string terminations

[Stephen Birkett]

Hi,

I collected up a few snippets for comments rather 
than contributing a stack of mini-messages by 
replying to each one separately .....

Ric:
>  I hope you will leave them up for a while as 
>they are certainly worth looking over several 
>times.  Nothing like a bit of real equipment for 
>observations to get things <<vibrating>> as it 
>were.

Sure. The videos on the website have actually 
been there for a while now (same with the SEM 
photos of felt hammers - did you see those yet?). 
The intent is to keep them there as samples for 
discussion. They're there mainly to show the 
sorts of images that are possible with the 
equipment. As I said previously, I wouldn't try 
to theorize too much about what's going on from a 
mechanism point of view (e.g. terminations etc), 
other than concluding that the effects are quite 
complicated and probably not  subject to simple 
generalizations. Also, three strings unisons need 
to be studied, since the interactions between 
strings via the bridge are quite extreme (as we 
all know already from experience and instinct). 
The piano used in the videos was not even tuned 
prior to taking those videos and unisons were 
almost certainly out of tune at the time.

I have some super macro shots too from when we 
investigated trichord phasing comparing with and 
without pitchloc installed. These were taken on a 
three-string monochord constructed to a piano 
configuration, including full tensions, but only 
a short segment of bridge on a narrow 
"soundboard". In effect, this is equivalent to a 
low impedance bridge termination and, in this 
scenario, you can see the bridge move as it's 
pumped by the string in the video, something 
which is not visible at the resolution used in 
the real piano bridge video at C64.  Again, it's 
more suggestive than equivocal as far as testing 
theories, because a careful sequence of targeted 
experiments is needed to make valid conclusions. 
But it does show how the equipment can be used.

It's also possible to extract quantitative data 
from these videos, by tracking points. We've used 
that method, for instance, to measure the 
rotation of the jack with respect to the whippen, 
which is otherwise impossible since the pivot 
moves and defeats normal measurement techniques. 
With the video tracking you can monitor, and then 
eliminate, the motion of the pivot to obtain the 
rotation of the jack in the frame of reference of 
the whippen.

Andrew:
>I wonder if the bridge would move visibly for the lower monochord notes.
Stéphane Collin:
>I noticed that you can't see the bridge move 
>under influence of the strings.  Is this because 
>the magnitude of bridge motion is invisible at 
>this scale, or because the bridge doesn't move 
>that much, being after all a node in the whole 
>story ?
Ric:
>Very interesting to watch the c64 bridge avi.  I 
>ran this through extra slow motion on my viewer 
>and for the life of me I can not see hardly any 
>downward deflection at the bridge pin by the 
>string. The strings upward motion at the pin 
>(and immediatly away from it as well) is quite 
>obvious to see, but if you mark the initial 
>position of the string, it never seems to go 
>below that plane. More then likely some effect 
>from the angle.... or.. ?  It would be nice to 
>see this same shot from a perpendicular angle to 
>the string on the string plane.

All very interesting questions. The wheels are 
already set in motion on this stuff, but it's 
going to take some time to do it properly. All 
input gladly welcomed. And the lab is always open 
for visitors if you're close enough to drop in.

Andrew:
>What I found really telling was the high speed 
>video of the felt punchings under the key.  I'm 
>seeing several rebounds there.  Any change we 
>can get a high speed video of the action doing 
>its thing?  This is fascinating.

I have stacks of high speed videos of actions 
taken for validating the predictions of the 
dynamic simulation model that's been under 
development for the past 3 years. That's one 
possibility that could be explored in Rochester 
next year. It would be nice to coordinate with 
others who are doing relevant classes, say, and 
show effects of some process of technique using 
the camera (e.g. before and after). It's feasible 
to set this up on the spot since the camera 
equipment is portable.

Ron N:
>Cool stuff, I WANT one! One thing - either my 
>system is psychotic (distinct possibility) or 
>the C64 damper video brings up the D6 damper 
>instead. Also, a question about that C64 bridge 
>sequence. How fast is that unison beating?

And thanks to others too who noticed the screwed 
up link. It should be fixed now. Continue to use 
the mirrorr site please at 
http://www.fortepianos.com, since ftp on my 
regular website is still out of action.

Last, Ron asks about how fast. Both D6 and C64 
were taken at 2000fps. The D6 = 38Hz fundamental 
is played back at 30fps, so the apparent 
frequency observed in the video is (30/2000)*38 = 
0.57Hz. The C64 = 1047Hz fundamental is played 
back at 4fps, so apparent observed freq in the 
video is (4/2000)*1047 = 2.1Hz.

Stephen
-- 
Dr Stephen Birkett
Piano Design Lab
Department of Systems Design Engineering
University of Waterloo, Waterloo ON Canada N2L 3G1
tel: 519-888-4567 Ext. 3792
Lab room E3-3160 Ext. 7115
mailto: sbirkett[at]real.uwaterloo.ca
http://real.uwaterloo.ca/~sbirkett