machine tuning

[Robert Scott]

James Turner wrote:

>>>>
I have been thinking about getting the SAT lll, RCT or the TuneLab.
When one tunes aurally, we listen to every note on the piano, intervals
and so on.  What puzzels me is how a machine can measure only 3 or 6
notes and compute an optimum tuning for a piano.  It seems to me that
for any machine or computer to create a really good tuning, it would
have to sample many more notes than 3 or 6?  Wouldn't a machine that
sampled every note on the piano be a better tuning? Isn't this what
aural tuning does to a degree?
<<<<

While it is true that an aural tuner listens to every note in the
course of tuning a piano, I don't think he needs to listens to
every note before he starts actual tuning.  In fact, don't
aural tuners integrate the tuning and measuring functions as
they go?  What the machine can do is make measurements before
any tuning occurs. This is a potential time savings.

If experience shows that inharmonicities of a certain group of notes
can be well represented by the inharmonicites of only a few
representatives, then it makes sense to just measure those
representatives.  You can debate whether 3 or 6 or 15 notes is
sufficient to describe the piano, but I doubt that you ever
will need to separately measure the characteristics of more than
a few notes.

If you measure these notes in TuneLab, then you can see a graph that
shows how the beats of various intervals will work out based on the
amount and the shape of the stretch you want to use.  Then you can
experiment with that graph by pulling and pushing various parts of
it around.  For any modification of the stretch curve, the
consequences in various intervals can be instantly seen.  You can
see, for example, how one degree of stretch will make the single octaves
beatless, but will cause the double octaves to beat narrow.  Then
you can nudge the stretch a little higher and see how you can make
the double octaves beatless and the single octaves a little wide.
You can even experiment with perfect fifth tuning and see the
consequences in the octaves, all without lifting a tuning hammer.
If you have a Windows 95 computer and would like to give it a
try, point your web browser at  http://www.wwnet.net/~rscott
All it costs to try it out is your time.

Robert Scott
Detroit-Windsor Chapter