impedance

[Ron Nossaman]

>Let's set up some standards so we are all sure we are measuring the same
>things.

Hi Don, 
Yes, that's why I declined offers for measurements at the beginning, since
I didn't have any idea what I was looking for, exactly, and didn't really
want folks wasting their time producing numbers that didn't point to anything. 


>I would submit that we can use *proper* real time analysers from the
>posting I made from that thesis. 

I skimmed back, and found mention that you considered RCT to be at the edge
of usability for this kind of thing. If this is going to require RCT or
better, you'll have to carry on without me. I tune aurally, and that sort
of expenditure isn't in my budget for casual research. I'd sure like access
to the spectrum analysis function in RCT, but I can afford to spend $34 on
intellectual curiosity for Tunelab. We slay our dragons with what swords
are at hand. The check is in the mail Robert.


>Failing that, we should measure each partial *atleast* five times and throw
>out the high number and the low number, then average the results. And
>measure the atleast the first six partials. Obviously the measure devices
>we have are pretty crude.

Ah, like a diving event. Not a bad idea. Now that I've proposed a concept
model of sorts, there's even a target to try to shoot down. This looks like
a reasonable point to call in some offers for data measurement along Don's
proposed standards, to pitch against my preliminary hypothesis and see what
falls out. I would also like to see figures representing the overall pitch
drop of the unison when the individual strings are tuned identically, at an
agreed upon partial as Roger outlined, so we can attempt to get a
correlation between partial frequencies and pitch drop, as I proposed.
Also, I'd like the make. model, and year of the piano tested, for reasons
that should be obvious. 

Pick a partial for the killer octave.


>The piano should be a decent quality instrument. Strings should be seated
>on the bridge just before the unison is tested, and the strings level and
>hammer mating done.

If it's a decent piano, with the strings leveled and mated to the hammer,
seating shouldn't be necessary unless there is obvious falseness in a
string. Even so, I would consider strings false enough to require seating
to be compromised beyond the scope of the data set. We just need a profile,
not every single unison, and I don't wish to condone seating strings to
bridges even in the interest of illuminating this little mystery. It's just me.



>
>I believe we should try to measure the first 12 partials if using the
>*correct* real time analyser software. 
>Regards,
>Don Rose, B.Mus., A.M.U.S., A.MUS., R.M.T., R.P.T.

Possibly, but aren't most of the audible partials in the first six?

Ron N