You are here

Steve De Furia: Behind The Scenes At Lexicon

Interview | Manufacturer By Paul White
Published June 1995

Paul White discusses the future of digital effects processing with Lexicon's Steve De Furia.

Despite stiff competition from other major manufacturers, Lexicon remain the undisputed champion when it comes to professional digital reverb. Even so, they realised that the audio professional wasn't the only potential customer in the marketplace, and that if they continued to build only high‑end products, they'd miss out on the booming project studio and MI markets. Initially, mid‑range products such as the LXP series brought Lexicon quality within reach of the serious non‑professional and project studio owner, but it wasn't until the Alex hit the shops that you could honestly say the Lexicon for the rest of us had arrived. Two similarly packaged, but functionally different products have also emerged from the Lexicon stable, namely the JamMan delay/sampler/looper and, more recently, the Vortex Audio Morphing effects processor. So, does the non‑reverb nature of these processors signify a new direction for Lexicon? I put these questions and more to Lexicon's Steve De Furia at the recent NAMM show.

Steve has a diverse background in music and high technology. He has taught Electronic Music and Audio Production at Berklee School of Music, he worked as a session player and programmer for a number of years, and as a columnist for the US Keyboard magazine, he wrote a featured series on performance applications and computer programming for musicians. He has also written several books on synthesis, sampling and writing MIDI software.

Product Development

Could you tell us exactly when and how you came to be working for Lexicon full time, and in what capacity?

"I've been a Product Development Manager at Lexicon since 1990. I felt that Lexicon's products were very interesting, and I was quite excited at the prospect of getting involved in product development. There were things not being done that I thought obviously should be done, so it was a nice opportunity for me to apply my experience."

How involved do you get in product development? Do you restrict yourself to proposing concepts, or do you get in there with the engineers and talk about code and hardware design?

"I've written code, but I'm not an engineer. My background is mainly as a hands‑on user, but I've taken courses in DSP and written a lot of music software. My first experience as a software designer was writing a set of composition programs for Frank Zappa. Those programs got me very deep into computer programming. Before that, I'd never really written any software. I know enough about code and hardware to be able to translate marketing concepts and new product ideas into a format that the engineers can work with, and I'm also able to converse with engineers to translate their ideas into marketing concepts."

Where do the ideas for new products come from? Are they individual concepts or the result of meetings and pooled ideas?

"We get our ideas from anywhere we can! There are certain things that make sense to us individually, but we try very hard to listen to what our customers say. We look at what our customers are doing, but we also try to think ahead a bit and extrapolate, to provide them with things that perhaps they haven't thought of yet. JamMan and Vortex fall into that category."

It must be a real problem designing low‑cost MI gear that doesn't compromise the Lexicon name for high‑end products.

"It's a challenge, but we're good at it. Our inexpensive stuff still costs a little more than other inexpensive stuff, but ours has higher standards in terms of both audio performance and mechanical design. You can drive a car over an LXP1, and I've seen a JamMan that was burned in a studio fire, doused with water, thrown out into the street — and still worked! We get agency approval on all our products, CE in Europe and UL and FCC in the United States; not all our competitors are that rigorous about quality and safety. It costs money, but we feel it's important and have a very sharp crew who know how to design for agency approval."

Which of the MI products have you had most input on, and how do they evolve before a production version is finally decided on?

"I did a lot with Vortex and the PCM80, but I've been involved in Alex, JamMan, and more recently Reflex as well. We do a lot of initial spec'ing where we sit down and come up with feature sets based on the input from our users, and the project team. It's a large negotiating process where we balance the price against the features, and then we get to a stage where we're ready to start building. At the point, things aren't actually frozen, but they're pretty fixed. If something happens along the way and we realise there's a feature we should include, or something new comes up, we'll make changes, but we have a very fixed target because that's the only way to bring things to the market on time.

"You have to come out with MI products pretty quickly. A lot depends on whether it's a 'from the ground up' new product with new DSP and mechanical packaging, or if it's a new application for DSP and mechanical packaging that already exists."

With the Alex/Vortex range of products, the packaging is clearly similar, but what goes on inside the box? Is the hardware similar with different software, or is each product quite different?

"The JamMan, Vortex, Alex and Reflex packages are very similar. The Vortex package is different internally from the other three products, but we still tried to keep as many common elements as possible. That's one way of keeping the price down."

Market Perception

I guess that you've had time to analyse the market response to these units, so is there anything that you think can be taken forward and improved upon in the next generation?

"I think our approach to tapped delays, where it's not just a single fixed rhythm, but you get to set divisions of the tap to create flexible rhythmic delays — that concept is moving forward. Also, the looping idea, where you can create a loop, then instantly play it back. The PCM80 'Dual FX' card that we announced here at NAMM has some elements that are very similar to some of the morphing stuff on Vortex, in that you'll be able to re‑route the effects ordering within an algorithm from, for example, serial to parallel, on the fly. A general concept we're carrying forward is performance control. Vortex, JamMan and the PCM80 are not passive effects processors. They are interactive, in that many of their parameters are tied to performance attributes, like dynamics, tap, and a variety of control sources."

