wjm@whuxk.UUCP (11/28/83)
I'm glad to see this article since it confirms my personal experience about allegedly ultrasonic frequencies that can be heard by at least some Homo Sapiens. While musical fundamentals don't go much above 8KHz, there are significant harmonics that define the quality of the sound (bells don't sound like timpani for example) in the 15KHz+ region. In addition, a sharp cut-off at 22 KHz or so is likely to have measurable effects (especially frequency-dependent phase shifts in the region below 20KHz. This may be why some CD players sound different than others, since they use different designs of low-pass filters. There may be a warning here - do we want to lock ourselves into a digital standard with a 44.1KHz sampling rate (with an upper frequency limit of 22.05KHz? This is the present CD standard. Also this may explain why digital recordings using the Soundstream system (Telarc and Delos) sound better than those using the EIAJ system (like Denon) - Soundstream uses a 50KHz sampling rate. I can't quantify it but I've never heard a Denon recording that sounded "right" to me - there always was a certain "roughness" to the sound. Also, there are very few digital recordings that sound as good as the best Sheffield direct-to- disk high-quality analog ones. Sure, the digitals may be quieter, but I'm talking about musical quality. End of flame. Bill Mitchell Bell Laboratories Whippany, NJ (whuxk!wjm)