Developer Interview:



Eric Persing of Spectrasonics



In the second in a series of interviews with V.I. developers, one of the most highly respected programmers in our industry discusses his own evolution, the evolution of his company and its very popular products, and the evolution of sampling and synthesis in general



By Nick Batzdorf







What led you to sound design?


I never really pursued being a sound designer, because at the time there wasn't such a thing-you were a synthesizer player. Growing up, I always thought there were two aspects to it. One was that you play, the other was that you create your own sounds. I pursued both and loved both aspects of it.

It wasn't until 1980 or so when I came to Los Angeles and started meeting a lot of my heroes, when I was really surprised that there were very few people who actually knew how to create sounds-how to use the gear. So I found myself with a unique skill that I didn't realize I had.

And then I ended up working for Roland and going to the NAMM show in Chicago in 1984 just a little after MIDI was introduced in 1983. It was a big show-the SBX-80, the introduction of Octopads, the whole idea of MIDI modules, controllers-a lot pretty big things. That was how I got my gig with Roland, kind of introducing MIDI to everyone.

A lot of the Japanese engineers were creating the sounds, and they weren't even really musicians or sound designers. Most of what they'd put in would be sounds that made them laugh! There would be cow sounds and cats, that kind of stuff. They had no idea what people in the United States were interested in. The American synthesizers were doing much better because they had presets that were designed more seriously.

I had created my own sounds for the JX-3P and some of the early MIDI synthesizers...

Those were digitally-controlled analog, right? That was before the D-50 and others with sampled attacks?

...a set of sounds that were a lot more current and useful. That made a big difference. But they were inserting my sounds manually one unit at a time! So the president of Roland US sent me over to Japan-it would be a lot better if we could just get these put in in Japan. So that was how I got connected with the Japanese, and that became my primary role there at Roland: overseeing all the sound development.

That was a very long and fruitful collaboration. Up until just recently I was still a consultant with them. It was a great experience.

We started an R&D division in Culver City in the early '90s-it was an arm of Roland Japan. The idea was that the Japanese engineers would come and study with us, go meet people like Hans Zimmer, and get a little bit more sense of how people were actually using the instruments.

We were also in charge of all the sampling. So we went all over the world and did all kinds of sampling for years. It was one of the biggest sampling efforts of the time. Right around the same time the third-party developer thing was starting to happen, with Optical Media, Invision East West was just getting started, Best Service...so I was also the liason between Roland and these companies.

I was excited about what I was started to see about this idea of sound libraries, the sounds themselves being exciting and not just the hardware.

There a period when there was constantly a keyboard du jour. That couldn't be sustained, and people needed to get more out of the instruments before moving on to the next ones.

Yeah, and I was pushing Roland in that direction in a lot of ways-the idea of making the instruments expandable. You'd buy one significant piece and then create more sounds for it over time and gradually develop it.


So we'd done all this work creating these libraries, a lot more serious effort than the other hardware manufacturers had done. E-mu was really the only that had serious library development.

Not even Kurzweil?


Not really. We were taking sort a VSL approach where you have a very large team of people working on it, doing stuff every day-it was really pretty serious. And it took years to make these things, because back then it took eight hours to save 16 megabytes! Making a library was really tough-you had these primitive editing tools...

So after we did all this work, it was frustrating that we'd create the libraries, and the marketing at Roland didn't really understand what to do with them. It would be in the back of the catalog with the Boss socks, the belt buckles, the string ties...they just didn't understand that kind of product or how exciting the sounds themselves were.

But there were benefits when we would cut them down and do the mini-versions that went in the Sound Canvas, the JV-1080, the [8990?] and things like that. Those were the cut down versions; only the top pros knew about the full ones.

It was neat to see them coming out in the instruments, but it was frustrating at the time to see all that R&D get cut down. They were a different kind of companies-they were a hardware company.

So I had this idea. I had done the Orchestral Family series and the Project series, and I wanted to do a Legends series with well-known players. I had this idea for a Bass Legends, because I was friends with all the bass players.

