Verituner Update

[Dave Carpenter]

--- Richard Brekne <Richard.Brekne@grieg.uib.no> wrote:

> Curious about something David. We discuss back and
> forth a bit about ETD's and
> measurments. No one seems to feel that measurements
> are all that dependable
> unless you take a few samples for each note and
> average them out.
>
> Does the Verituner utilize some new and more
> dependable form for measuring
> inharmonicities, or do you not consider this to be a
> problem ?


Richard,

This is very true. There is a lot of uncertainty built into these
measurements, because you are trying to measure a moving target that really
may not even have one right answer, or even one definition of what the right
answer should be.

That "problem" cannot be (nor should be) solved, but we can improve the
measurement process as much as possible. The Verituner makes all of its
measurements in real time, i.e. there are not really any "samples". This
allows it to better measure the instantaneous values of the pitches of
partials. Since inharmonicity values are really the ratio of pitches, the
Verituner measures these instantaneous ratios directly, rather than
individually measuring each partial's pitch over some period of time. This
eliminates an extra step which may introduce error, and allows it to better
track the rise and fall in pitch of all partials as you are yanking on the
tuning pin!

Each of these measurements is evaluated as to its quality by a scoring
algorithm, and then the result is continually refined based on the weighted
average of individual measurements as they are available.

But we should consider this issue as it relates to calculating tunings. If
the inharmonicity is not very consistent when being carefully measured in
isolation, then certainly it is not very consistent when playing the piano.
For purposes of tuning, surely there is some point of diminishing returns as
to improving the accuracy of these measurements.


Dave Carpenter, RPT
Veritune, Inc.