Digital Synthesis of Plucked-String and Drum Timbres

Abstract

There are many techniques currently used for digital music synthesis, including frequency modulation (FM) synthesis, waveshaping, additive synthesis, and subtractive synthesis. To achieve rich, natural sounds, all of them require fast arithmetic capability, such as is found on expensive computers or digital synthesizers. For musicians and experimenters without access to these machines, musically interesting digital synthesis has been almost impossible. The techniques described in this paper can be implemented quite cheaply on almost any computer. Real-time synthesis implementations have been done for Intel 8080A (by Alex Strong), Texas Instruments TMS9900 (by Kevin Karplus), and SC/MP (by Mike Plass) microprocessors. David Jaffe and Julius Smith have programmed the Systems Concept Digital Synthesizer at the Center for Computer Research in Music and Acoustics (CCRMA) to perform several variants of the algorithms (Jaffe and Smith 1983). Not only are the algorithms simple to implement in software, but hardware realizations are easily done. The authors have designed and tested a custom n-channel metal-oxide semiconductor (nMOS) chip (the Digitar chip), which computes 16 independent notes, each with a sampling rate of 20 KHz.