The problem with synthesizers

February 23, 2009

Last week I got an email from a student conductor who is performing a piece by John Adams, needing a synthesizer, specifically a Kurzweil K2500 or 2600. I knew that Robert Winter had a K2500 in his studio, otherwise I had no idea what to recommend. I asked him to tell me what patches the score called for, thinking we could come up with some generic sounds that would come close to what John wanted. The problem is that both synthesizers are no longer in production, and, although there are still some used ones around, there probably won’t be any in a few years.

Our world-famous composer is now tasting a bit of what I experienced: investing time and energy in writing music for instruments that are dead.

Between 1992 and 1994, I wrote three cantatas for gay choruses whose core instrumentation included three or four live synthesizers: “Hidden Legacies”, “Letters to the Future”, and “Flashpoint/Stonewall”. The synthesizers I used were the Yamaha DX7, the Yamaha TX802, the Roland D50, and the Kurzweil K2000 — all now long dead. I still own these synthesizers, but there is no software available that can download the sounds into them. I have held onto old computers with old operating systems in hopes that someday I can resurrect them. Many of the sounds that I used were sounds I designed myself that have no equivalent in traditional orchestral instruments or synthesizers, so there really is no substitute for these old dead sounds.

An alternative is to create a recording of the accompaniment in a kind of Music Minus One format. The problem here is that conductors are control freaks: they are opinionated about tempo, balance, and dynamics. They don’t like being a slave to a click track or the shackles of a prerecorded accompaniment, and for this reason alone, this doesn’t seem to be a viable option.

Synthesizer manufacturers treat their products like cars: they are hot for a year, then they take them out of production and replace them with a new, updated, better and improved version. These are financial decisions, not artistic decisions. So, composers who have composed music for them, and I mean notated music in the grand classical tradition, are just out of luck. What does John Adams plan to do about this? or his publisher, Boosey and Hawkes? Will the piece die because of this problem?

I have always hoped that someday a Glass Bead Game-type instrument would appear one day where one could punch in a number and have any historical synthesizer available at their fingertips. The problem seems to be that Roland owns its sound architecture, and Yamaha owns its, and so on.

Recently, virtual synthesizers have replaced physical ones. But the problem has been that we can’t access ANY of the sounds we created on the original instruments, much less import our old sound libraries into them. What good was that? Why has no one tried to update Opcode’s librarian software, or its suite of programming software designed for individual synthesizers? Part of the answer may be that certain sounds sound dated: but since when has that stopped a musician? We have a panoply of “early instruments” carefully maintained and used to faithfully reproduce the old traditions of performance and repertoire, but new music gets relegated to the chopped liver category.

Musicians who work in popular media use synthesizers for a CD, and then use them on the tour. After that, they move on to the next album. Their final product is a CD, not a score. But for those of us trained to be composers who create scores, so that people can re-create our music in years to come, we are out of luck.

With this current reality, I would advise all composers to NOT write for synthesizers in the traditional way, meaning notating music with patch changes and specifying synthesizers. Rather, they should specify “Synthesizer” and use generic terms that describe the kind of sounds you have in mind.

The instrumentation of the orchestra is closed. The only section whose instrumentation still evolves is the percussion section. Sure, you can put synthesizers in your score, just as you can a sitar, or erhu, or koto, but there is no guarantee that, in the future, anyone will be able to play those parts.

{ 6 comments… read them below or add one }

daviddas February 23, 2009 at 8:18 am

Thanks for the article, Roger. I would say this, though: it’s a shame to encourage composers to limit themselves only to the instruments that are commonly available. That’s kind of like saying web programmers shouldn’t use Flash because not everyone has Flash.

You make a valid point that modern synths don’t have the longevity that traditional reliable orchestral instruments do. But there are a lot of web sites out there that offer sampled versions of classic synths, and at least two or three pieces of software that can automatically sample a piece of gear and map it in a sampler (Autosampler, etc.). Samplers, too, will change over time, but there are some universal formats that will certainly be readable and convertable for decades (or longer) to come – formats like Giga, EXS, or Kontakt.

Maybe composers should begin offering downloads of the patches required to play their pieces. It’s certainly not hard to do.

I like this solution better than telling composers, “Only compose for the tried-and-true instruments!” While there will always be a place for traditional orchestral writing, there’s definitely a place for expanding the sound palette of the traditional orchestra as we look forward. I think it’s part of orchestral music evolving with the times.

I do kind of think that John Adams’ (and similar) attempts to add a synth part to the orchestra weren’t extremely well developed. The synth parts are kind of “token” and yes, now they sound dated. Also, they’re somewhat unnecessary — most of his pieces play fine without the synth part, or (hypothetically) those synth parts could be reorchestrated to other instruments in the orchestra with no ill effects to the music. However, as a composer and sound designer myself, I think of the amazing things that would be possible in combining a synth like Absynth with traditional orchestral writing.

