'Till We Have Faces', the new album from Steve Hackett, is rock music's first flirtation with the world of Ambisonics — a system of surround sound which attempts to create around the listener a realistic illusion of the soundfield that existed at the original recording session. To discover more about the system and the way it was used on the album, HSR sent Nicholas Webb off to Marcus Music Studios, London where he spoke first to recording engineer Richard Elen — a longtime exponent of Ambisonics — and to Steve Hackett himself. Is it just another fad? Or are we looking at a new sound for the future?

Ambisonics

RICHARD ELEN: Ambisonics is basically a system of encoding 'directionality'. In other words, getting recorded sounds to appear from a desired direction. You're not limited, as you are with stereo, to a sixty degree spread of sound. You can position things all around the room, round the edge of the room or in the centre. To do that you need to replay the encoded recording through a decoder which would usually be put after the preamp and into a pair of power amplifiers and four speakers.

If you don't have a decoder and you play the Ambisonic recording back in stereo, you will get an enhanced stereo effect. The image will be rather wider than the speakers and you'll have better localisation or a great deal more clarity as to where the instruments are actually placed in the mix. Also, if you collapse it to mono you get all the 'punch' as if you'd originally mixed it in mono, so it's deliberately compatible between surround, stereo and mono systems. This is a very good aspect of Ambisonics.

Do you then mix down an album with this in mind?

ELEN: It's a good idea, yes. Your multitrack recording techniques can be pretty standard. You can use the Calrec Soundfield microphones which are designed to work with the system very nicely. Equally you can just take your ordinary 24-track tape or whatever, put your instruments through either an outboard unit which gives you 360° pan control or you can use the mixing console pan pots as part of an Ambisonic system.

You just re-assign your sounds slightly when mixing. Just as you do with stereo, you can localise your individual instruments in the mix more or less anywhere you like - but instead of being limited to 60° you can do it anywhere around and inside a complete circle.

It's hard to visualise 360° of sound - to see myself in the middle of a circle, as opposed to 2 or 4 separate sound sources.

ELEN: Think of yourself as being in a real acoustic environment. Think of the fact that if you are in a concert hall listening to a band you'll be hearing sounds coming off stage, out of the PA, some reflections off the walls, and you're hearing the audience behind you.

We can recreate all of that electronically. When you're listening with a decoder and four speakers, you can arrange the speakers in just about any rectangle you like. Basically, it's designed for domestic replay, it's not a system where you have to be a super audiophile to experience it.

In ordinary stereo recording, or 'pan-potted' mono which is a more accurate description, you have a pan pot on the mixer which controls the amount of signal level of a track going to left or right speakers. Now level is only one of the means we use in hearing for localising where sound is coming from. Another important aspect is phase and what Ambisonics does is to give you control of both level and phase. As a result you can recreate a sound coming from any direction you like in the horizontal plane. In fact, we can recreate the impression of height as well, but it's difficult to get it on records.

Ambisonics is not the same as 'quadrophonic sound' because the old quad system still used level to localise between four speakers. So what tended to happen was that sound was sucked into the speakers. If you panned round the room in a circle it wouldn't go in a circle, you'd get sound dashing from one speaker to another.

So 'quad' was just a pan pot that waggled four levels about instead of two.

ELEN: Exactly. What we've got here are pan pots that waggle level and phase around. The old quad idea was that to get 'surround sound' you had to have four different signals stuffed into four different tracks of a tape recorder and four different speakers when you listened to it. That idea was actually wrong. What you're really trying to do is to re-create the wave fronts that come to you from different directions and that doesn't have much to do with just waggling level about. It's much more to do with other effects...

It also gives the advantage then that you don't have to be listening in exactly one place as you used to with quadrophonic sound.

ELEN: That's correct. You can stand outside the speakers and there's still that image sitting inside the room.

People have used what's called 'phase shift' panning in the studio for years. It's quite a common technique. You know how you do tape flanging, for example, where you take a signal, split it, put one side through a fixed delay and the other side through a variable delay, mix them back together again and as you move the variable delay past the fixed one, you get that flanging effect.

Now if you were to take those two outputs and listen to them in stereo, instead of the flanging effect, you'd hear the sound source pan from left to right and back again. That's just an example, leaving the level alone and changing only the phase. Now if you take level and phase together you've obviously got a much more powerful medium for controlling direction. That technique works in stereo as well as it does in surround, so when you take your Ambisonic album and you listen to it straight, say you're an ordinary guy at home who hasn't heard of Ambisonics or decoders, you'll still find that one track on Steve's album is different and that's the track which isn't in fact encoded. It's a straight stereo mix on 'A Doll That's Made In Japan' and you'll notice the difference straight away.

