Joe Bryan

Interview by Brian Cowell.

From Ensoniq to KORG to Universal Audio, Hardware Engineer Joe Bryan has become renowned for being a key component in some of the music industry's most legendary products. From his idea of an "analog to digital" input on the Korg Wavestation rack, to his work on the legendary unreleased keyboard seen at NAMM (1993) known as the OASys, Sonikmatter found Joe hard at work producing more "cutting edge" technology.

Located in Santa Cruz in California, Joe Bryan is now Vice President of Engineering at Universal Audio. Sonikmatter caught up with Joe before WINTER NAMM 2002, this is what he had to say:

SONIK : How did you end up becoming employed in the music industry?

JOE : I was crawling around my mom's piano one night while she was playing Moonlight Sonata, and managed to electrocute myself by sticking my soggy fingers into an uncovered power outlet. That pretty much sealed my fate. I think the adult language translation would be something like, "Holy fucking shit! What the fuck was that?!" I'm sure there's some matted tangle of nerve cells nestled deep in my brain that's still ringing from the combined sound.

I had the usual pile-o-wire in my playroom as a kid, and every electrical device in the house was at my mercy. Fortunately for my dad, I could put everything back together. I made my first guitar in shop class, painted it black, and painted a bloody severed skeleton hand with a tarantula crawling over it on the back. It was too cool. Unfortunately, I used a type of marine boat paint from the garage that never dried, so every time you played it you ended up looking like a shoe-shine boy. I think that guitar was sacrificed in a fiery Hendrix-inspired ritual a few years later.

I studied electrical engineering and computer systems in school, and I had a few hardcore engineering jobs before I managed to land a job running the computer system at Ensoniq. My brother Marc was a software engineer there, and he put in a good word for me. In addition to keeping their VAXCluster running, I designed some DSP algos (the Rotary Speaker on the VFX), fixed the clacky EPS keyboard, and designed the SQ-series user-interface.

My former boss at Ensoniq, Alex Limberis, formed Korg's R&D group in California out of the ashes of Sequential. After making a short trip to CA to see the place, I basically said, "Send for my stuff, I'm staying." The next 6 years at Korg R&D formed the main part of my MI career.

SONIK : Who have been your musical influences?

JOE : I heard Wendy Carlos' "Switched-On Bach" when I was 6, and it completely blew my mind (that happens a lot). I couldn't have been more surprised if I'd found a space ship in my morning cereal. I was totally fascinated.

Stevie Wonder filled my craving for more of that electronic synth sound, and introduced me to the idea that you could create an image of a place with melody and tone, and that yo' feets got to boogy with the rhythm!

I had a little transistor radio that could pick up WRKO out of Boston, and I was the proverbial empty vessel. I made "stereo" headphones out of a coat hanger and two little crappy speakers, and "Sunshine of your Love" happened to be playing when I tried them on. Bam! Got to get a guitar and make THAT SOUND. Electric guitar became my primary purpose in life.

Here are some of my influences in random order:

1) Jimmy Page - He uses tone to create new places, he's really a sculptor who uses sound. Great clanging pieces of steel under fingertip control. Awesome!

2) Miles Davis - He introduced the use of space as sonic element, and used it to create new emotions. He also had balls bigger than Jupiter's moons. One of the true pioneers of music.

3) Allan Holdsworth - He took a Strat and a Marshal stack and made them sound like a violin and a saxophone. No one else has ever managed to fundamentally change the very physics of the guitar's envelope, and Holdsworth added an embouchure.

4) Brian Eno - Space, emotion and environment. He created ambient music, and for that we can't forgive him! His work with Robert Frip was especially influential.

5) Bill Evans - There's a trend here. Space and emotion have never been so beautifully enunciated with such fluidity and grace. His compositions formed in front of you, out of the space around you, and all in real time with no overdubs.

6) David Gilmour - Fat fingers and fat tone. Not too many notes, just the right ones.

7) Jon Anderson - And the rest of Yes created new worlds in acoustic space. Their live shows were transcendent. One of my first live concerts was trip to Boston Garden. Listening to Relayer performed live was a very liquid experience. Whoa!

8) Frank Zappa - Another true pioneer, visionary band leader, and personal hero. He told the idiots of Tipper Gore's PMRC to go to hell, and got an "explicit lyrics" warning sticker on an all instrumental album to show for it! Talk about performance art.

