cables to patch them together. I don’t even bother with 75 ohm cable – it works fine (maybe because the runs are short).
Do you have a lot of plug-ins?
So many I haven’t tried them all yet! I also got dozens of free VST effects and instruments from the links on the Sydec site.
Have you tried other systems?
I went very carefully into buying a DAW, as they were not common ten years ago. KGM let me spend a day in their demo room comparing Pro Tools and Soundscape, both of which were very new. I played around with some binaural recordings, checking whether they still worked on headphones after going through the two systems. I also put a track in, copied it, and flipped the phase to see if it disappeared. Soundscape did this perfectly, but it was still just audible on Pro Tools. I’d heard Pro Tools was prone to crashing, so I had a go at hitting handfuls of keys – and sure enough it crashed. What finally convinced me to go with Soundscape was that I couldn’t make it crash. I switched off the PC while it was playing, and when it re-booted the track was still playing…I’ve always been impressed with the audio quality of Soundscape. DAWs shouldn’t have a “sound” but they do. Most of the other systems I’ve heard have a “plastic-ky” quality that I find hard to define - but it’s there, and to my ears it makes the sound somehow two-dimensional. The audio quality of Soundscape can’t be faulted. It’s logical and easy to use, and totally reliable.
Do you generally mix your own music yourself?
I prefer not to, for the same reason as cameramen shouldn’t edit their own films. You tend to leave in the bits that were difficult to do – whether they work in context or not. What I prefer to do is build the mix as a I record, get something that works and then get in a fresh pair of ears at the end to do a final mix. It’s so easy to recall a mix, and you can easily keep different versions anyway. But saying that, I did mix “Bilocation” on my own. It was so exciting to do, I didn’t want to let anyone else in!
Do you tend to mix entirely inside Soundscape?
Most of the mixing is done entirely within Soundscape, but I sometimes use the SS8I0s to hook up to my old analogue racks. I’ve got several Joe Meek units, and other boxes with valves in, so I like to use them. I like the idea of having a state-of-the-art digital system hooked up to a 1960’s spring-line reverb…
 How was Soundscape set up for Bilocation?
I made a mixer with a mixture of stereo and mono strips. I’d already edited and EQ’d the raw material and put it onto ten CDs, so I didn’t include any EQ. Each strip had six aux sends, going to the six channels of 5.1. Some of the original stereo tracks were fed into two separate mono strips, so I could place them diagonally across the sound stage. I also had a couple of strips with the Soundscape 5.1 module in them – for placing things anywhere within the soundfield. I had two sets of transaural strips too. At the end of the mixer were six output strips, one for each of the surround channels, and a stereo output strip. The surround channels went straight out to a surround amp and speakers. All the fades were drawn on screen. The only thing I used automation for was changing the delays in the transaural channels, using snapshot. When I was happy with the mix I dropped channel modules into the six surround output strips, and recorded the six channels in one pass.
What are your future projects?
The stereo version of the Howard Hughes Suite will be finished in a month or so, but then we have to do the surround mix…
I’ve got another surround project planned for early next year which I’m very excited about. I’m working with three other musicians on an album of acoustic music improvised and recorded in different locations. The plan is to make my Soundscape mobile by using a laptop, and the four of us will play together in a church, all placed around a Soundfield mic. Then we’ll go to a different church, with a different reverb time, and record more parts, and so on. The idea is to create spatial hocketing rhythms that use the space as an element of the music. To me, that’s a good way to use surround sound and I can’t wait to get started!
Find out more about Bilocation
An in-depth article about the making of Bilocation is featured in the October 2004 issue of Sound On Sound.
|
More inFocus articles
| Back to the homepage
|