[pianotech] Another ETD check

[Ron Nossaman]

On 5/14/2012 12:04 AM, Jim Moy wrote:
> In the spirit of Ron Koval's multi-octave "unison" challenge, here's an
> experiment I've done a half-dozen times or so over the past couple
> years, on "nice" pianos, Yamaha U1 or better that have stabilized so
> pitch adjustment is less than 10 cents:
>
> - Quick tuning to bring it back to the saved tuning from last time.
> - Copy the .tun file.
> - Re-tune by ear to what I think is "best," and capture the override.
> - Go home and study the offsets.
>
> For me, it's +/- 3 or 4 cents, and it's a jagged, random set of offsets.
> Usually only a few in the temperament, but a lot offsets elsewhere. No,
> most of my customers can't tell. I think some may be able to.
>
> But I know.
>
> I'm a relatively new RPT, passed the tuning exam with good scores, not
> CTE level, but pretty good, with some 100's, and the rest 90's, and high
> 80's. But I have a conundrum now, are what I perceive my best tunings to
> be, "better" than the ETD ones? And who cares? It's an interesting puzzle.
>
> I've recently purchased Verituner for my iPod/iPhone recently and plan
> to repeat the experiments. (Can I compare its calculated and measured
> tunings? Tunelab highlights the diffs in red.)
>
> Jim

It is an interesting puzzle, and I think should be approached as a 
puzzle rather than a pair of adrenal glands. In the interest of sticking 
with my observation that ETDs let us worry over numbers beyond the 
resolution of most of the pianos we tune, it has also occurred to me 
that one piano may be a different piano under different circumstances. 
Saving the best tuning for replication with an ETD is a good test case. 
While I think it's an unprecedentedly valuable feature, I also think it 
needs more scrutiny. Now, I know this isn't a new thought, and I'm sure 
someone has already done this, but I don't recall it having been 
mentioned on list so here goes.

A master tuning test, whether it's gotten close by a first pass with an 
ETD or not, is in the end a super refined aural tuning. I'm using this 
as a conceded starting point, because I think it's a quite reasonable, 
and the most practical and nearly possible approach. So I'm not 
questioning the master tuning on a given piano at a given time. This is, 
therefor, *NOT* a criticism of the testing procedure! This tuning is 
then recorded by an ETD for comparison against those of testing 
applicants. Now, were the master tuning on a given piano tuned in 35% RH 
to be compared with a master tuning four months later in 65% (what would 
here be considered a controlled environment), How would the "point 
spread" fall? This goes with the eternal question of what makes a piano 
go out of tune, and an observation by Michael Jorgensen during one of 
the list discussions about a fundamental and long observed point of 
physics concerning vibrating strings. A string at a given length and 
diameter and at a given tension will vibrate at a lower frequency when 
terminated by a less rigid terminus and at a higher frequency when 
terminated by a more terminus. While we know by bearing offset 
measurements and doing the math that soundboards don't rise and fall 
nearly enough to account for the difference we hear in pitch, we can't 
directly measure the differences in soundboard assembly stiffness. All 
we have is the cumulative effect of everything together as measured in 
pitch change. It seems to me that comparing two or more master tunings 
at minimum time between, and maximum RH% differences would indicate what 
affect RH% has on the measured effect of tuning differences - if there 
are any. Since a difference would indicate that the recorded "best shot" 
isn't necessarily the best shot for those conditions (though pretty 
darned close, I'd think) this experiment (or posting the result of 
records comparisons already made) might cast a lot more light on tuning 
in general than what we typically wade through in these discussions. I 
left temperature out here because I presume the piano being master tuned 
for testing will have been in a stable enough temperature for long 
enough to make such a tuning possible.

There are obvious anomalies being skirted in these tuning discussions. 
Likely not intentionally, but the more detailed objective information we 
can compile for comparison, the better our chances of actually learning 
something useful.

Now let's hear from the CTEs and master tuning teams who have the data. 
Has anyone done this?

Ron N