Peter Gabriel, man of a thousand disguises and a hundred musical styles, discusses the effect modern technology has had on his work, and gives some advice to adventurous musicians hoping to follow in his footsteps.

After a career spanning nearly two decades, Peter Gabriel is still one of rock music's major talents. His calm, intelligent songwriting - comprising an appealing sense of melody with lyrics that range from the celebratory to the disquieting - has been an inspiration to generations of musicians, even though Gabriel himself would claim to be proficient on only one instrument: the flute.

He began his career fronting a five-piece public-school pop combo in the late sixties. The band was called Genesis, and under the auspices of entrepreneur Jonathan King, it was an almost unmitigated disaster.

However, the arrival of progressive rock appealed to the creative instincts of Gabriel and keyboardist Tony Banks, and with the addition of Steve Hackett on guitar and Phil Collins on drums.

Genesis became one of the most adventurous bands of their era. In the early seventies, the band unleashed a succession of albums containing complex, classically-inspired rock pieces the like of which had scarcely been heard before. The discs were accompanied by some ground-breaking live shows, which saw Gabriel placing the emphasis on image and theatre, more than a decade before the video promo 'discovered' those elements and gave them a permanent place in the marketing of pop.

But The Lamb Lies Down on Broadway proved to be Gabriel's last album with Genesis when it appeared in 1974. Constricted by the band's massive instrumental virtuosity which was gradually pulling the emphasis away from soulful performance, Gabriel quit the band just as it was beginning to enjoy mass success.

His first album without Genesis, titled simply Peter Gabriel, was a cosmopolitan mixture of songs from a huge range of musical sources, robbed of any real continuity by the production excesses of Bob Ezrin. Some songs - notably the apocalyptic 'Here Comes the Flood' - survived Ezrin's penchant for kitsch gimmickry and over-elaborate arrangement, and a single - 'Solsbury Hill' - did sufficiently well to bring the name of Peter Gabriel into the homes of thousands, outside the band with which he had always been associated.

Gabriel's second LP, which carried the same title as its predecessor and featured a similar front-cover design in keeping with the artist's intention to produce a series of records similar to a volume of magazines, saw a radical change in style, with Robert Fripp at the faders, and a bigger emphasis on the emerging synth technology of the late seventies. Yet if anything, the second album didn't have quite the songwriting genius that Gabriel had exhibited in the past, and with the honourable exceptions of 'Mother of Violence' and 'White Shadow', it was a case of too much technique, not enough artistry.

All those ills were cured with the arrival of the third PG album. With Steve Lillywhite co-producing, and the likes of Kate Bush, Phil Collins and Paul Weller playing, the third incarnation of Peter Gabriel was a magnificent collection of haunting, delicate and occasionally beautiful material. Collins' full-frontal gated drum sound was thrown into light relief by the ingenious omission of all hi-hats and cymbals, and synthesisers - with US session keyboardsman Larry Fast at the controls - played a bigger part than ever.

The irresistibly catchy 'Games Without Frontiers' gave Gabriel his biggest hit yet in the singles charts, but on the album, the two most significant tracks were the marimba-jive of 'No Self Control' and the long, ethnic lament of 'Biko'. Those two songs were evidence of Gabriel's new-found desire to investigate the possibilities of 'other musics' - ancient African, Latin, Indian, and Balinese forms, among others, that would form the basis of his fourth album, released in the autumn of 1982.

Before that, Gabriel had been a guiding light behind the formation of WOMAD (World of Music Arts and Dance), whose aim was to bring musical cultures from throughout the world into people's homes in Britain, Europe and the US. WOMAD'S first major live event was a disaster, but a one-off reunion gig with Genesis a year later eased Gabriel's financial problems, and now WOMAD is flourishing.

Paradoxically, it was Gabriel's acquisition of the highest musical technology then available - the Fairlight CMI - that enabled him to realise his dream of combining ancient ethnic musical forms with one another. His fourth album saw what was almost the abandonment of traditional song structures, in favour of a looser, but still rigidly composed, style in which no musical influence - from whichever era and whichever part of the globe - seemed out of place.

Yet the album was panned by the critics who had decided they'd had enough of Gabriel's western intellectualism, and there was no sequel to 'Games Without Frontiers' in sight.

