Magazine Archive

Home -> Magazines -> Issues -> Articles in this issue -> View

Back To Bach

The Making of an Album

Article from Sound On Sound, July 1992

When Julian Colbeck went back to Bach to find musical inspiration, little did he think he'd end up recording a whole album.

I don't remember the precise day, but it was one of those times when it seems like you've played everything a zillion times before. Same old chords, same old sounds, same old grooves. It's the price we pay for today's sky's-the-limit technology. In the old days when you had to make do with just a piano or a guitar, composition was somehow facilitated by the limited physical tools at your disposal. Current technology, on the other hand, has now become so all-encompassing that those of us who like to think of ourselves as musicians have become, unwittingly perhaps, engineers, programmers, and computer operators. The net result is that so much of today's music can be summarised as, to borrow the punchline from a joke I've long forgotten: shit — but beautifully cooked.

In previous times of disenchantment I have found comfort and inspiration in going back to some Chopin, or Beethoven, or Debussy. On this particular day I went back to Bach. To the 48 fugues and preludes, that most mind-cleansing, illuminating collection of keyboard pieces written in the early 1700s to demonstrate the benefits of a new system of tuning, called equal tempered, that allowed a single instrument to be played satisfactorily in all 12 keys.

Now my classical chops were always a bit skimpy but I could negotiate Prelude One and stumble through one or two of the others if I put my mind to it. Then on went a red light, as I realised that maybe technology wasn't completely redundant. If I simply recorded each piece into the sequencer I could practise each hand at a time at my own pace.

So I set about inputting Prelude III, the galloping C# major, into my MC500 sequencer. I was careful about inputting the music, without being too pernickety. Since much of the piece is a series of equal-length 16th notes, it was actually quicker, never mind easier, to write in step time. On the MC, writing in step time is also a doddle. Once you've specified your note length, you can just keep banging away as quickly or as slowly as you like.

Playing it back was a revelation. This music had been input in the most seemingly sterile fashion. No dynamics to speak of, no varying note lengths, no rubato. Yet the piece sang as if freshly composed. The music, inherently, had life. This so intrigued me that I quite forgot about whipping out the right hand in order to bash on and get to grips with that murderous left hand part. My foot simply started tapping on the offbeat. By the time the piece had got to the series of dramatic question and answers at bar 28, I was completely hooked. I swung around and began to add some drums. These neither sounded inappropriate, nor that bad! Almost instinctively, I just slammed the MC into record and put down an off-the-wall rhythm track using my standard M1 drums. Some of it worked, and some of it was a bit crass. I went back and just blipped out the crass bits and listened to what was left. I found myself humming a counter melody that seemed like it should be played on a sort of ghostly electric piano.

It took me a while to rationalise what I felt, into a part that actually worked. Unlike in most pop or rock music, where even the most convoluted songs will linger for suitably reviving periods of time on something like E major, your average classical composer in general, and Bach in particular, seldom rests on anything approaching a normal chord for longer than a few beats. Slowly I built up the piece, deliberately going against the grain; taking breaks where no break in the prelude actually existed, adding a manic cod jazz acoustic guitar I'd just programmed on a recently arrived Wavestation.

Bach's second prelude is busy enough as it is. With these extra parts, and no attention being paid to placement or arrangement, it soon began to get out of hand, and, merely to give myself a bit of space for a couple of minutes, I temporarily blipped out the Bach. Amazingly, the piece still kind of held together.

I carried on in this vein, adding a few more parts, and by the end of the evening I'd got this curious new 'composition' on my hands, along with the germ of an idea that perhaps I could do the same with a couple more.

Next day I looked at Prelude III. Again, inputting it didn't take at all long, though a full track took longer to put down than I anticipated, the odd bar falling into place quite swiftly but at other times the whole thing slid wildly off the rails. By careful listening, plus the odd bit of careful reading and the odd bit of luck, I completed what I felt was a reasonable obligato type part. Other parts quickly revealed themselves as being necessary — a picky electric guitar part inspired by the piece's general 'up' feel and tempo, and a bell-based melody line to go with it. Some bass, the odd drum, some heroic piano lines, and finally some oompah brass. This is German music after all.

Gingerly I removed JSB. It was weird, but the track stood up, alone.