I still have it!

Good!

But I didn't really feel good about giving it to Roland, because I didn't think they'd promote it, and I didn't want to get my friends involved if it wasn't going to be treated seriously. I was kind of moaning about it to her, and she said, "Well why don't we just do it ourselves?"

That's a weird idea! Maybe we could sell 200 copies and get ourselves a new Honda.

And that was the beginning of Spectrasonics.

So while we were in the process of doing that I was still working at Roland. The guy who was kind of like my protegÇ/trainee at the time, Bob Daspitt, started working with Hans Zimmer and doing a lot of sampling for him. He had this giant collection of guitars but he's not a guitar player. But Bob is, so he asked him to sample them.

Bob had been doing a lot of pretty innovative stuff for the time, like velocity-controlled string bending, 12-way velocity switching on a hardware sampler-that kind of stuff. He showed me what he was doing, and Hans approved and said he could use his name, he was looking for a distributor...so we decided, well why don't we just do it together, we'll split the advertising, and we had that product.

Then I met these guys from Singapore who were working on this crazy Asian library...

Heart of Asia.

Yeah. They had all the greatest performers from China and all over the Asian world coming to Singapore all the time. So they would record them, but they didn't know what to do with it, and they wanted to be part of this too.

It was pretty cool, because we were able to start with three really outstanding products. Our first advertisement was those three products.

It was very humble: my wife and me in our kitchen. But it ended up being a lot bigger deal than I thought it was going to be. Because I didn't set out to be a businessman-that whole aspect. But you have to be to make this work.

Then of course the whole development and evolution of Spectrasonics from starting as a sample library company to being a full-blown software company today.

It seems like we're at the point where the major developers are taking that step. Sonivox has their own player, EastWest has theirs, of course VSL has lots of software, and both those companies have their own studios. You have your own players and STEAM engine...how do you see this going?

I can only answer for us, of course, but I've always been interested in making an instrument. Take a library like Distorted Reality. I'm trying to design an experience for a person so that when they get it, it takes them to a certain place musically, creatively, and emotionally; they think of it as experience the same way you do with a great acoustic instrument.

That part of working with Roland was always very exiting. I got to work with a D50 a year before anybody else did! And I got to help design it, to have some of my ideas in it.

That aspect as always appealed to me-the product design; the user interface, what kind of synthesis, all the components and how they work together, how you make it simple and make it inspiring...how come a Mini Moog is still so great, what makes it tick.

That's one of the things I like about being in this business. I don't think everyone is coming from that point of view; for a lot of people it is about the sounds, and that's fine-I don't think there's anything wrong with that.

But we'd make our libraries and we'd have to make them for all the different sampler platforms. We had Kurzweil, E-mu, Ensoniq, Akai, Roland, Sample Cell...when these platforms would go under because there weren't enough users, it was really a shame after we had spent so much time crafting the experience for that platform.

And then Giga came along. We had another platform. Now that's going away. I got really tired of that-re-making the same product over and over again just so I could keep selling it.

So that's why in 2002 we came up with the idea of the large-scale virtual instrument that's sample-based, it's got an interface...you can use it as a starting point for your creativity. I'm talking about Atmosphere, Trilogy, and the original Stylus.

That was the first stage. And at that point there weren't really instruments like that. There was IK Multimedia Sampletank, Bitheadz Unity-nothing like a dedicated module that had its own interface and its own set of sounds.

I like the idea of being able to make the experience one, and then port that software-wise so that it keeps working on different platforms. It's platform-agnostic-it doesn't matter if you're a Pro Tools guy, a PC guy, a Mac guy, a Performer guy, or whatever-we don't care. You create the product once.

It was frustrating when there would be features on the Kurzweil, but we couldn't take advantage of it, because it only worked on the Kurzweil. It wouldn't work on the Akai, which was the base standard for everybody but it was the least-featured. And it was designed on a Roland sampler!