9) Joni Mitchel - She's another great musical sculptor. Her voicings are simultaneously original, beautiful, and highly personal.

10) Billy Corgan (Smashing Pumkins) - One of the most interesting of the recent crop of artists. He also wears his Jimmy Page influence on his sleeve. Check out the extended ending of "Drown", it's a feedback solo worthy of Jimmy Hendrix and Robert Frip.

11) John Lydon (aka Jonny Rotten) - After a certain incredibly obscene and hysterically funny act at a P.I.L. concert I can't possibly describe without offending everyone, he will remain rock-and-roll incarnate for me.

12) n'Sync - Ha ha, just kidding!

SONIK : What kind of music do you like to play?

JOE : When I get a rare change to play these days, I like to play eclectic improvisational music. I've playing in traditional rock and blues bands, but there are a few guys I've played with for years who simply don't fit into any categories. We like to stretch out when we play. One guy has a Chapman Stick with a MIDI pickup that he likes to play with a slide. Let me know when you come up with a name for that.

SONIK : Can you tell us about the "OASys" that was originally previewed in 1993 at NAMM?

JOE : That was quite a beast. Unlike the OASys PCI card released several years later, the original OASys was an actual standalone keyboard, and the story is quite involved.

The OAsys project was a continuation of the work Steve O'Connell had started with his SynthKit program back in 1989. It was very powerful, but it was Mac software based, and users needed some DSP theory to use it. It was primarily a research tool. Even though Physical Modeling synthesis was the main topic in the research community, no one had access to real-time rendering systems, and some simulations would take hours or days to generate their sounds. Because of this inability to interact with them, most of the algorithms didn't sound very good. SynthKit changed that by making algorithm development an interactive, real-time process, and one that advanced voicing guys could be trained to work with.

Andy Leary, Charlie Bright, John Bowen, Ben Dowling, and a few others spent most of their time developing algorithms with SynthKit. The interesting thing about it was they threw out nearly 99% of their work. That was what made their results so good: they could audition and refine their algos quickly.

As good as SynthKit was, it wasn't practical as a musical instrument. We needed a system that could play the algos polyphonically and in real-time, just like any other synth. The challenges of this simple concept essentially meant we had to develop a full-blown general-purpose computer system from scratch. The founder of Korg, Katoh-san Sr., was impressed with our Wavestation work, so he basically gave us carte blanc with the OAsys.

My responsibilities included designing the hardware and software system architecture, and developing the technology that let the algorithms be ported from SynthKit and loaded and played in real-time. I expanded the real-time dynamic resource allocation and modulation techniques I'd started using on the Wavestation, and in the process, developed file systems, user-interface design tools, modular cross platform techniques, and object-based compilers. I wrote a paper describing it for the AES, which our VP Alex Limberis presented, and we got a few patents in the process.

Unfortunately, the project took on a life of its own, as each group tried to out-do the other with features. The user interface started out relatively simple, but ballooned into the mothership of midi controllers. We out-did ourselves making it flexible, and because of the very long rope we had, in the end we succeeded in making mostly a big noose out of it. Ultimately, we realized the custom component costs Japan was paying wouldn't justify the required sales price, and what started as a $3000 synth became a $15000 boat anchor, So Katoh-san was reluctantly forced to cancel his pet project.

Despite the eventual realities that doomed the original, Korg eventually came out with OAsys/SynthKit based products like the WaveDrum, the Prophecy and Z-1, and finally the OAsys PCI. The closest thing to the original OAsys keyboard is the Trinity. It uses the same case David Goldstein designed for the OAsys, has the same touch-sensitive controllers and touch screen with the bright-blue LCD, and even some of the same signal routing structures developed for the OAsys keyboard. Korg's best stuff was developed as a result of Korg Japan and Korg R&D engineers listening to the MPB team.

The first time anyone heard the OAsys was at NAMM in a private suite in the Hilton Hotel (they showed it again the next year on the floor). I remember holding the door for Keith Emerson and Bob Moog on their way in to hear it. One of the coolest moments was observing a jaded LA session pro hear the OAsys for the first time. I overheard him telling someone how tired he was of all the look-alike, sound-alike synths out there, and how it was starting to become a chore checking everything out. After listening to the OAsys, he started jumping up and down like a kid saying, "Let me play it, let me play it!" It was great to see him regain his excitement for something that had become a chore.

