When it comes to how keyboard players make music, Dave Smith just may be the most influential person alive. His early experiments building sequencers led to his first company, Sequential Circuits, releasing the first programmable polyphonic slab synth, the Prophet-5. His work on wavetable and vector synthesis in the Prophet-VS was reborn in the enduring Korg Wavestation. His Seer Systems Reality is widely acknowledged as the first commercially available soft synth. And though he tells the story more modestly, it’s perfectly acceptable shorthand to say that he invented MIDI.
At the North Beach, San Francisco offices of today’s Dave Smith Instruments—which makes some of the most desired analog and hybrid synths on the market—we settled in for a long conversation with Dave about where he’s been, where he sees the synth industry going, and what it’s like to finally feel like a rock star.
What was your point of entry into engineering before you were a synth designer?
When I got out of high school it was a time where if you didn’t go to college you went to Vietnam. It was kind of a default to go into engineering. I always had kind of a technical interest. I started at Santa Clara, then I went to UC Santa Barbara for a year, and the last two years were at [UC] Berkeley. I had a day job as an engineer for about six years until I quit. I started Sequential Circuits after the first three years. I realized I didn’t want to work in “regular” engineering the rest of my life, and that’s when I got interested in synthesizers—when I bought my first Minimoog in, I think, 1972.
Your first product was a sequencer, correct?
Yeah. It was an analog sequencer—16 steps by three rows, a bunch of knobs, kind of the style everybody’s building again these days. I wanted a sequencer to go with the Minimoog and the Moog ones were too expensive. So I just built it for my own use. Then I realized, maybe other people might want to have one, too. Sequential sold a grand total of four! That was 1974-ish.
What was the path from there to the Prophet-5?
It wasn’t as direct a path as you might think. The main idea was programmability. I found out that there was going to be a set of integrated circuits from Solid State Music that did synthesizer functions. I’d been working on microprocessors in Silicon Valley for a few years and this was still early in the game, but I knew how they worked. So it was obvious to me that if you put a bunch of these synthesizer chips on a board with a microprocessor, you could build a fully programmable polyphonic synth. For a while I said, “I don’t want to build a keyboard. Moog and ARP are going to do it.” Then in 1977, two things happened. One, I finally quit my day job. Two, I thought maybe nobody is going to build this thing after all, so I’m going to do it. I started in probably spring of ’77 working on what turned out to be the Prophet-5 and had it available to demonstrate at NAMM the following January. I basically did everything—the metal, the wood, the circuit board, the schematics, all the software, just the whole thing.
Today, musicians tend to see you, Tom Oberheim, Alan Pearlman, Roger Linn, and the memory of Bob Moog as one big happy family. Back in the day, though, what was the competition like?
We were lucky that the Prophet-5 had over a year with no competition because it took everybody by surprise. Oberheim was the fastest to respond with the OB-X, in 1979. Basically, after a year and a half . . . everybody was either an Oberheim guy or a Prophet guy. It wasn’t until the early ’80s when the Jupiter-8 and Memorymoog came out that the market got more crowded with polysynths. What’s really cool now is, fast-forward to 2016 and people have the same choice! They can buy a Prophet-6 or an OB-6. It was surprising being at NAMM and seeing how many people just immediately gravitated towards one or the other. It’s like it’s 1979!
You were the principal developer of MIDI. Before that, brands had proprietary interfaces but their gear couldn’t talk to other brands. What was the tipping point?
The synthesizer market was tiny in the late ’70s. No one was selling 50,000 of these things. It wasn’t until the Yamaha DX7 came out that a company shipped 100,000-plus synths. But the Prophet-5 was the first synth with the microprocessor so when everybody started copying it, they used microprocessors. As soon as you have one in an instrument, you realize it’s pretty easy for them to talk to each other digitally. It was around 1981 that we all started realizing the interface needed to be common. It wasn’t like any one person had a bright idea. I remember talking to Tom Oberheim and he had talked to [Roland founder] Kakehashi about it. That was when I gave a talk at the AES convention in New York presenting the USI—Universal Synthesizer Interface. I get tagged as “the person who started MIDI” because I just initiated all of that. I followed up three months later at NAMM. I’m pretty sure it was Roland, Korg, Yamaha, and Kawai who all said “Let’s do it.” That was the start of what became MIDI.
Was there a business case made that someone who owns synth brand X might also buy brand Y if the two can talk to each other?