Three-and-a-half years later, Gabriel is just about to release his fifth album, which breaks with tradition by actually having a name - So. By comparison with the third and fourth LPs, it's a more conventional album, with Gabriel's strength in traditional songwriting coming to the fore on the single, 'Sledgehammer' and the raucous, scathing dance workout of 'Big Time'.

Quieter, more elusive tracks such as the immaculate 'Mercy Street' and the desperate 'Milgrims 37', show more obvious evidence of Gabriel's collaboration with Eno engineer Daniel Lanois, who produced the latest album after doing the same job on last year's Birdy film soundtrack, which saw Gabriel craftily remixing and manipulating previously-recorded material to create a hypnotic, atmospheric melange of different sound textures.

In person, Gabriel is a quiet man, unassuming and softly spoken. His conversation is a little disjointed at first, but as soon as a favourite topic comes up, his face lights up with anticipation, and the words start to flow more freely.

Above all else, he gives the impression of a man who is proud of his achievements, but who has yet to become so conceited as to feel he has done all there is to do. Gabriel is honest enough to admit that he wishes he had done many things in his career a little differently, but in an artist so full of conflicting ideas and motivations, perhaps that's not surprising.

The new album has been a long time coming. Can you give us some idea of why it took so long to produce?

Well, there's a variety of reasons. The first is that I've been in the music business for a long time now, and I've got no real desire to be part of the rock 'n' roll production circus, which is album, tour, album, tour. I wanted to follow up other projects. In this case I was doing some touring, then I mixed the live album, and then I did a bit more touring. Then I did the soundtrack for Birdy, and after that I did a little bit of travelling: I got to Brazil a couple of times, and then to Senegal, which was fantastically exciting for me.

Armed with a tape recorder, no doubt.

That's right, yes: 'Whitey With Machine'. You can do a lot of that sampling in your armchair at home, but it's a lot more fun to travel a bit.

Only one of the things I recorded while I was away actually ended up on the album — the Brazilian percussion on 'Mercy Street'.

That's based on a traditional rhythm called Forro, which apparently originated from parties which British and Irish immigrant railway workers used to hold when they were building the railway lines in Brazil. These parties were 'for all', and the Brazilians, who couldn't speak English that well, turned it into 'Forro'.

'Don't Break This Rhythm' is the song in its original form: the two share the same rhythm track. I completely re-wrote the verse and the words, and I felt that 'Mercy Street' and 'Don't Break This Rhythm' were different enough to warrant putting out separately, so that's why one is on the album while the other is the B-side of 'Sledgehammer'.

Do you have any problems coming up with material, or is there an excess of it?

There's always an excess of ideas. But finishing material is a bitch for me. I'm very bad at finishing lyrics off. Getting the ideas and scribbling the first few lines is easy, but from there on, it takes me a long while because I'm constantly re-writing. I think I'm my own best critic when it comes to music, but lyrically, I feel I have a few shortcomings.

"I've always needed someone to bounce ideas off. I've never liked the idea of calling my records 'solo albums', because all the people I work with contribute ideas."

The new album is more conventional than the last: it doesn't spring as many surprises. Did you consciously go back to more traditional forms and structures?

In songwriting, the answer is yes. That's partly as a result of doing the Birdy soundtrack, where I was working purely with atmosphere and no lyric. Having done a bit of that, I was more turned on by the idea of getting back to the song side of things.

'Sledgehammer' is definitely an attempt to re-create sixties soul style. Certain sections of 'Don't Give Up' have gospel elements. In some ways I'm still very keen on the craft of songwriting, but my loyalties are split between trying to explore and develop the songwriting aspects, and then the other side which deals more with soundscape and atmosphere.

In the past you've succeeded in blending those two things very well. It seems to me that this time, the two interests have been split.

The things on the new album are subtler, yes. But there are some sound things there that I'm proud of, and I think there's some good work that Daniel Lanois and the other musicians did in terms of building pictures.

What sort of ideas did Dan Lanois give you ? His forte seems very much to be sound treatments.

I'd been recommended him for Birdy for basically those reasons, yes. And because that worked well and we built up a good working relationship, I asked him to do this album — though that wasn't the original intention.

But he also functioned as a traditional producer as well: criticising material, encouraging me to work in certain areas and so on.

I've always needed someone to bounce ideas off. I've never liked the idea of calling my records 'solo albums', because all the people I work with contribute ideas. I've done one or two sessions on my own, but for the four albums I've always worked with separate producers. And in reality, even if you work only with an engineer, that engineer is contributing ideas anyway, so it's just a matter of how you label it.

If I wasn't writing the stuff, that might be different. But I don't like having to worry too much about the mechanics. Because I normally take so long over the writing, I need some feedback in the more general production areas.

You mentioned soundscaping. Where was the album recorded?

All at my home studio, except for one week that we did in New York. The real pleasure of having a studio setup of my own is that I can experiment in a way that I could never afford to do in a commercial studio.

The setup is a 24-track now. We had two 24-track machines at one point, mainly to do a sort of alternate reel system, which gives you a reel on which to develop percussion or vocal ideas, and then you fly two or four tracks back from that onto the main working reel.

But I had horrendous synchronisation problems that probably cost about four weeks of the album. It was incredibly frustrating and upsetting.

In the end we got so fed up with it that we hired in a Mitsubishi 32-track digital, which had sufficient tracks for us to do everything on one reel. The Mitsubishi worked well for us. I'd love to do an A-B against the Sony system.

For all the bullshit, the only thing I trust is being blindfolded and choosing between two or three sound sources. With the tests we ran, I actually preferred bass and drums on the analogue; whether it was distortion or not, there was something giving that sense of power which I enjoyed better. But the clarity on the highs, the presence and the transparency of the digital wins out in those areas.

I think the digital people still have some way to go - Probably what will happen is that the techniques will be further refined, and then people will hark back to the sound of the early digital.

The album wasn't mastered digitally. We started off working with a Sony PCM-F1, but we had some problems putting clicks on it, and the half-inch analogue proved more reliable, so we used that.

I got into the effects processing side much more heavily when I did Birdy. But I'm still mainly using the AMSs and Quantecs that I've had for a while. I'm not a great fan of the Lexicon, though the Yamaha REV1 I like: it has some things the Quantec doesn't, even though it covers much the same areas.

I'd never really entered the world of treatment chains the way I did with Daniel. The drum sound at the start of 'Milgrims 37' is one of the most striking examples of that, I think; it's something Dave Botteril set up with the AMS because Daniel was away. It's fascinating to see how the character of treatments changes as they are put through other treatments, and how their position in the chain affects the end result. But you need music that has enough space to hear that type of detail; it has to be sparse, and I think the album is quite sparse in a way.

Sticking with technology, but moving onto the music side, is there anything you used this time around that you didn't use three years ago ? Technology has moved ahead a fair bit...

The only new things were the Emulator II and the Linn 9000, though I didn't have time to get into the Linn properly before I started recording. Part of the catch with these things is that you do need time to get to know them. If the equipment arrives after you've started recording, you have to take it out of the studio to get to grips with it, which I didn't feel like doing in this instance.

"The CS80 has a great breathiness to it - human breath and its musical uses have always fascinated me..."

There's one synth that I'd used a little bit on my first album, but which Daniel Lanois re-introduced me to, and that is the Yamaha CS80. It has a great breathiness to it — human breath and its musical uses have always fascinated me — and a kind of organic character which I like.

But otherwise it's mainly Fairlight and Prophet 5. There's quite a lot of what I'd call a cheap organ sound, which I'm very fond of. And there's also a fair bit of 12-string guitar, which I'd avoided using since Genesis days, but which I have a different attitude towards now. I've always been a great fan of any instrument which can create two sounds that are slightly out of tune with each other, and the beating which results from that.

Similarly, there are some things recorded at two different speeds on 'Mercy Street', which ring in a magical way to me.

You were one of the first people to realise the potential of computer technology in making music. What kindled your interest initially?

Well, I'd always dreamed of a machine that allowed you to sample sounds and then play them from a keyboard, long before the Fairlight had even appeared.

I can remember Larry Fast telling me about this Australian guy who was working on a Carly Simon session and trying to sell this strange box. That guy was Peter Vogel, and he was having a really hard time because no-one was interested in the concept of sampling then. But Larry picked up on it, and for me it was a sort of fantasy come true.

Are there any tasks which you feel today's technology isn't accomplishing, but which might be possible soon?

There's one thing in particular that's concerned with samplers. Now I think that, particularly if you listen to a lot of tacky records, there's not enough performance being put into samples. That's partly the fault of the machines themselves, and partly the fault of the people using them.