I think keeping that flame alive is the ultimate goal for all of us in this industry.

SONIK : Do you think that an OASys type of keyboard will be possible in the future sometime?

JOE : An OASys style keyboard is still a dream, although it'll take a serious commitment to pull it off correctly. I wouldn't be surprised if Korg was busy working on one right now. If they limited the scope of the project to just playing killer sounds instead of being the end-all be-all synth/MIDI controller it started out as, the OASys could be a reality today.

Unfortunately for the rest of the MI community, there's a barrier to wide spread adoption of physical modeling (PM) synths. Anyone who wants to develop one has to pay Yamaha/Stanford for the Sondius license and co-brand their work with it. If you haven't read the Sondius agreement, it's pretty interesting. Suffice to say no one's rushing out to develop PM synths anymore, and those that have, have all been forced to pay the fee. Even Korg was forced to pay, and I'd say they're clearly the group with the best sounding algos.

Despite all that, there are workarounds. The larger problem is the processor requirements for algorithmic synthesis are quite high, and currently there's little need for anything that powerful in the consumer space to make them affordable (yet). PM chips will either have to use custom DSPs, or off-the-shelf DSPs, both of which are expensive and limited. A PM synth would either be limited in its algorithms, or limited in its polyphony. Unless of course you've found something useful that's popped out of a crashed UFO or something.

Right now, the algorithmic synthesis field is at a similar point as the DSP effects field: retro emulations. There's a lot to be learned from the old gear because it sounds great. The real-world shmutz that gets into the process provides the secrets. Two of my favorite examples are the type of spruce Stradivarius used in his violins, and UA's LA-2A's El-Op device. Stradivarius used spruce that was unlike any other because of a bacterium that lived in the water where the logs were transported. He just used what was available, and it just happened to be extraordinary. The classic LA-2A compressor uses a simple nightlight and a type of photo resistor you can get at Radio Shack, but the combination is truly magical. It took a huge effort by our DSP team, including consultations with quantum device physicists, to figure out exactly how it works.

Simple stuff produces great results, but the reasons are profoundly complex. This same magic will continue to happen in the design of future of musical instruments. It's inevitable.

SONIK : What happened to the OASys prototypes?

JOE : As far as I know they're still at Korg R&D in Milpitas. There was one prototype encased in a blue version of what's now the Triton case, and there were a few others encased in steel boxes. The blue one was called the Blue Bomber, and the others were called the Perestroika Boxes. There were one or two others mounted on boards that were used in the lab for development. I suspect they're all gathering dust these days.

SONIK : How did you come up with the idea of the WAVESTATION A/D?

JOE : There's always a lull in activity after a product goes out and we scratch our heads asking, "What's next?" I had suggested we put analog inputs on a synth for a few months, but no one was very interested. My motivations were purely selfish. I'd just developed what I thought was a cool effects processor for the Wavestation, and I really just wanted to play my guitar though it.

I didn't want to spend a lot of time developing a converter circuit for a quick proof of concept piece, so I started rummaging around the lab for some existing converters. The lab at Korg R&D is pretty interesting. Besides an old Mac Classic used as a door stop, there's lots of old Sequential stuff piled up (things like rare 10-voice single rank Prophet 5's, protos, etc.). I found a broken Prophet 2000 sampler board with some ADCs on it, and noticed something useful about its board layout. So one summer day while everyone was sort of kicking back, I started work on the first prototype Wavestation A/D.

One minute there was relative quiet, and the next there was this loud screeching fingernails-on-a-blackboard noise out in the lab, and a few guys came in to see what was up. It must have looked pretty odd seeing me hacksawing away at a circuit board clamped in a bench vise, because there were the usual "What could you possibly be doing?" questions. It so happened that the A/D circuit was neatly arranged on the board so I could literally cut it out with a saw. Pretty soon we had the converters hooked into the voice engine, and then everyone started getting interested in putting them on the rack module. In the end we didn't use the same Sequential A/D circuit, but I thought it was a nice homage to the Sequential legacy that it influenced the Wavestation in several different ways.