I don’t remember it being quite that market-driven. Kakehashi was an engineer and a product guy, just like me. The same goes for Bob, Tom, Roger—everything was driven by someone where, it was their company. I think that’s what’s gone missing as things have become more corporate. The work was mostly done by Roland and Sequential. We were the only non-Japanese company so we wrote the spec and handled the English-language side of things. Yamaha, Korg, and Kawai came onboard, and I knew then that it was going to work.
You also came up with the name “MIDI”?
I remember a meeting where Kakehashi visited Sequential in San Jose and Jim Mothersbaugh [of Roland] was with him. They wanted to talk about the name. We all decided USI wasn’t good. So they’d suggested Universal Musical Instrument Interface, pronounced “you-me,” as in “we all connect together.” I thought it sounded a little corny, but I liked the idea of musical instrument instead of just synthesizer. For some reason “Musical Instrument Digital Interface” popped in my head and I said, “How about MIDI?” Everybody said, “Okay.”
The original promise of analog synthesis was to create any sound by getting under its hood, so to speak. Starting in the ’80s, there was the idea that digital was better. Were digital Sequential synths like the Prophet-VS part of that narrative?
The Prophet-VS was kind of in between—hybrid digital oscillators and analog filters with the whole vector synthesis thing. It wasn’t an attempt to “be more digital.” That was what we were doing with samplers like the Prophet 2000. [The VS] was more that a couple of guys came up with the idea of this four-way envelope-modulated vector synthesis idea, controlled by a joystick. We all just said, “Hey, that sounds like fun.” I’ve said this many times before, but to me, one of the reasons the Prophet-5 was so successful is, it was the first time people could actually have an emulative instrument on which they could play brass and strings and organ, some electric piano-like sounds, flutes . . . Of course, we would’ve sold ten times more if it was $1,000 instead of $5,000. One of the main reasons the DX7 did so well is it was $2,000, it had 16 voices, it had velocity, but it was a better emulative instrument. That Rhodes sound was, like, 90 percent of it!
Was that the beginning of analog’s long slumber?
The real death blow was when the Korg M1 came out, which was by far the most popular keyboard ever made. It even outsold the DX7. Finally, here was what keyboard players always wanted—real piano, brass, strings, organs, basses, leads. This is somewhat unfair and I’ll say why, but it put synthesis innovation into a 20-year dark ages, because ever since the M1 every company just kept building M1s. More voices, more and better sounds, more precision—just more, more, more.
In some ways, they’re still doing it. So why was that unfair to say?
Because it’s what 90 percent of keyboard players need to play gigs, which is different from players who are into synths for their own sake. What’s cool and different now is people are once again playing synths as synths because they’ve already got their Nords and Motifs and so forth to cover all the other sounds they need. So if you buy a synth now, it’s because you actually want to play a synth. That’s why I think this time it’s going to be different than last time. There’s not going to be something digital that comes in and makes true synthesizers go away again.
Fast-forward a few years. A lot of people credit Seer Systems Reality as the first soft synth. What was the impetus for your getting into software?
Well, whenever you talk firsts, there’s always somebody earlier. They were in a university, or they built one and nobody knew about it, or they sold 20 of them, whatever. But at Seer we did three generations of soft synths. The first one was for Intel. Intel was just starting the idea of native signal processing actually being able to do real stuff on a computer, and Andy Grove even used it for demos at Comdex. The second generation, we licensed to Creative Labs to put in one of their sound cards—this was mainly sample-based with General MIDI. A million of those things shipped. But what we’ll call the first real professional soft synth was Reality. It had sampling, it had subtractive, it had FM, it had modeling—all of this in one soft synth.
So we called it the future of music synthesis and to a degree, we were correct. But the first time we showed it at NAMM, people would come in and we’d have a keyboard connected to it and they’d go, “What do you mean, software synth?” “We’ve got this PC here. It’s running software that actually is the synthesizer.” “I don’t understand.” It was one of those cases where you’re too early to the marketplace.
What are the main challenges in designing a good analog polysynth at a reasonable price?
Well, the Prophet-6 has over 100 control voltages flying around, so we have high-speed processors generating control voltages at audio rates. We were able to do it at a reasonable price point. It’s a tricky design. We know all of our competitors probably bought one last summer and opened it up. It actually started with the Poly Evolver—the idea of having a high-speed processor per voice to do everything efficiently, and that’s of course what’s in the Korg Minilogue. It has four voices, each one with a fast processor controlling it. This is to be expected once people figure out how to do things.