On the third night I tackled Prelude I, the Ave Maria, a piece you really can break down into recognisable chords for much of its duration. Precisely because a chord structure was so self-evident, I decided to ignore that completely and set about recording a series of interweaving lines, single notes, and effects. In particular I tried to throw myself completely off the rhythmic scent by continually turning around the emphases of the percussion. The glue, I felt, should be vocals, and Chris Macleod had programmed some excellent vocals on a Proteus. Coupled with some suitably ethereal M1 vox, they did the trick nicely, except in the parts where I wanted some across the beat crochet vocal stabs. I wanted more attack, more of a 'Hal' sound. Fruitless programming hours later we resorted to the most simple and effective solution. Sample myself going 'Ha!', snip off just the breath part and trigger it, as a single note event, alongside each of the vocal chord stabs. It worked a treat.

I listened to all three pieces with and without their respective preludes underneath. Two things struck me: first that I seemed to be learning more about how Bach came to write these pieces by exploring them from the inside. Secondly, that after a lifetime of appeasing A&R departments by getting to the chorus quicker, simplifying the chords, taking out half the lyrics and all of that stuff, seldom had I heard anything less likely to get even the most fleeting of hearings from a record company.

Needless to say, I was offered a deal on the project almost immediately.


My goal had become to record as many of the fugues and preludes as could reasonably fit onto a CD. But exactly how, which, and what? I felt that if I just used my own compositions, leaving the Bach element simply as a sort of ghost, then only people capable of playing the fugues and preludes themselves would be able to hear how the old and new music fitted together. On the other hand, the full versions made extremely demanding listening — sometimes following two totally separate rhythms, melodies and counter melodies. Would I be chancing my arm just that bit too far by doing this? Yes, I figured, I would.

The only solution was to feature both, one comprising just my original music, without the Bach underneath, and one full version, complete with the fugue or prelude.

That threw up the first problem: how to physically present the music. In the old days I'd have put all the ex-Bach versions on side 1 and the inc-Bach versions on side 2. But CDs don't work like that. And EG (the record label who offered me the deal) told me right from the start that this would only be released on CD and tape. Having at least made the decision to use two versions per piece, I decided to wait and see how they turned out before I committed to a running order.

Swift mental arithmetic revealed that I could use no more than a dozen fugues or preludes, making 24 tracks in all. But which ones?

Eventually, I came up with the simplest solution, that of going through each piece in turn, starting at Prelude I.

I'd planned to crack on with the project, for sure, but nothing prepared me for EG's bombshell that they were hoping it could be delivered by, let me see... would mid August be okay?

Since this was late June, that gave me less than two months to write, record, mix, and master 24 pieces of almost impossibly complex music using a technique that, so far as I was aware, no-one had ever attempted before. Yes, I heard myself say, that should be alright!

I made a fundamental decision that although I would play most if not all my own compositions, I would find a concert pianist to play the Bach. I did this for two reasons — firstly there just wasn't time enough for me to learn all the fugues and preludes up to recording standard, and secondly because I wanted the input and perspective of a working classical player. Only one name appeared on my list: Jonathan Cohen, currently the musical brains behind Channel 4's The Music Show, but also a prodigiously talented classical keyboard player, a Royal Academy of Music scholar, and one the youngest ever Fellows of the Royal College of Organists.

The method of construction we agreed upon was to go over all the pieces first, to settle upon tempi and the overall approach I'd be taking on each new composition. Then I would disappear and complete the writing, working to a Bach track that I had simply input on computer, which Jonathan would subsequently replace in real time, and on an acoustic grand piano, in the studio.

Another difficult decision was the choice of recording technology. Although the project was basically a blend of MIDI-driven keyboard rig and (stereo) grand piano, I did not want to rule out the possibility of having guest appearances by other real-time players. In other words, could I risk a straightforward time code DAT-recorded grand piano? In the end, the cost, flexibility, and relative security of a regular studio and regular 24-track analogue tape won the day. Jonathan figured he could record the lot in two days. Who was I to argue?

Time at our disposal dictated a streamlined and straightforward approach to working between Chris Macleod and myself. In fact when the project was signed Chris was still on tour in America with Yes, and time was so tight that I picked him up from Heathrow and drove him straight back to the studio.


All the initial sequencing was done on the MC500, but once the basic tracks had been recorded, we downloaded the sequence onto Cubase. Aside from the unsettling micro-second tempo changes as the comparative coarseness of the MC's timing converted itself into the highly analytic Cubase approach, there seemed to be no downside to working in this manner.

I wish I could reveal some grand design in my initial approach to the pieces. Aside from wanting them all to sound different — a determination to explore this 18th century music from as many 'current' musical viewpoints as possible, I'm afraid there was none.

Writing a fugue is, by nature, a linear exercise, various themes overlapping, intertwining, inverting, repeating, modulating. I tried, in Fugue I, to develop my own themes regardless of the detail of the original. Thus Bach's developments and my own forge their own passage through the piece as if in fugue form between themselves. With this two-pronged attack Fugue I possibly makes most demands upon the ear harmonically out of the whole album. To counteract this I deliberately injected a relatively sane rhythm section, with Chris Macleod's programming going into overdrive on quite the largest and most rubbery bass sound I (or the cutting engineer at Townhouse!) had ever come across.

Fugue II came across as military from the moment I first heard it. By the time of writing I had also committed to using a Roland RSS system on the mix and was very excited by the prospect of working, for want of a better word, in another dimension. One track, I decided, had to be exposed to the point of hypothermia, in order to move an instrument around as if the player were actually walking around the room. A military snare drum seemed the obvious choice, and for accompaniment I chose a simple pipe flute, and a sort of Spanish civil war guitar.

From the start, the idea was to fuse the ex-Bach and inc-Bach pieces at the hip on this one, and have the Bach appear to join in and so form a smooth and logical progression from what had gone before. My original intention had been to move the military snare drum, via RSS, from a position in the distance directly behind you, through you, and on out in front of you, as the piece progressed. I was soon to learn that RSS is not yet quite that definable.

The accompanying instruments were played by a wonderfully expressive and breathing Ensoniq SD-1 flute, and a guitar patch I programmed on a Wavestation.

"Prelude V had always seemed the most song-like track of the project, and I had long felt that a humanising voice was needed somewhere on the album."

It was summer when I was writing. Occasionally there'd be one of those hot, languid days when what I really fancied doing was cruising along the coast in the South of France. Prelude IV transported me there, via a 'sequenced' ripple-effect M1 patch Chris Macleod called 'Osmosed', a program he'd written on a tour bus a few weeks earlier. I just found the right rhythm and simply programmed any notes that fitted the scale. To add that South of France feel, I used an Ensoniq patch we called 'French Cup'. One splash of that program and suddenly you were in that 1960s low-budget French movie, watching the girl in the headscarf snake through the hills above Cap d'Antibes in her open-topped MG.

Ever since enduring Jon Anderson's pathological obsession with choirs a couple of years ago, I've wanted to construct a piece of music entirely out of different choir samples. The Great Fugue, Fugue IV, seemed to provide that opportunity. I felt the complexity of the piece, where at times no less than five parts run concurrently, warranted the minimum of harmonic intrusion from me. Instead I tried to illuminate the piece by using different choir samples on each of the fugue's main themes. Chris Macleod and I assembled every conceivable choir we'd ever used, from a massive 8MB self-sampled choir I'd recorded in Virginia two years earlier, to M1 programs, new Proteus programs, and U110 patches (surprisingly good). At the time of recording, we were disappointed with the outcome. The range was just too great for even our most flexible samples. It just didn't, as they say, hang together.

Making a record is a succession of last chances. There's always another stage at which you can have another go from (fixing it in) the mix, to the remix, to the cut, to the pressings... Of all stages, the cut is probably the least understood. But on Fugue IV, time spent in the cutting room simply made all the difference, the power of EQ providing the glue that no amount of effects and samples seemed capable of delivering.

The point at which most of the fixing took place was during the recording of the acoustic piano. Prior to the session, Chris Macleod spent time cobbling together an approximation of each arrangement that could be played by nothing more than a succession of M1 Combinations. Still Cubase driven, these could simply run alongside the code in order to make sure everything was fitting properly. On the mix, we would then revert to triggering the full keyboard rig live.

Recording the piano was fascinating, not least because Jonathan had no option but to follow the internal timings set down by each sequence. The nature of the pieces meant that these were invariably strict. On the face of it, erasing all thoughts of human feel seems a depressing prospect. Quite the reverse, though. The more Jonathan tried to inject overt emotion, the less musical and more computerised the track sounded. On the other hand, if he just put his head down and played as strictly as he could, the track immediately sprang to life. The music itself was providing the emotion. A double dose was overkill.

The only hiccup over the two days was a strange clicking noise we started to hear on solo playback. Hours went by until we found the culprit: Jonathan's fingernails! On some tracks a trace can still be found too, since even a pair of wirecutters from maintenance could not remove all trace of nail from JC's fingers.

I had begun this project with few golden rules but one of them was that on no account would I interfere or take liberties with Bach's written music. Each piece that I wrote must fit exactly over its fugue or prelude as written. Although tempi are inevitably discretionary, for the most part I kept to within normally accepted bounds — except for Prelude V, that is.

The finished CD sleeve.


Prelude V was designed to be played at breakneck speed, and as such requires a deceptively advanced technique and/or giant hands to play. At half speed the piece becomes entirely playable and, to my way of thinking, a complete gem. First I set up a lolloping groove on two connected drum sources, a Cheetah MD16R and M1 drums, and then slowly built the track up from a Wavestation string pad. The observant may already be wondering what happened to the drums. They lasted a couple of remixes and then I'll admit to buckling to pressure from all sides that they were entirely superfluous to what the track had become.

Prelude V became guest appearance track. It started, simply enough, with a desire to infiltrate some of my synthesized guitar work with some real guitar, played by a real guitarist. I'd played with Steve Hackett earlier in the year and we'd discovered a mutual fascination with Baroque music. I'd also heard some of Steve's excellent Spanish guitar playing; he seemed the obvious candidate. Protesting wildly when I calmly set the score in front of him in the studio, Steve proceeded to deliver a warm and human Spanish guitar track that made me blush to my roots when I considered the, ahem, hack job I would have made of the part on a synth. There are times when keyboard players should resist the offers technology now affords us.

Prelude V had always seemed the most song-like track of the project, and I had long felt that a humanising voice was needed somewhere on the album. I toyed with the idea of simply hiring an actor to recite. I even contacted a few notables of my acquaintance with generally complicated and ultimately fruitless results. As with so many things, you end up working with friends because by doing so you can slip under the electric fence of 'business' and just get on with the job.

On the face of it, Mary Hopkin may seem a strange choice of vocalist for such a job. But behind her lilting and evocative contralto lies an immensely knowledgeable and imaginative musical brain, and what I needed was not simply a singer for a song, but for devising a tune to make sense of lyrics I had already selected. Very few people could even contemplate such a task.

The fact that Mary's sensibility was spot-on first time around is now irrelevant, because you'll never get to hear it. What finally made the album is something else entirely.

Running concurrent with this, an opportunity had presented itself to meet Brian Keenan, who, it turned out, had written a considerable amount of fine poetry during his extended period of captivity in the Middle East and who might, as a lover of things Bachian, be interested in reciting some of his work personally.

Understandably, Brian Keenan does not spend his time sitting in his front room at the Irish equivalent of 3, Railway Cuttings, County Wherever, with a listed phone number, waiting for the phone to ring. But finally, after months of delays, and when the rest of the album had long been finished, the day arrived when my patience gave out and I decided to pull the plug. Next day Brian called to "say how about this Saturday, in Ireland?"


Sitting at Heathrow, armed with a selection of microphones, a stand, and my trusty DAT recorder, jubilation at the prospect not only of meeting the man and hearing his poetry but finishing the album was quickly embraced and enveloped by the sea of dense fog currently rolling over the entire British Isles. Eventually, a flight to Dublin emerged out of the pea soup. Dublin, yes, but would our connecting flight have waited?

"No, I'm afraid it couldn't," said the girl on the Aer Lingus counter. "But we can get you quite near, and then we'll lay on surface transportation to your final destination."

The Republic of Ireland is one of the most wonderful places on earth. But as any visitor would testify, it is not a great place for getting from A to B in a hurry. By the time we made our final destination in order to pick up the rental car, said rental car part of the garage was well and truly closed. Resourceful to the last, our minibus driver simply drove around town looking for the garage owner's house! Ireland is like that. And sure enough we did find it. Our car, too.

We arrived with seconds to spare at our rendezvous with Brian Keenan. And he was nowhere to be seen. Guinness was the only solution. Several pints later, food made its way to the top of the list, and we were recommended a special bar/restaurant by the local police chief. We turned to go and ran straight into Brian Keenan. "Let me buy you a Guinness" was his opening gambit. Several Guinnae later, we set out once again for our bar/restaurant, this time with Brian, in his car, with me squeezed onto the backseat next to a large salmon he'd just been given. Ireland is like that, too.