SONIK : Were there any features that were left off the Wavestation series that you would like to have implemented?

JOE : Resonant filters. There wasn't a choice though, the voice chips available from Korg just didn't have them. I added some sweepable resonant filters to the effects (Distortion, Overdrive and Parametric EQ) that are usable for mono voices. You can select envelope, key number or any of the other controllers as the filter cutoff mod source.

I think the other big thing people miss is downloadable samples, but adding the RAM for that would have made it too expensive.

SONIK : Did you use any more of the technology from the SEQUENTIAL gear (while you were at KORG)?

JOE : We didn't look at specific designs or any design info, but we did use some of the concepts, since two ex-Sequential guys, John Bowen (voicing/concepts) and Scott Peterson (software development), were heavily involved at Korg R&D. Also, Dave Smith was involved in the concept development for the Wavestation.

SONIK : Can you tell us what you do at Universal Audio?

JOE : I shuffle about picking up scraps of paper saying, "That's not it." One of these days I'll find the right scrap of paper.

UA bought out a company I formed in 1998 called Hyperactive Audio Systems, and I became UA's VP of Engineering. Hyperactive developed the DSP technology used in the UAD-1. That effort mainly involved scouring the desert near Roswell, NM for parts from downed UFOs, and negotiating with the Chinese military for their schematics.

I mainly write lots of code, and manage the development and technical support groups. I'll occasionally get to do a bit of hardware engineering for grins. UA's pretty small, so everyone has a lot of hats. We make both classic analog gear and modern DSP gear, so it's really interesting. There are some spectacularly talented people in our development teams, and everyone gets to learn from each other. It's great fun.

My primary goal is to create the right environment structure so the creative process is streamlined. There's no better waste of time than reinventing the wheel or getting bogged in the mundane.

SONIK : What is the goal of Universal Audio?

JOE : UA has a unique reputation in the audio industry. You can read all about it on-line at: http://www.uaudio.com/history/index.html.

Because of that legacy, we have a duty to meet the highest standards for everything we produce, and maintaining quality is the key. We try to make the best possible gear in a way that makes it affordable, and since most of us are musicians, we also try to make everything easy to use. We really want as many people as possible to enjoy making great music.

Besides the LA-2A and 1176LN reissues of the classic compressors, we're especially fond of our new 2-610 and M610 tube mic-pre's. We also just introduced a new transistor-based design called the 2108 that will fill out our analog products line.

With UA's purchase of Kind of Loud Technologies (DSP algos) and Hyperactive Audio Systems (DSP hardware), we also have a killer digital products team, and we're launching a bunch of new products based on those technologies. We just signed a deal with Mackie Designs for distributing the UAD-1, and that will help us reach more of our customers than ever before.

It's really a great time to be a part of the new audio technology revolution. As advanced processing becomes cheaper, better, easier to use, and more reliable, more and more people are able to produce high quality music than ever before.

SONIK : Can you tell us about the group you founded at Universal Audio?

JOE : The Wavestation was a big creative success for Korg, but it wasn't a big commercial success. It was a huge risk to make a keyboard that didn't have drums or piano at the time it came out, and sales suffered somewhat because of it. Back then, everyone wanted canned band-in-a-box stuff, so we recanted and added some drums and a piano for the EX version. Now, everyone's into electronica and trance, so there's more interest in new sounds. We were lucky Katoh-san, Sr. was interested in taking that risk. He's a rare breed.

It's very difficult to continue doing what we do. Sometimes it involves waiting for the market to come around to a new idea (like electronica), or woodshedding something that advances the state of the art (like the OASys and UAD-1). It takes a lot of determination to forge through the personal, professional, economic and technical obstacles to make it all worthwhile. Failure and risk are a very real part of the process.

I came up with the name "Stubborn Fuckers" for the group at UA that's persevered in the face of incredible challenges and risks to bring the UAD-1 out, but its membership is open to anyone who's persistent enough to try and make a vision become reality. Older guys like Tom Oberheim, Dave Smith, Roger Linn, and Bob Moog completely define this approach, so even though I might have coined the term, I can't possibly call myself a founder.

SONIK : Do any design "accidents" become features in you products?