If you're listening on an ordinary stereo system you will find that there are lots of advantages of being able to use level and phase for your localisation. The big advantage of Ambisonics is that it's not nearly so position dependant when you're listening to it. You don't have to remain glued to the spot as you did with quadrophonic sound.

You've worked closely with Geoff Barton (Ambisonics inventor), but what is your own connection with this new system?

ELEN: As an engineer. Until recently I was also Editor of Studio Sound magazine. The reason I became involved with Ambisonics is purely because I think it's a nice system. I like using it. I'm not interested in promoting the idea so much, I just think it's good. It's a new thing for most people, particularly in the rock field, but it's been utilised in the classical field for some years.

I find it interesting that the classical boys have picked up on Ambisonics and yet the rock scene is just coming round to it. Why?

ELEN: Well the system was originally devised by a group of people from the Mathematical Institute in Oxford and some people at the University of Reading. In fact, quite a disparate group. The system was picked up by the British National Research Development Corporation (NRDC) which is now the British Technology Group, and they have been funding it since the early Seventies.

The system was originally envisaged as a means of capturing, with a Soundfield microphone, a live concert performance. Specifically a live classical performance. The microphone virtually came first as a means of capturing the sound from all directions then came the encoding scheme to get it on to two channels so that you could actually put it on records. Then came the system of decoding it so that you could listen back to it again.

The reason that the classical boys picked up on it, unusually, is that it is a very, very simple technique. For example, if you go into the Albert Hall you will see a Soundfield microphone hanging from the ceiling which the BBC keep there. It's a very easy microphone to use. You just hang it up and, there it is. In fact, the microphone is so clever that you can actually record it on four tracks of tape, move the microphone around and change its polar pattern after you've recorded. You can physically move the microphone backwards and forwards.

This obviously had a great appeal to the purist element in the classical recording fraternity because it's based on a very simple technique developed directly from Alan Blumlein's original stereo recording techniques of the Twenties. Blumlein developed the original concept where you had an omnidirectional microphone and superimposed on it a figure-of-eight polar pattern pointing to the left.

What we do with the Soundfield mic is we have an omnidirectional microphone and a figure-of-eight pointing to the left, which is left minus right signals. So your omni-signal is a sum of all the directions, left minus right in your horizontal, front minus back and also up minus down, because with this particular microphone you can record height information as well.

That composite signal, which is three different signals really, is called 'B' format and is the usual studio format used for Ambisonic recordings as it happens. The mono signal is called 'W', the front to back signal is called 'X', the left to right signal is called 'Y' and the up to down signal is called 'Z'. We don't normally deal with 'Z' signals because a complete three-dimensional surround system is a bit more difficult to set up for replay. That's called 'Periphery' which means sound around the edge. So we take away the 'Z' information and we get left with three channels of sound. Those can then be encoded into two.

The old quad system where you had four channels encoded into two and got four back again didn't actually work. One reason was that you can't get four into two and get four out again. Three you can (see Figure 1).

We, in fact, drive four speakers with the resultant signal because it's a convenient number of speakers to use.

So, when I came on the scene in the mid Seventies I thought 'wouldn't it be nice to do this with rock music'. The question was, 'how do we simulate all this electronically?' So what was produced was a set of electronic units for the studio which are intended to take mono signals in and to encode their direction in the Ambisonic format. And that was what was used for this Steve Hackett album, 'Till We Have Faces'.

Can I ask you a general question about the use of Ambisonics? Is it really going to make much difference or is it just a luxury, a passing fad?

ELEN: Well, firstly, there has been quite a long time lag between the possibilities being realised and the manufacturers picking up on it. This is why we haven't seen it in the Rock and Roll field until very recently.

So, do I think it's a fad that'll pass away? I don't know, it has been bubbling along merrily since the early Seventies. Albums have been produced by record companies like Nimbus, Unicorn and others in the classical market and that's a Hi-fi freak's field obviously, but when it comes to Rock and Roll it's a system that can, in fact, stand on its own without decoding. You are already getting more from an Ambisonic recording if you're just listening in stereo. It's not a very expensive system, compared with other studio equipment, the whole package costs about a third of the price of a Lexicon digital reverb unit. Just operating as a stereo effects unit though, it's quite cost effective.

If the software exists, as it's now starting to, then there's a fair chance that people will be interested in getting the best out of the ambisonically encoded records by getting their own decoder. Because you don't need big speakers, and because the decoders are very cheap, starting at about fifty pounds, then you're not talking about the same sort of investment as you require for a compact disc system. It does, in fact, go nicely with compact disc as a concept.

How many of these Ambisonics systems are in commercial use at the moment?

ELEN: About a dozen are in use by recording engineers. There's a system here (Marcus Music Studios), a system in my studio, and I believe Neil Young has one in the States. Stevie Wonder and some other people are also experimenting with it. It's generally producers and engineers who are being turned on to the technology and I think it will follow the same sort of profile as compact disc.

Is the information easily transferable to cassette?

ELEN: Yes. You can use it on everything. It can also be used for broadcasting. Radio signals can even be encoded. In fact, the BBC are doing a programme on this album (Steve Hackett's 'Till We Have Faces') which will go out entirely encoded.

I don't think technology should be sold for technology's sake. Real people don't buy technology, they buy music. If the Ambisonics system can offer existing listeners extra benefits in terms of sound quality, then that's a good thing.

Steve Hacket

Why did you employ the Ambisonics system on your new album 'Till We Have Faces'?

The idea was suggested to me relatively late in the album's conception. The Ambisonic system happens to be a facility of Marcus Music Studios where we recorded part of the album and which is also my favourite studio in the world. The system was explained to me and I asked if it would affect normal stereo sound and they assured me it would give a thirty percent wider spread on the stereo image.

I was quite sceptical about using it right up to the moment we started mixing. Geoff Barton, the inventor, came down to the studio, set the thing up and we found it worked very effectively. We found that by using the effects in a special way we gave an impression of being in a hall. We managed to create the effect of sound bouncing off the back wall and things like that.

Did you have to mix down in a different way, or do several mixes, then stand back and see what it was doing to your music?

We just mixed ambisonically and found that every time we checked back in stereo everything worked perfectly. I think the system has been well worked out because I had doubts about the record cutting stage but there was never any problem there. It's got to be a contender for a way of listening in the future.

How did you mix down?

We used four speakers and kept checking back on two. Although we had Auratones (standard studio reference monitors) in the studio we didn't actually use them. I've changed my whole attitude about mixing. The old idea was always to aim your mix to sound good on a transistor radio, but I no longer want to make music for garage mechanics to listen to on their transistors blaring away at a distance of thirty feet. I'm actually thinking less these days in terms of mass consumption and more in terms of an elitist thing.

I'm personally bored with stereo. I had this conversation with Peter Gabriel recently and he's been checking out the Holophonic system, which is Ambisonic's direct competitor. I'm going to build a room in my house which will be set up ambisonically and start designing music for it. Not many people are going to be able to hear my music as it's supposed to be, but the ones who do will be bowled over by it. I'm going out on that limb, as it were.

Why?

Well, I think all science is trying to produce something as fine as the human body, so I think Ambisonics is the aural area of it. I've decided to abort the domestic sound system as a yardstick and stop thinking 'stereo'.

Since it's not possible to reproduce the 'live' experience of your music perfectly, and probably not desirable anyway, what exactly are you trying to produce?

Something different! I've tried a lot of different ways of recording and always found something lacking in all of them. I've now come to the conclusion that what's needed is better placement of sounds and I'm starting to write things with that in mind. I'm now thinking ambisonically, I wasn't before.

For example, I'm thinking of rhythmic patterns which could be split into eight or so different sound sources. I'm even thinking of producing quite gauche pieces of music in order to show off the effect to advantage, and make it very obvious for the listening public. All I know is that if musicians are excited about something new, then soon everybody else will be. If I can excite one other musician with this system, even if it means keeping 'open house'... then I'll feel that I've achieved something.

Did you use the Calrec Soundfield microphone on this album?

No. We only applied Ambisonic treatment in the mixing stage. As an idea, it was almost an appendage.

The album was recorded both here in London and in Rio de Janeiro. Why did you go to Rio?

Because I wanted to do an album that was rhythmically based, and they've got the best drummers in the world. It's a society which is built on rhythm, unlike here. For instance, you can get a barfull of people who'll all start playing matchboxes... you can actually get a samba out of a matchbox, but it has to be heard to be believed. In Brazil, they tend to think 'one man, one drum' and they're used to playing in large ensembles. I simply wanted to capture some of that spirit for the album and I think I have if you listen to the track called 'The Rio Connection'.

What are your future plans?

I'm always thinking in terms of several projects. I'm not thinking of touring with this album yet, I'd prefer to record a live album later. I've been talking to Geoff Barton about mixing a live album ambisonically. That'll be a different thing again. I think you can only push this Ambisonics thing by direct illustration, and as a musician that's my job, that's what I'm trying to do. The proof of the pudding is in the listening.