So it was frustrating to have to cut down the experience. I wanted people to have the same experience I had on mine. In order to do that, the sounds, the GUI and everything all have to go together - at least for what I'm trying to do as far as making more than a sound library, trying to make a real instrument.

Total nirvana in that department must still be illusory, since you still have to chase the operating systems and the capability of the computers. There are rumors that the next Macs are going to have their own processing chips.

But I would still rather have those issues.

Because you still have your own interface that lets the instruments do what you need them to do?

Yes, and that was the second metamorphosis of the company: when we worked with licensing with USB to get the UVI engine, they helped us evolve that engine. That was a really good first step, but we didn't own the technology.

We had a whole bunch of other ideas about where we wanted to go with this, and of course those were just the first-generation instruments. So we found that it made a lot more sense for us not to be a sound company licensing from a software company; again you're tied to somebody else, and they went their own direction and that's fine.

But as a company you have to be able to control your own destiny as much as you can. That's why we started our own software division, and the first product for that was Stylus RMX. It was released in 2004.

I think that really showed people what we could do as far as our ideas and technology and sounds all together. It's been a long process since them, which is the STEAM engine. That's what we've been working on: synthesis, multisamples, high-definition, streaming engine for the future of our playable instruments. SAGE was the first stage of our own development.

It's been an incredibly long process to get to this point. With Omnisphere people can see what we're able to do outside the groove realm, being able to apply our ideas on synthesis and how a synthesizer should be set up, how it combines with the world of sampling. It really is a synthesizer, not just a playback engine. Of course we need that too, but we wanted to go a lot further, we wanted to marry those two worlds.

Going back a little, do you think the days of the guy in the garage developing libraries are numbered, or is it going to require major resources to keep up?

I don't think so. I think that what happened is when Giga came along, it lowered the requirements for what it took to create a sample library. It would cut up things, automatically map things, you didn't have to loop things necessarily. So you could produce a sample library a lot faster.

So then you got this explosion of all these new developers. That was a great thing, and there are still a of things coming along. It all depends on what somebody wants to do. If somebody has a good idea for a library of sounds, producing it in Kontakt format - one of the things that's good about this time period; Kontakt kind of won the war with the samplers, which had been going on since the Fairlight was introduced, and it's kind of over now. It's like, okay, it's going to be pretty hard to top Kontakt. That's a very good thing for library developers, because they can develop in one format.

My only question was brought up by a friend who's developing a library: EastWest PLAY Pro has been announced, but with less competition will they still have a fire under their rear ends?

Oh I think so. If they don't somebody else will come along. We've become used to such an insane pace of development, it might not be such a bad thing to have sample development itself calm down a little bit. The past seven years have been pretty nuts as far as the number of formats and duplication of features...

And I think they are going to have plenty of competition. If somebody sees an opening, there's a good market there.

So I think library development is always going to exist; there's a need for it. And a lot of composers want to work with one interface primarily. At the end of the day it didn't turn out to be Giga, it turned out to be Kontakt.

They did a better job of managing all the different factors, they made it cross-platform. So I think it's going to be very healthy. Certainly the bigger sample developers are going through the same realization we had in 2002.

No question. VSL needed their player for their library.

Right, and I think that's been real successful for them too.

I hope it's been successful financially, but I have to say that creatively it's a success.

Generally speaking that's what's driving all of this development; I don't think the innovation is going to stop anytime soon. At the same time, are we going to see the same radical amount of development since computers started being used seriously for sampling? No. That was a unique time.

But I don't think it'll be like MIDI, where MIDI settled and didn't go anywhere. I think it's going to continue to grow and get a lot of innovation, and I think the concept of a straight-ahead sampler is going to be around for a while. People need that sort of thing.

And because that's there, there's going to be more opportunities for a guy to create a library. He doesn't have to pay any license fees, and he's able to do it on a website.

And that's a great thing for that: there's people who have come out of the woodwork in the last year or two who weren't even there before that. They're doing great work, and that inspires us! There's no resting on your laurels - you've gotta push yourself creatively.