JOE : Engineering is the process of paying attention to extremely minute details with excruciating deliberation in order to realize a creative goal, it doesn't make a very good game of chance. Most accidents are either bugs or anomalies at the very least, but there are the occasional serendipitous mistakes. One that comes to mind is the FM-like ring mod effect caused by reversing the slow and fast rotor speeds in the Wavestation's Rotary Speaker effect. The QC guys thought it was weird, but the MPB guys said, "No, it's interesting, it has to stay."

SONIK : What is your view of the "software versus hardware" debate?

JOE : I've never taken sides. I don't see the distinction really, since every piece of hardware has some form of software in it, and vice versa.

At all the MI companies I've worked, we made synths and effects processors known in engineering jargon as "embedded systems". Structurally, they're no different from a desktop PC: they both have power supplies, processors, storage devices, and human interface devices. They're computers with LCDs instead of CRTs, piano keys instead of typewriter keys, and buttons instead of a mouse. That they have connectors that output voltages that are converted into sound instead of characters on a printed page is a happy coincidence.

The real question is, "What's works for you personally?" Embedded systems are nice because they're reliable and built for a specific purpose. They're made from specific matched components that aren't usually upgradeable. That's also the problem with them. The components are expensive and obsolete before the product ships, their interfaces and operation are inconsistent with other devices, and they aren't very flexible. Computers are inexpensive and flexible, but they're too complicated and aren't reliable.

The challenge for any engineer is to optimize the system design so that the device meets its requirements as inexpensively as possible without sacrificing reliability or usability. Specifically, we try to provide maximum quality for minimum cost. For me, it's an economic problem, not an engineering one.

To paraphrase Frank Zappa, the appliance is your friend. We're already seeing more and more designs that blur the distinction between "software" and "hardware". The MI device of the future is a computing appliance, not unlike a video game player. It has the advantages of both hardware and software: purpose-built, reliable, flexible.

The bottom line is this: Whatever it is, you have to be able to use it when you're fucked up. Anything else gets in the way of the creative process. As Jack Hotop once said, "EQ, IQ, whatever it takes."

SONIK : What would be the funniest thing you have seen happen in the music industry?

JOE : There are way too many stories, and most of them would either make you blush, or get someone indicted. So it's hard to pick a funniest one.

One of the more unusually experiences I had was riding the elevator at NAMM with an Alice Cooper wannabe, the two barely dressed spandex-clad babes draping him, and a pair of horrified yet crisply pressed older ladies. They happened to be attending the National Evangelical Minister's Association convention that overlapped NAMM that year. I seriously doubt they made that same mistake twice, since there's never been another booking mishap quite like that at NAMM. Most folks have the good sense to get the hell out of town.

The juxtaposition of straight life and music industry life is always amusing to me, and few people augment that division better than the characters in Korg's MPB. They're like a barrel of monkeys on acid when they get together.

Since we spend a lot of time together during the intense phases of product development, we get somewhat immune to their antics, sort of like building up mental calluses.

One typical day at Korg R&D, a few of us software engineers were having a serious technical meeting with an outside vendor in the conference room. All of a sudden, there's this increasingly loud pounding and impassioned moaning and yelling coming from the sound room down the hall. Ben Dowling was doing yet another amusing impersonation, and this time it was his version of a slightly older Heidi enjoying a good shag. He's pounding away on a table with his hand, yelling in a German falsetto, "ya... Ya!... Ye-Ah!... Ye-AH!... Yee-aaaaaahhh!!" as Heidi reaches Nirvana with a loud yodel. We didn't even blink, and it didn't occur to us that anything particularly unusual was happening until one of the guys in suits asked us very concerned, "Exactly what the hell is going on back there?" We just managed to say "Oh, it's just Ben," before breaking up laughing. Like that was a valid explanation.

SONIK : Looking back over all the products you have done; which one stands out the most?

JOE : That's hard to answer. There's the smarmy, "I like all of my children!" answer, but in reality, little Jimmy's really a brat.

The one that had the most influence on me personally was the original OASys, but since that product was never actually released, it doesn't really count. Still, that project transformed everyone involved with it.

SONIK : Do you ever put forward names for products?

