source file: m1396.txt
Date: Sun, 26 Apr 1998 10:23:04 -0400

Subject: TUNING digest 1395: JI Tuning Resolution

From: Daniel Wolf <DJWOLF_MATERIAL@compuserve.com>

Unless you can get a tuning resolution in hertz at least approaching that=

of the Rayna card (with double precision, I typically use 20 voices with
frequency units of .0000044139 Hz.) I would recommend a cents-resolution
for a synthesizer for work in aproximating JI. The question of what that
resolution should be can be best answered on the basis of finding a large=

equal temperament in which the limit you want to work in is consistantly
represented. =


Whether manufacturers come close to the best cent resolutions or not is a=
n
entirely different story:  I bought the Rayna because I was frustrated wi=
th
the inconsistant representation of 3's and 9's on the TX81Z (resolution
around 1.5625 cents or 768TET); several others have pointed out that a
resolution of 1 cent (1200TET) leads to similar if not worse problems. So=
,
it is a matter of lobbying and educating manufactures that if they are
willing to add microtonal capacity, this is better addressed not
necessarily by a larger subdivision but rather by a more consistant one.

Since many manufacturers will probably continue to specify tunings in ter=
ms
of deviations from 12tet pitch classes, it would probably be best to lobb=
y
on the basis of two large equal temperaments, one chose entirely on the
basis of prime consistancy and another on this basis, but limited to thos=
e
temperaments divisible by 12.

       =