MD equipment query

[Doug Whalen]

   Paul Garrett has made the decision to go with minidisc.  The
website mentioned (www.minidisc.org) is quite useful, but convinces
me that I would never use minidisc.  For phonetic purposes, any
compression algorithm leaves you with unknowns in the signal, and
those will be a concern for every reader of every paper based on
those signals.  Since compression algorithms are typically written by
speakers of English and tested on English, they may do more damage to
distinctive sounds in other languages than they do to English, thus
making the "acceptable" level of compression questionable.  Further,
the duplication of minidiscs to later generations only compounds this
problem with every copy, while duplication of DAT or digitized audio
is essentially error-free for many generations.  (Only when large
stretches of sound are unrecoverable is there any degradation.)
   I know that many researchers are not concerned with phonetic
structure, but it would strike me as a shame to preclude phonetic
analysis by using compression technology.
   Doug Whalen DhW
--
Doug Whalen (whalen at haskins.yale.edu)
Haskins Laboratories
270 Crown St.
New Haven, CT 06511
203-865-6163, ext. 234
FAX:  203-865-8963
http://www.haskins.yale.edu/