JOE : There's product names, and there's project names. Product names are often conservative compromises arrived at in committee meetings, but project names are usually fun. Sometime though, individuals will propose product names that stick. I usually get overruled, but I managed to get Nigel to be the name for UA's new guitar processor effects suite. Dan Phillips came up with the name OASys as an extension of a tradition started at Ensoniq. First there was the Mirage, then there was the OASys. Dave Smith called his second company Seer Systems, and you'll notice Seer follows Prophet under "Oracle" in Roget's Thesaurus. Let's hope his next product isn't called "Soothsayer".

Ensoniq had related names for everything. For instance, all their computers were named after dead musicians (mostly rock stars), and Korg R&D followed that tradition. The main server was Elvis (of course), and we even had a Keith Richards, since he's really an honorary dead rock star considering all the blood transfusions he's had.

SONIK : What forms of synthesis do you think could be explored more?

JOE : Like I was saying earlier, algorithmic synthesis is really where we're all headed. It sounds like a generic term, since all DSP based synthesis is algorithmic, but I use it to mean the integrated real-time control and processing model I described above, not just DSP versions of fixed algos. The OASys was one of the first real-time performance algorithmic synthesizers, but it's certainly not the last. There's a class of similar algorithms I loosely call Extended Karplus-Strong that combines filters in and around delay loops with commuted synthesis, that has a relatively rigid signal processing architecture that's well suited to efficient implementations. I'd like to see this explored more fully, because the sounds this class of algorithms is capable of making are very broad and self-interactive. Every plucked or hammered electric or acoustic string and percussion instrument falls in or near this class, so it provides a large bang for the processing buck.

I was working on the real-time modulation code for the OASys many years ago, and I was using a simple plucked string model someone had made to test the pitch controller and mod wheel. The keyboard was setup to retrigger existing voices, which adds signals into existing resonating strings, while the pitch controller was setup to give a dive-bomb whammy effect. As long as you held the sustain pedal, it essentially had "live" strings. One of the completely unexpected side effects was the mechanical feedback caused by the controller handle hitting the stop in the case. When you bottomed-out the pitch quickly, it would introduce a type of discontinuity into the signal that re-excited the existing string resonances. It was just like whacking a real whammy bar. It totally blew me away that an all-digital synth could get the same sound. Mechanical feedback? I thought, no way, but there it was.

We joked we should add a temperature sensor, a barometer, and vibration sensor to the OASys for more random influences. Those were included in an email I sent to some friends at Ensoniq that listed the features of the new secret project we were working on. I also said it had a laser for a head's-up display that would create holograms above the instrument for interacting with it. I also think I suggested we should target the Dub music community because they were more likely to have the necessary "particulates suspended in air" to make the laser more visible. Ha ha (cough)...

SONIK : Do you think that the "workstation" concept is dying?

JOE : If you define a workstation as a synth with effects and a sequencer, then I must admit I've never been a big fan of them. Most people use a computer for their sequencing, so why complicate the synth's UI and compromise its software reliability with all that extra stuff? I'd rather have a synth that's dedicated to making sounds. I realize there are plenty of people who need all-in-one synths, but I'm not one of them.

On the other hand, I think the trend towards dedicated music computers is great. The best of both worlds. Record all your tracks, mix and process them in one consistent environment with all your synthesized and sequenced material, done using the best components for the task. I like to blur the line between recorded, processed and synthesized music, and I want to be able to do everything easily in my intuitive creative mode, not my rational analytical mode. I don't care if it's a computer that gives me that experience or a dedicated "workstation". I think it's a combination of the two.

SONIK : If you were developing a synthesizer now Joe; What features would you insist be on it?

JOE : I still want to make a simple synth that provides the dynamic real-time interactive tones you get from algorithmic synthesis. The Korg team has done an incredible job of creating a lot of emulations of real instruments, but I'm more interested in new sounds. If I need a trumpet, I'd rather hear a real trumpet player than an approximation from a keyboard. Besides, the better the algorithm, the harder it is to play, and you end up almost needing to be a trumpet player to articulate it correctly.

I like integrating the effects into the actual voice. I've never made a big distinction between signal generators and signal processors, because even though one of them doesn't have audio inputs, they both take control input from the player and contribute to the final audio result. I also don't make much of a distinction between control processors and signal processors for the same reason. Most of the important advances these days are in the control processing side of things, but they usually require advances in audio processing to make them work correctly (like needing more DSP horsepower for smoothing the parameters).