That's the way it should be.

It would be nice if everything was stable and steady and nothing changed - it would be very reliable! - but then you wouldn't have all the benefit of all the exciting new ideas.

Is your engine going to be opened up to third-party developers?

Not initially. Especially when people see what you can do with it, there'll be a lot of interest in people being able to put their own samples in. One of the things we wanted to avoid is giving the impression that it's a sampler. That's not what it's about, it's really a synthesizer. We know how useful that would be and we're certainly not closing the door on that idea, but it won't be part of the initial experience of Omnisphere. It's really about all of what you can do with what we're giving you to work with.

There's a tendency for people to think an instrument is only good if you can add samples or add libraries to it, a content filler kind of idea. What we want to do is encourage people to really learn the synthesis part of it, and we've made it really easy so you can take what we've done and turn it into your own thing with very little knowledge. If you're an expert synthesist you can use it that way too. But that's or focus.

It's sort of sad - people know our instruments are going to be popular and think "we're going to hear those sounds all over the place." That's not what we're trying to do! Of course it's going to come with great sounds, but the idea has always been to push you to the next level of creativity and to get in there and have fun with it. It's a catalyst for creativity.

There's a place for "what's this week's sound library - I'll use it in my production," but I think usually it's more satisfying when you evolve. One of the really positive aspects to the revolution we've had in home recording is that you can be involved in what a microphone does, what a compressor does - it's no longer the engineer's role, and that applies to synthesis too. You can jump in and you don't have to be a brilliant sound designer to create pretty interesting sounds that nobody else has.

We tried to design the interface so that you can you can jump in at whatever level you're at. If you're primarily a presets guy, there's some really clever things you can do if you don't have any synthesis knowledge. And if you have a little synthesis knowledge but you're not 100% with complex modulation or multibreakpoint envelopes, we have presets for all those different things.

And then with the envelopes you can use sliders ADSR-style, and then you can zoom in and you can do anything, you can have 1000 breakpoints if you want. So there are two interfaces. We actually have the same envelope, but you can approach it different ways.

Easy edit.

Yeah, except it's more gradual, the whole interface moves that way. It's like a simple page. You have an overview that's kind of intermediate level, it's very comfortable; it looks similar to something like Atmosphere. There's more to it for sure, but then when you zoom in you can see how much more there is there, you can go deeper and deeper. So as opposed to having a complicated synthesizer and a simple mode, it's more like there's a view that's accessible to everybody and then you go deeper and deeper depending on your skill and knowledge. It's a fun process too.

We did the EZ Edit page on RMX, but we found that people rarely use it - because the rest of the instrument is fairly easy to use anyway. This is a different scale of things because it's really deep. We set it up so you don't have to approach it that way. A lot of synthesizers have presets or you're a heavy-duty synthesizer programmer, and there's not a lot in between. We've tried to make it a seamless thing; at every level you're encouraged to go further and build techniques so that you feel in control of the instrument.

Then you won't have that feeling of "I'll just use the presets," and then you sound like the other guy who used that preset and it's like "I've got to get the next synthesizer of the week." That's never been what synthesis was about in the first place!

I see more to that, at least historically. For example in the 1983 the DX-7 came out. But it could only play one sound a time. Then instruments became multitimbral, and that made a huge difference to what you could do; if you wanted that, you bought the next keyboard du jour.

The DX-7 is a great example because it introduced a whole new approach and so many possibilities, but the interface was terrible so that it encouraged people the other way! It encouraged to just use presets, then get more presets, and then get other units. We've been conditioned that way for a long time!

But now that we have the opportunity to create these better interfaces I'm seeing a lot of people being a lot more creative. That concept we provided in Stylus RMX, it was a very cool thing to see how people were using the loops. I was surprised. People did some very clever things.

It's starting to happen. The tools are so powerful musically, it's less about buying something new and more about knowing what you have.