Where do you think the whole Eurorack craze is going to end up?
I refuse to predict, but I think it’s awesome. The fact that people are getting into synthesizers, getting hooked on buying modules, and hopefully actually using the modules they buy to make cool new sounds. Personally, I like having a “save” button in whatever instrument I’m working with. I’d put a Prophet 12 or Pro 2 up against most modular systems. Obviously it doesn’t give you the advantage of adding some awesome wacko little module somebody built in their basement. I’d like to see more real music made with them, but that’s difficult because any sound you get, you have to use it right then because you won’t get it again! Of course, modulars look great. Having all those wires gives you that synth cred! But just philosophically, I prefer to build complete instruments.
MPE—Multiple Polyphonic Expression—is showing up in more and more controllers like Roger’s LinnStrument, the Roli Seaboards, and Keith McMillen’s stuff. Is this the future of playing the synthesizer, a solution in search of a problem, or something in between?
It’s something in between. Roger just asked me very specifically about MPE. We have customer wish lists for all our products. MPE is something like that. We’ve had a handful of people request it. I told Roger if it gets to the point where a lot of people want it, we’ll consider it. It’s like alternate tunings. People have been bugging us for years about that—usually the same small group. We finally did implement it—I think the Prophet 12 was first. This gets back to our limited engineering time. Of course, people have been interested in alternate controllers for years—go back to the original Moog versus Buchla debates in the ’60s!
The “east coast” versus “west coast” schools of synth design?
Right. [East coast] has a keyboard because that’s what most people want to play. [West coast] didn’t because synthesizers should have a controller that connects to their unique sound possibilities. Well, we know which one won the Darwinism battle. The main thing, though, is chicken-or-egg; getting enough people willing to spend the hours to develop the muscle memory to play your alternate controller, and they’re not going to do that unless they’re pretty much guaranteed that the thing is going to be around. I used to point to the Chapman Stick as an example. Here is this cool bass-guitar hybrid thing that Tony Levin played, but aside from a small group of enthusiasts it’s not really in the musical instrument repertoire now.
What do you think of the prospects for alternate controllers in general?
Well, the Seaboard is better because it’s keyboard-based. It’s kind of like black-and-white keyboard on steroids. Don’t get me wrong—I really want to see things like the Seaboard and the LinnStrument take off. One of the first accessories I ever built for my Minimoog was a light-based controller, in fact. It’s similar to what I’ve always said about the idea of MIDI 2.0: If you could get Yamaha, Korg, Roland, and maybe Nord to do it, game over. Everybody else would have to sign on, just like the first time around. But I also think that as-is, MIDI is good enough to do anything you need.
I’ve seen how your booth at NAMM just mobbed. I’ve seen the adulation when 20-year-olds in an indie band recognize you. I’ve seen this happen to Tom and Roger. In the first heyday of synthesis, it didn’t seem like the synth designer was a personality—except possibly Bob Moog. Now, you guys are treated like rock stars. Is that a new feeling?
A lot of it is about longevity and history. Bob started ten years before we did, and this is going to sound kind of silly but his name was simple to pronounce—even though most people mispronounced it! It became the Kleenex of synths. People looked at any synth and went, “Is that a Moog?” When I came out with the Prophet-5, I didn’t have that history, and I was actually the only one that didn’t use my name. Moog, ARP, Linn, Oberheim—pretty much everybody else did. So of course nobody knew who I was, even the second time. “Dave Smith Instruments, what’s that? The Prophet guy? Oh, cool.” Just like everybody loves old synths, they like the old synth designers, too!
It’s funny you compare us to rock stars, because there’s a big parallel between bands and synth designers. How many bands did something 20 or 30 years ago that was really cool and ever since, they’ve been trying to record new albums? Go to a show and they want to play the new stuff but the audience only wants to hear the old stuff. I get that even now. People come up to me at the booth and only want to talk about the Prophet-5, even as they’re surrounded by new instruments. It’s having to prove later in life that you can still be relevant, just like somebody in a band. Some bands like Radiohead or U2 do this successfully, but many still tour based on what they did 20 or more years ago.
So I think why we get treated like that now is because if we haven’t so much reinvented ourselves, we’ve been able to reproduce what we did before with instruments that people actually like and use and that have modern convenience and stability. It’s a good combination, and we also have the name recognition that comes from history, since the mindset in our industry is very “vintage.” It shouldn’t be, but I’ll take it!