What I would like to see developed is the idea of a layered performance. It starts from the theory that when you're playing a piece of music in real time, you respond differently because of your adrenalin and because you're functioning in a different way than you would if you were analysing a sequence on Page R or whatever.

When player pianos were really popular, there were great pianola players who were able to express a piano roll very well. They had certain parameters — volume, sustain, speed and so on — that they could influence.

With the sort of system I'm talking about, the first pass would let you sort out the basic composition and correction work, while the second pass would let you use a keyboard simply as an interface for performance: instead of giving you normal keyboard information, it could give you, say, vibrato. Or you could have two sounds or 16 sounds, so that the melody is switching between voices, and the internal composition is continually changing.

If these parameters were built up layer by layer and in real time, it would be like doing a dub mix: you'd have an idea and then just go for it. That kind of smash 'n' grab energy could give sound-sampling personality, and help to define character through performance.

If you have the same four-note sequence and you're using the same DX7 preset or whatever, then your version of that is going to be very similar to the next person's. But if you had 20 or 30 different expression parameters that you could define in this sort of real time, layered performance, then your version could be radically different to your friend's version.

I haven't expressed it very well, but there's definitely something there that's just crying out to be developed.

That sort of thing is starting to happen with MIDI master keyboards like the Yamaha KX88, which offer a whole load of parameters that can be assigned to different sorts of controllers - pedals, wheels and so on.

Yes. But that's still a little way from what I'm talking about because it takes so much time to program all those parameters for each sound or set of sounds. What you really need to be able to do is press two buttons and say: 'Go!'. It also confines you to the keyboard, whereas my system would allow you to use whichever interface you were most proficient at using, be it the keyboard, guitar, drums, maybe even dance movement or something like that.

Putting MIDI on a grand piano is also a step in the right direction. I've only tried it a couple of times but it is very immediate, very oriented towards performance. But the system doesn't have the technology to cope with storing different sets of parameters. What I'd ideally like to see is a system that allowed you to store a whole set of parameters, remove them from their initial context, and then use them as a blueprint for future work, like a sort of stencil or mask. That could be quite interesting, I think.

So far, it seems a lot of the newer techniques haven't resulted in very much musical or technical change. Are you excited about things like sampling technology becoming accessible to a much broader range of people?

Oh yes. I think it's absolutely fantastic.

There's always a temptation to copy what already exists when you're presented with new tools for the first time, because you don't have any other set of references.

But through time, working with the equipment you have gives you some insight into its quirks and its personality, and that in turn can make you think in different ways.

The wonderful thing about all this technology — which has so far been available only to the fortunate few — becoming available to, we hope, everybody with a home computer, is that anybody who wants to be a musician can be a musician. You don't need to spend years training, and you can get great sounds to develop things with almost immediately.

I think as we go on from the sort of new ambient music that is beginning to come out from the fringes now, that there will be a sort of home-made hybrid which comes out of sampling technology and so on.

"What's happening is that the young musician is sitting at home with a Portastudio, surrounded by keyboards and drum machines, and not interacting with other people at all."

The most important thing is that more and more people become involved with creative processes. One idea is that you have your own mixing facilities with your home computer, so that instead of just getting someone's album and listening to it passively, you can override the master and start doing your own mix, your own intepretation of your favourite record.

Interactive art...

Exactly. It's sometimes difficult to draw the line between where your art ends and somebody else's art begins. That's a problem which is besetting sampling at the moment, and there are no easy answers to it. But I do know that if a composer creates a piece of music in four sections, and a listener then rearranges those sections so that they run in a different order, then that listener has done something more rewarding than simply sitting in front of the hi-fi and being spoonfed.

I think the same is going to happen in the visual arts as well, because we're just starting to see video synthesisers appearing now, and they'll soon be open to as wide a market as the audio synthesiser.

For kids and for any non-professional people, the opportunities are going to be there, and I think it'll be a lot more challenging. It should, in the long run, encourage the growth of some strange entertainments, outside of the main commercial marketplace.

That sounds almost like a return to the Victorian idea of the family gathering around the piano for an evening's entertainment. People could start to make their own entertainment in the home in a way they haven't done for some while...

Yes. I think that's one of the most positive things about it, that people start to become active consumers rather than passive ones.

At the moment I think there's a slight negative factor, though, in what I term the bedroom effect. What's happening is that the young musician is sitting at home with a Portastudio, surrounded by keyboards and drum machines, and not interacting with other people at all. I'd strongly encourage anyone who's doing that to work with other musicians, because I'm sure that what I do has been made a lot stronger just through working with other people. Creating music alone can work in some cases, but you always get this aura of someone's private room: I think sometimes a blast of fresh air tends to make for a more attractive picture.

The problem is that collaborating with other musicians isn't always that easy. Drums, for instance, are a critical thing. I've worked with some of the best drummers in the world and none of them can keep time as well as the cheapest drum machine; but what they put in as musicians, in terms of feel, can be a lot more than most drum machines. And having the patience to program that kind of feel into a drum machine, even where it's possible, isn't easy.

That leads us on to the problem of getting caught up in modern instruments that can be too complex for their own good. With a system such as the Fairlight, which is constantly offering the user another 16 sets of possibilities at every turn, do you feel the technology can end up being a distraction?

It certainly can be a distraction, yes. At the moment I'm still working with the Series II Fairlight, though I'm very keen to get into the new version as soon as I can. But I know that getting to grips with the Series III will be like going back to school again.

At the same time, it's partly the responsibility of the manufacturers to make their devices more and more accessible to dumbos such as myself.

What I really hate is the sort of technical elitism, whereby people ruffle their feathers because they have a few technical tricks up their sleeve that nobody else on the block can do. It's such bullshit. You don't have to understand how a car works to be a good driver and to get from A to B. So although I'd never take anything away from the people who do have great technical knowledge — and there are people like that who can deliver some great music, too — I don't think it should be an elitist role.

I happen to think that simplicity is the hallmark of good design. So even though the technical operations of a machine may be complex, it should still be child's play to use, because all it's really offering you is another set of choices each time, and there's nothing so difficult about that.

How closely do you still ally yourself with the movement that wants to bring world music together, and put it in front of as wide an audience as possible?

Oh, I'm very close still, yes. Instigating WOMAD is one of the things I'm most proud of, out of all the things I've done. I'm no longer involved in the organisation, but that isn't necessarily what I'm good at anyway. I'm very pleased because after a potential bankruptcy, there's now talk of big things happening.

I think, in more general terms, that our music will benefit enormously from having that presence around us — music culture from different countries. Reggae has influenced some of our music and now has a minority but firm hold over certain sectors, and it's good to see that the Virgin Megastore has a huge rack full of different sorts of African music.

I know that coming across new visions and new groups has liberated my writing. That's still happening now. Even 'Don't Give Up', which is essentially a ballad, still has a groove which I think is very unusual. It ends up sounding like a normal song, but the musical ideas behind it are quite strange.

There's so much to learn out there. I feel very much like a novice, a naive enthusiast. But I quite enjoy that. Going to Senegal was great. The people were very welcoming, and I felt a sense of community with the local musicians straight away. The music is vibrant and vital and un-self-conscious in a way that modern rock music isn't.

Moving closer to home, do you listen to a lot of contemporary chart or dance music?

Some of it. Tears for Fears brought me the Blue Nile tape which I think is great, they're a really good band. They've now got a second album which is due out sometime. But they're another act that takes its time — they're not exactly fast.

Actually, it annoys me sometimes that there's so much pressure on us to come up with new albums quickly. If a novelist wants to take seven years over writing a new book, then nobody gets on his back asking for it sooner. I don't see why rock musicians should be any different. Some people can turn out an album in seven days, others take seven years. And in the end that isn't too important, anyway; what matters is what's on the record, not how long it took to produce.

If you can see that far ahead, have you any idea what is your next move likely to be?

Well, I'd like to do a bit of playing live again, because I still enjoy it. I'll certainly be out touring by the end of the year, if not before. The question is whether I do a big visual show or whether I do something more down-to-earth. And there are a couple of benefit gigs I've been asked to play this summer, which I'll probably do.

On the recording side, do you see yourself continuing to split your musical personality in two with future releases?

Probably. There's a whole load of bits and pieces that have got left around in the last 18 months that I'd like to develop as atmosphere pieces. But I'd also like to do an album which is much more traditional. I like good pop songwriting, so it's possible that there could be more Peter Gabriel pop songs.