November 96 Champaign, IL - Sound designers will be able to make animals talk, change men into women, and execute other real-time audio special effects with some new tools that Symbolic Sound Corporation has developed for the Kyma Sound Design Workstation; both Kyma and these new tools will be on demonstration at the upcoming 101st Audio Engineering Society Convention, in Los Angeles November 8-11, 1996.
In additon to offering general demonstrations of Kyma, Symbolic Sound will also be introducing several new or improved spectral manipulation tools which operate on the spectrum (or frequency domain representation) of the sound rather than on the waveform (or time-domain representation) of the sound. Among the new tools and improved tools you will be able to see and hear at the AES show:
A vocoder with up to 77 bands which, in addition to providing all the traditional analog vocoder effects (albeit with digital accuracy and reliability), can also be set up so that the center frequencies and bandwidths of all the filters vary over time under the realtime control of MIDI faders or following the pitch of the input.
RE (resonator-exciter) analysis and synthesis is similar to the vocoder in that it applies the formants of one sound onto another. It differs from the vocoder in that it is a two step process: a non-realtime analysis step, followed by a realtime cross-synthesis step. (The vocoder does both the analysis and the synthesis simultaneously and in real time). The reward for doing a non-realtime analysis is a more accurate filter that produces highly intelligible results when cross-synthesizing samples with human speech. Another advantage of the RE method over the vocoder is that you can control the rate at which you move through the time-varying filter in real time, so you can achieve the effect of slowing down or speeding up the speech.
A real-time spectral analysis module lets you take any sample, disk track, synthesized source, or live microphone input, analyze its spectrum, and resynthesize it using classical additive sine wave synthesis all in real time--without having to store the analysis on disk first. What makes this interesting is that Kyma also provides modules for modifying the spectrum before it is resynthesized. You can select subsets of partials and modify their frequencies and/or amplitudes, scale the frequency without affecting the duration, process the even numbered harmonics while leaving the odd harmonics unchanged, jumble up the harmonics, etc. You can also save the spectra on your hard disk and do only the modifications and resynthesis in real time. This can be useful when you need repeatable results or if you are analyzing a recording and don't have the performer present.
With real-time spectral morphing you can create the effect of one sound source changing into another, controlling the rate of morphing in real time using MIDI faders or LFOs. Unlike traditional crossfading, which always sounds like two sound sources, spectral morphing sounds like a single sound source that is changing shape. Tape Gallery has used this technique in radio commercials for Esso and Smirnoff to morph a man's voice from a British to a New York accent, from a piano to a thunderstorm, from a woman to a man, and from women into cats. (See the related story in the Sound Design News section). Morphing features that are new this year include independent control over the rate of frequency and amplitude morphing and the introduction of a new kind of morph that maintains the spectral envelope (or formant regions) throughout the morph.
A graphical spectrum editor lets you view spectra that you have analyzed and saved on disk, listen to individual partials or groups of partials, delete partials, scrub, zoom in on frequency regions of interest, and save modified spectra back to the disk.
In addition to these spectral manipulation tools, Kyma also has several other new synthesis and processing modules, including a graphic equalizer, a module for synthesizing banks of formants, and some tools for generating synthetic spectra.
The idea of Kyma is to put a modular synthesizer, sampler, and effects processor entirely into software where it can be more easily manipulated and altered, and to dedicate a box full of DSPs to generating and processing the sound in real time. The user can design new sounds by hooking modules together graphically on the screen of a Macintosh or Windows personal computer and those sounds are generated and processed in real time on the Capybara-66: an audio co-processor stuffed with up to eight Motorola 56002 DSPs running at 66 MHz and connected to your computer via ISA, NuBus or PCI interface card. The result is fully reconfigurable hardware that is entirely dedicated to sound--no need to share cycles with the operating system or the graphics.
Kyma is being used to generate special sound effects for motion pictures, advertising, music, computer games, and multimedia. You can hear Kyma in such diverse applications as: the backgrounds for the upcoming Muppets version of Dr. Seuss books, morphing voiceovers in a London radio spot for Smirnoff vodka, the voice of SID 6.7 in Paramount Pictures' Virtuosity, live experimental music performances, the new system alert sounds for a well-known computer operating system, a major science fiction film scheduled for release this fall, and a wide range of other applications that require unique sounds and special sound effects.