I created the concept of voice effects and environmental effects in the original OASys to describe effects that were part of the voice itself, and those that were part of the environment surrounding the voice. Voice effects would duplicate themselves when you added voices, while environmental effects would merge. Track inserts and sends are the same concept in a mixing situation, so there's an aspect of self-similarity in the hierarchy of music production. Once you're familiar with that dimensional relationship, scaling it up or down should be easy.

To a guitarist, the pickups, amp, speaker, mic, etc. are the instrument for the tonal dimension of the sound, while each string is an instrument within the melodic dimension of the sound. To some guitarists, like Allan Holdsworth, there are many dimensions to his instrument. If you've ever seen him perform the tune Tokyo Dream live, you know what I mean. It's on an album called Road Games, and features a deep multilayered arrangement of what sounds like strings, breathy pads, bells, guitars, and other sounds. He plays everything at once on one guitar using hammer-ons, left- and right-handed tapping, harmonics, and other techniques, and just uses an analog ADA multitap delay processor. It's a complete trip.

Another thing I like is meta-controllers. Agents that process other parameters to create new dimensions of control. For example, UA's Nigel processor suite has a killer mod filter module that has a random LFO setting. Pure random is good for that old ARP-2600 sound, but we added a chromatic pitch quantizer and range controls to refine the effect and make it more melodic. For example, you can create a complex melodic counterpoint effect by adjusting one control in realtime. I'd like to extend that to include a transform sequencer for the filter cutoff, but it can already do that now since everything's MIDI controllable.

I think Korg's Karma is an interesting idea. Steven Kay's been working on that for years, so it's cool to see his Sorcerer technology realized in a real synth. He gave me a demo many years ago, and I was impressed.

SONIK : What do you think of the different products that the European keyboard manufacturers are doing, as against what the Asian/American countries are doing?

JOE : I don't really follow MI keyboards anymore, since I'm mostly involved in pro audio processors, but I did notice the emergence of the European synths. Nord has done a great job of reintroducing knobs, while the Italian MARS/ISIS group has done some impressive things with physical modeling. The Creamware cards are OK, and I understand they have quite a following in the techno music community. However, I think John Bowen's Zarg Music synths are the best thing to happen to them.

The Asian companies are incredibly good at manufacturing reliable products, and very supportive of innovation. Many of them have formed teams of American and European designers to keep their designs fresh, and to provide the necessary broad perspectives in the design process. I think the northern Europeans are less interested in that approach, and sometimes their products reflect a certain arrogance of viewpoint. That can be great when the viewpoint is innovative and fresh, or it can be bad when it's overly restrictive, so it depends on the product and the customer who buys it.

SONIK : What sorts of "effects" do you think need to be explored more for guitarist/keyboardist?

JOE : I think the two most useful effects would be an anti-suck filter and a free drink coupon copier. You'd need both, since the more free drinks you got, the more intense the suck filter would have to be set.

I've always liked multi-dimensional sounds where unusual elements are used to carry the melody and rhythm. One example is sending drums through a resonant filter and modulating the filter cutoff with a melodic note sequence. There are many dimensions within sounds that can be used like this.

UA's new Nigel effects suite for the UAD-1 has a lot of effects that provide this type of multi-dimensional control. One simple effect in Nigel is called Shimmer, which is basically tremolo with the intensity and speed controlled by an envelope triggered by the signal. You can control the delay time of a mod-delay with the same envelope, which lets you modulate the tone and pitch along with the amplitude. We also included the ability to morph one amp into another, not by crossfading, but by actually transforming them at the component level. The sound is simply stunning and must be heard to be believed. That's just a few of the many things we're doing in Nigel, which has 9 effects integrated into one ultra-high performance module. Right now, we can't even run the whole suite native on our fastest dual CPU systems, but we can run several instances on the UAD-1. It's opening up new worlds for us.

SONIK : What DSP processors do you think look promising?

JOE : I think the Z80 still has some kick in it. I also like the old PDP-8's for their sheer warmth. I once convinced a Motorola salesman at a trade show that the big-endian byte order used in the 68K sounded better than the little-endian byte order used by Intel processors. He seemed relieved to learn one more thing to use in his sales pitch.