I look forward to a day when people aren't constantly battling computers!

It's definitely a double-edged sword. You get this fantastic power in this package that's so inexpensive, but the rate of change is so intense.

I don't know whether it will settle down. I hope so! But the quality of the operating systems is getting so much better. I mean Mac OS X Leopard is really a good operating system, so I'm hoping that these things build on all the transitions that have happened.

We never did end up with any standard in the plug-in world. That's not going to change. Audio Units just added another one! And there are new hosts all the time.

That's the other aspect: the support aspect; it's so important when you have a product that it keep working, and when your own software team you can stay up on the changes. Like with Stylus RMX, we were the first software company to have an Intel Mac native plug-in. With Atmosphere and Trilogy - and every VI reader who has them knows that we weren't able to get the licensed engine to be Intel Mac native, so it had to wait until our own in-house technology was ready. The great news is that that's now!

So now Omnisphere, the successor to Atmosphere (it includes everything that Atmosphere has in it, although that's a very small part of what it is)...it took a while to be able to provide that. That was very frustrating to a lot of people, but now that Trillian (our bass module) is out, we can provide the same level of service that RMX users have experienced. We are in very close contact with all the various host manufacturers!

You can only do that when it's your own code base.

The last Omnisphere that nobody's done before is we've created an acoustic Rhodes. I've always loved the sound of a Rhodes, but it doesn't have the richness of sound of an acoustic piano.

The first thing we did was sample all the tines with mics, instead of using the DI sounds; that's an unusual sound, because you can't record a Rhodes that way - you'd have to use 88 microphones!


Then we're feeding that into an acoustic piano. In the piano we're holding the sustain pedal down, so the piano is acting as a resonator. We're miking the piano strings, but it's not being played by the piano, it's being played by the Rhodes. It's on different layers. When you put it all together, you get this really beautiful, sustaining, psychoacoustic Rhodes that doesn't exist in the real world. But it has the organic element that mechanical instruments have, instruments that move sound in the air.

Nobody was expecting a really interesting new Rhodes sound out of Omnisphere. Everyone was expecting it to be the new Atmosphere, which of course it is, but it's much more than that. It's really kind of a history of everything we've done too, because it includes the "best of" all the sample library library stuff as well. All the Symphony of Voices stuff is in there running through the STEAM engine, combining with all the new source...it's pretty fun stuff, creating instruments out of staplers and typewriters...all kinds of different things. We connected rubber bands to an acoustic guitar...it doesn't sound anything like an acoustic guitar, it sounds like a giant rubber thing...the idea is creating new sounds organically.

The thing you're starting with has an acoustic life. That's why we all love samplers: because there's this organic thing, the quality of a recording as opposed to an electronically generated sound. There'll always be a difference there.

So having that organic material and then using granular synthesis, you can apply FM, and polyphonic ring modulation - all kinds of cool real-time thing that the engine does. The sounds have a life to them that makes them very playable, dynamic, and expressive, because they're starting from a unique place. The sound is already rich and interesting.

I personally am the most interested in that. Acoustic instruments have an emotion, but then there are interesting new sounds.

And of course it has a DSP synthesizer too, so it can do all the virtual analog stuff too. What's neat is that it's all in the same environment so you an combine all these elements together - you can go from acoustic to extreme electronic and all those shades in between.

It's an idea that I think has a lot of potential, in fact there are so many more ideas that we want to do with this. We've been working on it for six years already; we started right when Atmosphere was finished. We've been sampling in earnest for three or four years. There's a lot of sound there. I don't know if there's been an instrument released that has as many different sounds. VSL I'm sure wins the prize for number of samples! But as far as each sound being a different idea, there's an awful lot. I haven't made a final count yet, but it's in the thousands!

It puts all those things together: my experience at Roland, the early days of Spectrasonics as a sampling company, what I wished I could do - putting it together into one big package that itself will continue to grow and evolve. I really like that idea about this software synthesizer world.


RETURN TO TOP