The Symbolic Sound Corporation designs, produces, and markets hardware and software for digital audio. The first Kyma system was delivered in January 1991, and since that time, they have produced three major software upgrades, two hardware upgrades and have ported the system from the original Macintosh platform to Windows machines. This year, they introduced Kyma version 4.1, doubled the clock speed on the Capybara to 66 megahertz, and in September, started shipping a PCI interface card for the newest Macintosh and Windows computers.
For additional information:
Contact Symbolic Sound Corporation
Toll free: 1-800-972-1749 (US)
Voice: +1-217-355-6273
Fax: +1-217-355-6562
Email: info-kyma@symbolicSound.com
URL: http://www.symbolicSound.com/
November 96 Champaign, IL - Symbolic Sound Corporation's Kyma Sound Design Workstation will be 10 years old in November, 1996. Kyma is a graphical signal flow language that runs on Macintosh and Windows computers and does real-time synthesis and processing on a multi-DSP signal processor called the Capybara.
The first version of Kyma was written by Carla Scaletti in November 1986, and it used the Platypus digital audio signal processor built by Lippold Haken and Kurt Hebel of the CERL Sound Group. In August 1987, Scaletti gave a live demonstration of the system during her paper presentation at the International Computer Music Conference. This presentation was written up by Bob Moog in Keyboard as one of the technologies to watch in the future (Kyma shared this designation with three other technologies in that article: "physical modelling", "granular synthesis", and the Motorola 56000 DSP).
In 1989 Scaletti and Hebel left the CERL Sound Group to form Kymatics, later to become Symbolic Sound Corporation. In 1990, Kyma was ported to the Capybara, Hebel's multiple-56000 replacement for the single-board Platypus, and the first Kyma system was delivered in January, 1991.
Since that time, the hardware has been upgraded first to the Capybara-33, and then this year to the Capybara-66. There have been three major software releases, three minor releases, and the software has been ported from the original Macintosh platform to include Windows platforms, and with the introduction of the PCI card this year, Power Macintoshes as well.
Symbolic Sound would like to thank all Kyma owners for their foresight, vision, and support of the future of sound and music on computers!
1 September 96 Champaign, IL - Symbolic Sound Corporation announces a new PCI bus interface card for connecting their Capybara-66 digital audio signal processor to Macintosh or Wintel computers. This makes the Capybara compatible with all of the new PowerPC computers and provides improved disk bandwidth on Wintel machines. Current Kyma owners can order a PCI interface card for US$225. New systems are being shipped with the choice of either an ISA, NuBus, or PCI interface card (with additional interface cards and software available for $225 each).
Kyma is a sound design environment for musicians, sound designers and researchers. Sounds are designed graphically on the screen of a Macintosh or Wintel machine and generated or processed in real time on the multiple-DSP Capybara. Because of the graphical "patch cords", it has the feel of a modular synthesizer, while also providing some advanced processing and synthesis techniques usually associated with non-realtime computer music: spectral morphing, cross-synthesis, spectral analysis and additive resynthesis, granular synthesis, and others.
Kyma is being used to generate special sound FX for motion pictures, advertising, music, computer games, and multimedia. You can hear Kyma in such diverse applications as: the backgrounds for the upcoming Muppets version of Dr. Seuss books, morphing voiceovers in a London radio spot for Smirnoff vodka, the voice of SID 6.7 in Paramount Pictures' Virtuosity, live experimental music in Rome, alert sounds for computer operating systems, and a wide range of other applications that require unique sounds and special sound effects.
16 May 96 Champaign, IL - Symbolic Sound Corporation announces the Capybara-66, a new sound processing and synthesis engine for its Kyma sound design workstation. Due to software optimizations and a doubling of the clock speed, this new hardware provides more than twice the polyphony of the previous (Capybara-33) hardware. Additionally, the Capybara-66 will come with 3 megabytes of sample RAM per card standard. All existing Capybaras will be upgradable, meaning that current Kyma users will be able to upgrade, rather than having to retire the old hardware.
The Capybara-66 is an external rack-mountable processor that has a motherboard and eight expansion slots. Each expansion card provides an additional Motorola 56002 DSP processor running at 66 Mhz plus an additional 3 megabytes of sample RAM. The Capybara has 16-bit oversampled stereo D/A and A/D converters standard with an optional AES/EBU and S/PDIF digital I/O. Internal arithmetic is done at 24-bits per sample, so it can take advantage of external, 20-bit converters.