The key to UA's UAD-1 DSP card is a processor I discovered several years before we even started the UAD-1 project. No one else saw its potential, or they dismissed it outright as impossible. We've got more gems like the UAD-1, because we view the technology development process as a core technology in itself, and we've got contacts in Silicon Valley and the research community that keep us informed of the latest opportunities.

I'm sorry I'm being evasive, but if I told you what we were planning to use next, no one would benefit. The advantage we've gained with the UAD-1 provides us the leverage to take the next step, and to undermine that would prohibit us from having the freedom to develop the algorithms we're developing without spending all our energy competing in a commodity market. Commodity markets require high volumes, and the MI and pro-audio industries aren't big enough to support that.

I think the results of our strategy speak for themselves. The plug-ins on the UAD-1 are considered by many seasoned professionals to be the best software plug-ins available on any platform, regardless of price. We plan to continue that approach because it produces the highest quality results for the lowest possible cost.

SONIK : Have you ever considered writing a book on DSP technology, Joe?

JOE : No, I haven't even thought about it. I think my brain might explode if I tried it, and I just repainted my home office, so I wouldn't want to clean up the mess.

SONIK : What would you recommend for young Engineers becoming involved in the music industry?

JOE : Number one, learn the basics: Math, physics, analog and digital electronics, differential equations and linear algebra, software structures and operating system fundamentals. The details of specific technologies you learn become obsolete very quickly, so learning how to learn new things rapidly on your own is essential.

Number two, learn how to listen to other people. No one has a monopoly on the secret sauce. Sometimes the best ideas come when you're least expecting them, and your ears need to be open.

Number three, learn how to communicate your ideas. This is more important than anything else. If you can't explain why an idea is worthwhile, it's unlikely anyone will just take your word for it.

Number four: Play an instrument and join a band. Teamwork and waking up in your car with cold cheap greasy pizza stuck to your date's butt are valuable learning experiences.

As for what to expect, don't expect to make a pile of money in audio. If you want to be rich, sell MP3 players. One guy I interviewed was fresh out of school, with no experience and limited skills, but he demanded a $100K salary and a signing bonus. We politely declined him a job offer.

SONIK : What instruments do you have in your own personal studio at home Joe?

JOE : You have to ask. About a year ago, my wife and I moved into a new house and had a baby, so I doubt I'll be setting up a home studio for a while.

My office at work is another story. In addition to all the computers and digital I/Os I use for development and testing our UAD-1 system with various DAWs (Nuendo, Logic, Cubase), I have a homemade basswood body guitar running into a pile of homemade guitar effects, and a Korg Z3 MIDI pickup that feeds a Wavestation A/D and a Wavestation SR. The guitar goes through the A/D (which is why I made it in the first place), and everything is fed into a small homemade mixer which feeds a stereo bi-amped guitar/bass rig. It's got way too many wires, but it's great for the out-there guitar sounds.

I don't try to get the synths to track the guitar too closely, because it's asking too much for the MIDI pickup, so I'll setup pads with a slow attack, and latch them while I play guitar on top. I like slide guitar, and I've got a connecting rod pin out of an old Triumph that makes a perfect slide. Sometimes I'll use it to get trippy middle eastern Kaval and Zurna sounds.

Most recently, I've been just playing the guitar straight through a UA 2-610 (the ultimate guitar DI), and into a DAW running the UAD-1. We have a plug-in suite called Nigel that basically does everything I've ever wanted for guitar. If there's something missing, we'll add it to the plug-ins.

There are several great guitar players at UA, so it gets pretty loud sometimes with the dog barking and the voicing guys doing their thing!

SONIK : What do you do outside of work to keep your mind fresh?

JOE : Fresh? I usually keep it in the fridge in the crisper drawer with the rest of the veggies. Sometimes I'll pat it down with some salt and baste it with a beer glaze to give it some flavor.

Actually, I'm lucky enough to live close to UA's building, which is located a few blocks from the ocean in Santa Cruz. So I usually ride my bike to work, and take a route that goes along the coast for a few miles. Depending on the season, there's plenty of birds, seals and otters; and in winter, dolphins and the occasional whale.

I used to spend clear nights observing the stars and planets with my telescope, but I've been too busy recently for that.

March 2002