FIG. 1: SuperNatural sounds continuously morph in response to changing Velocity or continuous Control Change messages. This provides greater expressive potential than simple sample-based programs, which rely on crossfading between only a few samples that cannot change their basic sonic character once they start playing.
Credit: Chuck Dahmer
When I worked as a product specialist at Roland (www.rolandus.com) more than 20 years ago, I saw several major new synthesizers arrive at the office before they were introduced to the world at various NAMM shows. Among them was a rackmount module called the MKS-20, which was dedicated strictly to preset piano sounds.
What is the big deal about a piano module, you ask? Well, this was not just another sample-based sound generator—it was the first to use Structured Adaptive (SA) synthesis. SA was an innovative combination of additive and sample-based synthesis, in which the sound and behavior of specific pianos was painstakingly analyzed and resynthesized using complex waveforms as additive elements rather than simple sine waves. The result was a level of expressive capability impossible to achieve using sampled sounds alone.
In the years that followed, Roland introduced several digital pianos that used SA synthesis. But SA soon disappeared from the product lineup, primarily because creating new sounds was very labor intensive, and musical microprocessors weren''t powerful enough to do much more with the technology.
Fast-forward 20 years. At the 2008 Winter NAMM show, I attended Roland''s press conference and heard company reps announce a new type of synthesis called SuperNatural, which will be featured in the RD-700GX digital piano and in some expansion boards for the Fantom G keyboard workstation. The reps were talking about enhanced expression, behavior modeling, and continuous timbre changes rather than crossfading on sounds ranging from violins to trombones to drums. SuperNatural sounded just like SA, but it applied to a much wider variety of instrumental sounds!
After the press conference, I talked with Toshio Yamabata, the director of Roland R&D; Shun Takai, an engineer who has been working on the technology directly; and Mike Kent, manager of technical relations for Roland R&D. They confirmed that SuperNatural, which the company has been quietly developing over the past two decades, is indeed an outgrowth of SA.
As was the case with SA, the first step in creating a SuperNatural program is separating the various elements of the target sound—for example, the string and bow of a violin as well as the frequency, time, and phase components. These elements are then utilized to reconstruct the sound using an additive technique with complex waveforms from the original sound, along with modeled components and wave-table synthesis.
Yamabata, Takai, and Kent explained that Super-Natural programs can smoothly morph from one sound to another in response to changing continuous Control Change or Velocity messages (see Fig. 1). By contrast, conventional samples must be crossfaded, and once a sample starts playing, it can''t be changed without sounding completely artificial.
Another advantage of SuperNatural is its ability to control various “acoustic” characteristics of the sound, much like physical modeling. For instance, you can change the shape of a snare drum body and the corrosion of the tines in an electric piano. Many programs also include modeled microphones and speaker cabinets that can be tweaked to produce different sounds. In the NAMM demo, the violins and cellos were quite realistic, as were the electric pianos. As a trombone player, I was not as impressed with the trombones (I know the real thing far too well), but the saxes were reasonably convincing, and the drums were fantastic.
The promise of this technology can finally be realized thanks to much more powerful processors than were available 20 years ago, though new sounds still take a long time to develop. I look forward to hearing more of what SuperNatural can do in the future.