With the introduction of the new hardware, Symbolic Sound also announces that the entry-level configuration has been changed from a 3-card to a 2-card system for a new entry-level price of US $4400. The new entry-level system is 33% more powerful than the previous, 3 card system, has four times the sampling RAM, and yet is 12% less expensive.
Even though additional expansion cards have twice the processing power and four times the sample RAM of the previous version, the $595 price of each additional expansion card has not changed.
The idea of Kyma is to put a modular synthesizer, sampler, and effects processor entirely into software where it can be easily integrated with other software such as sequencers, wave editors and digital audio workstations. Synthesis and processing algorithms are designed by hooking modules together graphically on the screen of your Macintosh or Windows personal computer. The graphical signal-flow diagrams are computed in real time by the DSPs in the Capybara-66.
In addition to providing standard synthesis and processing modules in an extremely flexible, modular form, Kyma also provides some unique, frequency-domain processing algorithms including spectral morphing and cross-synthesis.
Kyma is intended for sound designers in applications that require the design of new, unusual, and/or customized sounds rather than reliance on factory presets or existing sound libraries. Sound design cuts across several fields including music (especially in music where the sounds themselves are as important as the actual pitches and rhythms), sound for picture, post production, multimedia, computer games, psychoacoustic research, and data-driven sound for immersive environments.
Kyma was created as a tool for sound designers who need a maximum of flexibility, modularity, and reconfigurability for creating and manipulating sounds.
Kyma has been used in a wide range of contexts ranging from the voice processing on Sid 6.7 in Paramount Pictures' Virtuosity, to test signals used in psychoacoustics laboratories, to morphs in radio and television advertisments, to the sound synthesis and processing used in Public Organ (an interactive sound installation on the World Wide Web) to live performances and recordings of music in a variety of styles.
For more detailed stories on how Kyma is being used, see Symbolic Sound's online newsletter, the Eighth Nerve (http://www.symbolicsound.com/eighth.html).
The Symbolic Sound Corporation produces hardware and software for digital audio. The first Kyma system was shipped in January 1991. Work on the software was begun in 1986, and Kyma was first demonstrated publicly on a Macintosh with a Platypus signal processor at the International Computer Music Conference during the summer of 1987.
1 FEB 96 Champaign, IL - Symbolic Sound Corporation will release version 4.1 of the Kyma sound design workstation software on February 26, 1996. Some of the new features and additions include:
Built-in spectral analysis tools let you analyze your own samples and resynthesize them (with spectral morphing or warping) using additive synthesis. New tools for specifying your own tuning tables let you take advantage of Kyma's 0.0026 Hz frequency resolution to build microtonal scales and other alternate tunings by remapping MIDI note numbers to any arbitrary frequencies. A MIDI scripting language lets you generate MIDI sequences algorithmically (in addition to the standard features of being able to use MIDI files and live MIDI from sequencers or MIDI controllers).
RE-Resonators are time-varying filter banks derived by analyzing voice or instrument samples. By analyzing your voice and then feeding another sample through the resulting filter, you can get vocoder-like effects like "talking guitars" or drums.
Using the SOS (Sum-of-Sines) module introduced in Kyma last year, you can do additive synthesis using hundreds of sine wave oscillators. A new module called GA-Oscillators lets you do additive synthesis with many fewer oscillators (3-5), each with a more complex waveform rather than a sine wave. Both methods let you resynthesize from your own spectral analyses and both allow for independent time/frequency controls and spectral morphing.
The multi-file disk player lets you use MIDI to trigger audio tracks stored on your hard disk in real time. This turns Kyma into a disk-based sampler and means that there is no limit on the duration of samples (other than the amount of free disk space on your hard drive).
Other new modules include: an all pass filter for phasing effects, a dynamic range controller for real-time control over compression and expansion, and an iterative waveshaper for nonlinear synthesis.
Software optimizations have resulted in 20 to 130% increases in real-time polyphony.
A real-time evaluator in Kyma lets you use arithmetic combinations of real-time MIDI controls as parameters of your sounds, for example, to cause both pitch bend and key velocity to control the pitch of a sample, you could use:
(!Notenumber + !PitchBend + !KeyVelocity sqrt) nn
The evaluator now performs common subexpression elimination, with the result that these expressions can be up to 50% more complicated and still be evaluated in real time. Other additions to the evaluator include ramp generators, a "metronome" for generating triggers, and the ability use a MIDI value as an index into an array.
