The use of dither in digital sound production equipment such as electronic music systems is necessarily a compromise, required by practical limitations on wordlength, between production of signal dependent distortions and processing which introduces low level broadband noise. In this paper we investigate this compromise and some associated algorithms, which are particularly applicable to music synthesis. These optimize the psychoaccoustical effects of dither by minimizing the idle channel noise while still maintaining the benefits of dither during enveloping. We have found that these produce slightly superior results at minimal cost. Additionally, background on minimal cost dither hardware is given.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.