Yahoo Tuning Groups Ultimate Backup tuning Empirical study concerning 19ET

--- In tuning@yahoogroups.com, "Gene Ward Smith" <gwsmith@s...> wrote:
> --- In tuning@yahoogroups.com, "Martin Braun" <nombraun@t...> wrote:

> > In April this year there appeared an empirical study of consonance
> > and dissonance in 19ET:

> Here's a tinyurl: http://tinyurl.com/6wlpk

The authors don't present their data; they seem to think it sufficies
to draw pretty pictures. The only thing they give a table for is
something you don't need a table for, namely how many cents make up
the various 19-et intervals. Really, really sloppy; I know the
scientific community often errs inn this department but here it is
simply unacceptable in my opinion, since in this instance you cannot
tell for certain which interval is even being talked about simply by
eye. To my eye, for instance, it looks as if the 758 cent fifth is
very consonant, and the meantone fifth highly dissonant. Not only does
this result not seem credible, it contradicts what the authors say.
Aside from the results allegedly presented in figure 1, which I am
complaining of, figure 4 also presents results; these are more
comprehensible since you can at least tell which interval is which,
but a table would have been preferable. Moreover, the results I see do
not always accord with what the authors say, so it's hard to say for
sure what their data actually is.

From figure 4, the most consonant intervals were judged to be the
fourth and fifth (sharp and flat as they are), then the minor third
and major sixth (which in 19 are nearly pure, and this may explain why
they outscore the major third.) After this so far as I can see is 9/5,
but this contradicts the claim that neither of the two possible 7th
partial approximations score well. *After* 9/5 and its inversion, the
tone, if you believe the data alledgedly "presented" in this table,
and assuming I am not looking at it cross-eyed, always a possibility,
we finally get 5/4.

If you print it out figure 1 and use a ruler you can probably guess
correctly what the interval supposedly being depicited actually is,
and certainly can get a good idea of what the consonance measure is
assuming your guess is right.

Shame on the referee for this paper! He could have and should have
told the authors to tabulate some of this data.

Empirical study concerning 19ET

🔗Martin Braun <nombraun@telia.com>

🔗Gene Ward Smith <gwsmith@svpal.org>

🔗Gene Ward Smith <gwsmith@svpal.org>