All MIDI-controlled parameters show up in a "virtual control surface" on your computer screen. Virtual faders can now have log or linear taper, min and max values, and their current settings can be captured and saved.
The idea behind Kyma is put sound synthesis, processing, sampling, and editing operations entirely into software. Kyma runs under both the Macintosh and Windows operating systems and uses a multiple-DSP56002 mainframe called the Capybara to perform the sound synthesis and processing in real time. The Capybara attaches to your computer through either a NuBus or an ISA interface card. You design new sounds graphically, by manipulating signal flow diagrams on the screen of your Macintosh or PC.
Kyma 4.0 was reviewed in the July 1995 issue of Electronic Musician.
Kyma is intended for sound designers in applications that require the design of new, unusual, and/or customized sounds rather than reliance on factory presets or existing sound libraries. Sound design cuts across several fields including music (especially in music where the sounds themselves are as important as the actual pitches and rhythms), sound for picture, post production, multimedia, computer games, psychoacoustic research, data-driven sound for immersive environments.
Kyma was created as a tool for sound designers who need a maximum of flexibility, modularity, and reconfigurability for creating and manipulating sounds.
Kyma has been used in a wide range of contexts ranging from the voice processing on Sid 6.7 in Paramount Pictures' Virtuosity, to test signals used in psychoacoustics laboratories, to animal morphs in television advertisments, to the sound synthesis and processing used in Public Organ (an interactive sound installation on the World Wide Web) to live performances and recordings of music in a variety of styles.
The Symbolic Sound Corporation produces hardware and software for digital audio. The first Kyma system was shipped in January 1991. Work on the software was begun in 1986, and Kyma was first demonstrated publicly on a Macintosh with a Platypus signal processor at the International Computer Music Conference during the summer of 1987.
July 24, 1995 Champaign, IL - When you hear the voice of injured cyber-villain Sid 6.7 in Paramount Pictures' August 4th release Virtuosity, you are hearing the actor's voice processed through Symbolic Sound's Kyma sound design workstation.
Symbolic Sound's Kyma is to sound what Photoshop, Avid or an SGI workstation is to graphics. It's a tool for generating, shaping, processing, and morphing sounds on the computer. Sound designers use Kyma to create special sound effects to go along with the special visual effects you see in films, computer games, videos, multimedia and any kind of music that uses electronic sound or electronic processing of sound (including experimental, industrial, techno, ambient, computer music and other genres).
Frank Serafine, president of Serafine Sound Design in Venice, California and designer of special sound effects for feature films like Star Trek, The Lawnmower Man, and Hunt for Red October as well as commercials for Nike, Energizer, Microsoft and others, heard about Kyma from his engineer Francois Blaignan. They were in the middle of creating the effects for Virtuosity and were searching for a way to process the cyber-villain's voice so that it sounded menacing yet slightly glitchy, as though his nano-motors were starting to fail. Blaignan had seen a review of Kyma in the July issue of Electronic Musician magazine and told Serafine that they had to get Kyma.
With only three weeks left for post production, a frantic call went out to Symbolic Sound, and thanks to some overtime on the part of SSC engineer Barry O'Dell, overnight delivery by UPS, some telephone calls and Email to Kyma software and hardware designers Carla Scaletti and Kurt Hebel, and lots of file sharing via the company's FTP site, the perfect cyber-villain voice was created. In fact, Kyma ended up being used throughout the movie, whenever the cyber villain gets low on energy or when other people or animals have to sound machine-like or computerized.
"It's a cool box," according to Serafine, who hopes to use Kyma for the sound effects in an upcoming sci fi television series.
The Kyma sound designer's workstation includes software for graphically designing, shaping, and processing sound on the screen of a Macintosh or Windows OS computer plus SSC's own hardware accelerator for generating and processing the sound in real time. It is being used internationally by professional sound designers who need to be able to create their own unique sounds for film and video, computer games, multimedia, virtual reality, and music.
Symbolic Sound Corporation was founded in 1989 by Carla Scaletti and Kurt J. Hebel. For more information on Symbolic Sound, its products, its founders and its users, please refer to the World Wide Web press kit: http://www.SymbolicSound.com/pressKit.html