A Guitar's World

The Vortex seems well‑suited to both guitar and keyboard processing, but traditionally, hi‑tech processing seems to have been aimed at the keyboard player or home studio owner. Looking round the show, it's obvious that there are at least 10 guitar companies for every keyboard manufacturer, and the guitar shows no sign of going away. Does this mean you'll be looking further into this market in the future?

"I don't know if this is the peak year, but the guitar is absolutely king of the hill at the moment and I love it — it's great to see. A huge number of guitar players are already using our stuff but we're always looking for new things to do. A lot of guitar players are using Vortex, but then it has a lot of effects that process stereo inputs, so you can use in a mix situation too. The guitar market hasn't gone unnoticed."

Another interesting area is the creation of 3D sound illusions from conventional stereo. I know you have some stereo enhancements in the PCM80, so you see this as a worthwhile avenue for further research?

"As you say, the PCM80 has surround capabilities built into all of its algorithms. These can be used to produce a wider reverb sound or focus it anywhere between the centre, sides or surround speakers. The trouble with a lot of the so‑called 3D stuff on the market is that it doesn't image very well over two‑speaker systems unless you're sitting in the absolute 'sweet spot' between the speakers. Even then, without visual cues, some people might believe a sound is behind them, while others hear it as being in front.

"On the PCM80, we use a standard 'left minus right' technique which is continuously variable, and if you use it on a project intended for playback on a Dolby surround system, anything that's processed that way will come out of the surround channel. The fact that you can vary the effect width means that you can take stuff from the centre to the sides, or to the rear, and you can do it dynamically. You can put your echoes or repeats wherever you want as long as the result is played back over a Dolby surround system. One of the patches keeps the reverb focused up front when you're playing, but when you stop playing, the reverb tail sweeps out to the sides and then goes out behind you.

"We're seeing more surround music albums coming out, and lots of people have surround TV. One of our preset developers, Frank Serafine, is working on a television series with a virtual reality theme that's being mixed in surround. They're using the PCM80 constantly to generate the surround effects. Of course, I can't tell you exactly what we're working on now. Let's just say the future is going to be interesting."

A Model Future

Physical modelling is the emerging technology for synthesis, but I can foresee several applications of this in sound shaping. Is this a line that you are following?

"Well, sure. An ironic thing is that Lexicon was one of the first companies to get involved in physical modelling/virtual reality. Those buzz‑words weren't around back then, but Dave Griesinger's reverberation algorithms had to be among the first practical examples of 'virtual reality.' [See Recording Musician August 1992 for a full interview with David Griesinger]. In these algorithms, an existing signal is processed to place it within a simulated acoustic environment. Look at the Roland VG8 — it's an extension of the same idea. A set of existing signals (the six guitar strings) are processed to place them within different simulated environments, and out come different kinds of guitars, played through different kinds of amps and so forth. It's all audio processing. You could conceivably build a processor that could re‑process monophonic input sources, such as the human voice or a sax, to produce completely different sounds — or, for that matter, morph from one to the other in real time.

"One application that has been kicked around for quite a while is formant‑corrected pitch shifting so that you can keep the shifted voice sounding natural."

Using modelling techniques, it doesn't seem too implausible that you could map the characteristics of one singer's voice onto the voice of another. Maybe in a couple of years, we could all choose who we want to sound like?

"Absolutely. You could conceivably have a system where you dialled in parameters like the gender, perhaps the age, of the singer, the amount of whisky drunk over the years, how many packs of Marlboros a day and all that kind of stuff."

What aspect of effects processing most appeals to you at the moment?

"Control related to performance. In the case of Vortex, a lot of what's in there is standard: chorus, delay, modulation, filters and so on. What's different is that the effects are controllable by your performance, so that your style and phrasing completely influences the way they sound. That interaction with the effects is an extremely potent means of expression. I have done a lot of synthesis work, and still do. I've always liked synthesizers as opposed to samplers, because synths generally provide more direct 'handles' to the aspects of sound that are used for expression — pitch, timbre, loudness, etc. As a player, I've developed a vocabulary of techniques that allow me to put myself into the 'here and now' of a performance. Samplers don't offer this kind of sonic interaction. They will create impressive sounds at the push of a button, but basically, you're conjuring up someone else's performance.

"With audio processing now, because of all the power of DSP, you can make expressive changes to what you're doing based on how you play. At the moment, a lot of this is based on playing dynamics, but in the future, a lot of that can relate to other things, such as frequency content."

With Vortex, you took effects that had been around for years and gave them a new lease on life by enabling them to evolve dynamically. Presumably, when effects based on modelling become available, you'll also be able to control those in a similar way, to create evolving sound textures?

"That's true, and it's not just the weird stuff that you never heard before, it's also the stuff that you've heard lots of, that we're letting you hear it in a new way. For example, the Vortex 'Choir' patch isn't just a great chorus, it's a ducking chorus, so that as you play louder, the chorus effect gets less. That means that when you're playing a loud solo passage, the chorus isn't getting in the way — the notes are right out in front. As you play quieter or let a chord ring, the chorus comes up from behind."