When I first heard this I was thinking that it could almost be a @robertsyrett track.
I love Spacek! Here’s a sample from the album Curvatia:
I know people often dismiss digital synths but it’s hard to beat 24 oscillators for depth. (48 on the new Summit).
The biggest problem with the digital synths is that when I was younger all I thought existed was some awful plastic keyboards with salsa beats and tuba presets. I had no idea this whole time that there was a subtle artform in the background (digital or analog).
That being said, in many ways it has worked out the best. Yesterday I was using KQ Dixie (FM iOS synth), and I figured I would modulate some parameters over time like I would a subtractive synth. As soon as I got to the parameters I laughed, because I remembered how complicated it is to actually adjust FM parameters. This is probably a big reason we ended up here – we were too stupid back then (consumers) to have proper digital synths.
Interesting that you mention FM synthesis. FM synthesis wasn’t really practical until they developed digital oscillators. Analog oscillators just aren’t stable enough.
i don’t believe (even though it seems to have achieved the status of conventional wisdom)
that oscillator stability was the issue. I think there was a confluence of factors:
- the expense of having enough oscillators to do really interesting FM . There were a number of electronic music labs that had analog setups where FM was an important aspect of the sound. (This includes the first Buchla’s but in Britain they were exploring it, too)
- my understanding was that linear FM circuitry wasn’t cheap.
- Even though FM had been around, I don’t think it’s usefulness was appreciated till Chowning’s work both in terms of composition and technical research. Being freed from hardware limitations (his work was on computers), he could explore arbitrary numbers of oscillators.
I think the main factor that got in the way was expense. Until capable microprocessors of the ilk in a DX7 were cheap enough, there wasn’t an affordable way to create a practical complex FM synth.
Anyway, that’s my take.
You certainly have some valid points. I would agree that expense was the limiting factor. After all, frequency modulation as a method of radio signal modulation was introduced in the 1930s and was entirely analog. I probably should have said that analog VCOs stable enough for FM were too expensive to make them practical for a commercial product. Not only that, a typical FM operator has an envelope as well as an oscillator, which drives the cost up even more. In the case of FM radio the carrier was at a fixed frequency and the frequency could be stabilized with a crystal, but in the case of a voltage controlled oscillator, it is difficult (and expensive) to keep the unmodulated frequency from drifting. Since the ratio between carrier and modulator is critical, unstable oscillators make it difficult to achieve a reproducible sound. Useful monophonic FM doesn’t necessarily require more than a few oscillators. While the DX-7 had a max of 6 per voice, earlier Yamaha products used 2 or 4. The major benefit to very cheap digital oscillators was the ability to build polyphonic instruments. The OPL3 chip had 36 oscillators which could produce up to18 2 operator voices or 6 4 operator voices.
Chowning’s work was obviously instrumental in demonstrating the usefulness of FM (or PM) in musical synthesis, but it was Yamaha’s use of digital circuitry and VLSI that made it cheap enough to build a commercial product using this approach.
Great stuff but it should be in the FM synth mega thread
I agree, but I couldn’t figure out how to move them.
From what I understand, Yamaha was interested in FM entirely due to Chowning’s work and saw an opportunity to develop a synth that was in a totally new class. Stanford had patented the key breakthroughs in FM/PM algorithms. Chowning and another guy from CCRMA (who designed a lot of the DX series patches if I recall correctly) were key players through the whole process. Stanford (not Chowning—since he didn’t own the patents) earned a lot of money from the patents that they licensed to Yamaha.
It is to Yamaha’s credit that they recognized the potential of a synth focused in FM.
I really think it was cost and impracticality of all the envelope generators more than pitch stability that is the reason FM-focused commercial synths didn’t happen earlier. Although some would say that all the Buchla synths were pretty heavily FM-oriented from the beginning.
Yeah, it’s a little rich that Stanford claimed a patent on math that had already existed for hundreds of years but now could be used by computers, especially when the essentially the same technique was already in use by Buchla. But it helped out the music and art departments so I won’t complain too much.
I believe there was quite a bit more to the patents than codifying what was widely understood. As far as I know, the patents were never challenged. Given how lucrative they were and the number of synth manufacturers that would have loved to have profited, I have the sense that they were legit. Someone would likely have challenged them if there hadn’t been anything to them other than putting on paper what people already understood.
Like you said, a lot of that money benefited students of the arts and added to CCRMA’s cachet.
The patent can be viewed here:
Fascinating, thanks for posting that.