Several more Guinnae and several dozen oysters later, thoughts of recording that evening had long retreated into the mercifully brisk night air. But timing is everything. I couldn't possibly risk tomorrow. Back at the hotel, I hurriedly unpack my mobile studio in my room as Pete and Brian sort through possible material. There is one clear favourite: 'Faces From Old Photos Rediscovered', haunting, redolent with musical imagery — and the right length!

Recording in a hotel bedroom is not the easiest thing to do at the best of times. The current situation, ie. five pints down and counting, is not being helped by the fact that my DAT recorder is noisy. Not the internal workings, you understand, but externally — the motor. My solution is to record Brian in the bedroom, and use the bathroom as a control room. Except that with the bathroom light on, the air-conditioner automatically comes on — noisier than my DAT.

So I turn the light off, and kneel down beside the pan with the door cracked, one eye on Brian, the other squinting at the meters. Brian Keenan has a unique, distinctive voice; naturally sibilant, and very soft. All I could do was record every which way and hope to hell I'd got it. There was no limiting, no EQ, no nuffin'.

Recording took about an hour all told, after which we celebratingly returned to the evening's main sport of Guinness-sampling. Thankfully there were a few days in between recording and laying Brian's performance over the track. When we did, it was one take, one timing, and pure magic. The only problem came when Brian, as it were, met Mary. The juxtaposition of the lyrics I had selected for Mary to sing with Brian's poem was, in the words of the record company 'dynamite'. Not in the best sense, either.

We had no choice: write new lyrics to the now existing tune and re-record it. Which is what we did, Mary somehow managing to write not only new lyrics that fitted the existing scansion, but ones that made sense, sounded good, and were apposite. A migraine during the recording was the final hurdle, and no sooner had her final note died away than Mary collapsed like a row of ninepins on the studio floor.


Prelude V was a mind-altering experience all round. Some months earlier, we'd been in constant touch with mind-altering experiences in the shape of Roland's RSS 'surround sound' system.

RSS is not an exact science. Maybe Roland never intended it to be. But if so, that's a shame, because the potential to expand into other dimensions is one of the few exciting prospects on offer in the current technologically 'all bets on' soundscape. Sceptics, and there are many, say that RSS simply plays about with phase reversal. Perhaps so; I'm not sufficiently au fait with pure electronics to pass comment. I'm more of the school that says if it works, don't analyse why too deeply, and don't complain.

The planned straight line movement of the snare drum in Fugue II was not possible. The best we could produce was a starting point some way over your right shoulder, and a travel that arced round the right hand side, concluding high up on the left hand side. The effect, while not quite what I'd intended, was still impressive, despite the virtual drummer seeming to disappear behind a large building as he passed through 90 degrees.

The other danger is, of course, mono, which we failed to take into account on the first mix, Prelude I. On this track the vocals have been RSS processed, and on a stereo system they will appear to be coming from quite deep behind you. In mono, they're gone. Provided you check for mono beforehand, you can invariably avoid this problem by slight re-jigging.

Totally by chance, the final track is the relentless, slow moving, but triumphant Fugue VI, on which I could hope for no better send-off than Milton Macdonald's seamless, one-take guitar solo, accompanying not merely my progression of doleful chords coerced from an effect-free Oberheim Matrix 1000, but Jonathan Cohen's sympatico piano playing, and the genius that will forever be JS Bach.

More from these topics

Browse by Topic:

Composing / Art



Previous Article in this issue

Turtle Power

Next article in this issue

Roland DM80

Publisher: Sound On Sound - SOS Publications Ltd.
The contents of this magazine are re-published here with the kind permission of SOS Publications Ltd.

The current copyright owner/s of this content may differ from the originally published copyright notice.
More details on copyright ownership...


Sound On Sound - Jul 1992

Feature by Julian Colbeck

Previous article in this issue:

> Turtle Power

Next article in this issue:

> Roland DM80

Help Support The Things You Love

mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.

If you value this resource, you can support this project - it really helps!

Donations for May 2024
Issues donated this month: 0

New issues that have been donated or scanned for us this month.

Funds donated this month: £24.00

All donations and support are gratefully appreciated - thank you.

Magazines Needed - Can You Help?

Do you have any of these magazine issues?

> See all issues we need

If so, and you can donate, lend or scan them to help complete our archive, please get in touch via the Contribute page - thanks!

If you're enjoying the site, please consider supporting me to help build this archive...

...with a one time Donation, or a recurring Donation of just £2 a month. It really helps - thank you!

Small Print

Terms of usePrivacy