Just a thought…

Roger Bourland February 23, 2009 at 8:35 am

Thanks for the response David. If this article pisses off the right people, then maybe there will be hope that the orchestras in the future will be able to play old (now, new) orchestral pieces with unorthodox orchestration.

I hope you understand how sad, infuriating, and unfortunate it is to invest so much time into something you think is a great step in the evolution of live performance, only to see it discarded with last year’s models.

It occurs to me that another composer might say: Who cares? I’m writing for audiences right now; who cares about the future? And that is a good attitude to have as well. The old “I’m writing for future audiences” is just weird.

Brad Wood February 23, 2009 at 10:53 am

Something perverse in me is intrigued at the prospect of having to learn to orchestrate for traditional acoustical instruments to reproduce the sounds of obsolete synthesizers.

I guess virtually none of the old synths have been held in high enough regard to be preserved—while more specific older electronic instruments like the theremin and ondes Martenot are still available, in some limited way? That’s sad.

The other rather devilish thought is one almost voiced explicitly by Roger: that we may see the rise of a not-very-old instrument revival of sorts focused on recreating, restoring and maintaining old synths. I want to see those one-time-ubiquitous DX7s on the stage, with the Yamaha logo and all!

And despite the difficulties, surely a sufficiently virtuosic group of players could figure ways to duplicate the FM synthesis sounds—hilariously, developed to conserve computational resources in those olden times. Foss gave us some examples of just how unlikely some of the sounds out of woodwinds can be when he was on his multiphonics kick (I recall it took days to get some of the noises out of my system heard in a performance of his Cave of the Winds).

(cackles evilly)

Grant Damron - Boosey & Hawkes February 24, 2009 at 8:56 am

Thanks for the post, Roger. It illustrates out one of the major issues modern composers — as well as their publishers and the orchestras that present their music — have been trying to tackle in recent years. It’s true that, while the majority of orchestral instruments were perfected in previous centuries, the nature of synthesis and sampling ensures that it will continue to evolve for a very, very long time.

And this is something that John Adams has been aware of since he began experimenting with electronic sounds decades ago. You can practically trace the history of commercial synthesizers throught his work, The Wound Dresser: a synthesizer part originally composed on a Yamaha HX Electone, then adapted for SY-77, then SY-99, adapted again for Kurzweil K2000 (or higher), and now available for Kontakt 3.

Luckily, John has a fantastic sound designer on his team, Mark Grey, who has been working to reformat the sampler and synthesizer parts for Kontakt and has done so for John’s most performed works. It is, however, a laborious project and will take awhile to complete. Whenever the Kontakt 3 version is available, we send a copy of these samples alongside whatever version for hardware sampler that we have on hand — generally for K2000 or K2500 which are also compatible with newer Kurzweils and a wide variety of other keyboards — so that the orchestra can choose whichever format suits them. When we them formatted for Kontakt, we are also sending raw samples which can be mapped onto any sampler.

There a matter of economics too. We are in a period of transition, no question. For many orchestras with smaller budgets, it is not feasible to invest in a computer system powerful enuogh to reliably run a software synth, not to mention the cost of the program itself. For them, renting a K2500 is still the most affordable option. They are still reasonably easy to acquire through instrument rental agencies and many keyboard players still own one. Few orchestras have sound designers and engineers in house at this point, so we must also ensure there is an option that requires little technical knowledge so that the technical requirements do not become too daunting. This is changing, however, and we at Boosey & Hawkes look forward to the day when there is a universal sampler format and this continual game of updates is behind us, and an engineer is as indispensible as a concertmaster.

In the meantime, we will continue to keep the patches that composers like John Adams and Steve Reich spent so much time crafting up to date. We are certainly not in the practice of letting incredible music die, and this is incredible music we’re talking about, after all.

I look forward to the next insightful post.

David Ocker February 25, 2009 at 1:29 am

Are you aware of the Adams “technical” page at ?? I can’t say whether this specific piece – whatever it is – is addressed there. Maybe Mark Grey has provided something helpful.

grunin February 25, 2009 at 9:37 pm

I asked composer Annie Gosfield about this back in the mid-90s, as her work was build very specifically on the sampled sounds of detuned pianos.

After reviewing the problems and possibilities, she ended by saying “…or you can just accept the fact that some aspects of your music will be ephemeral.”

Actually, I think the electronic keyboards you discuss are rapidly discarded by musicians in favor of “this year’s model” for a very good reason: because the player has relatively little control over individual notes, any given patch is soon perceived as boring (‘dated’).

Just think of how many expressive choices a violinist has with every note. Compared to that, a synth is less expressive than your average mallet percussion.

Of course this is true of all keyboard instruments, with the unique exception of the piano (which compensates in ways too complex to state briefly).

Previous post:

Next post: