I have done a very unscientific survey with my last three hammer installs and have concluded that there is a general correlation between weight, density and pitch. So now I am doing (1) sort by weight, (2) channel, (3) sort by weight, (4) install on rail and check pitch, (5) rearrange or remove as necessary. I find not much to do (5).
I find that the difference in weight and in ptich is miniscule .1 or .2, so moving them around to suit the pitch is not affecting the overall strikeweight enough to care about touch wise but COULD be affecting the overall tone somewhat and I have no real data here, other than a hunch that Tim is right and it feels right. I can hear a difference when I don't sort shanks at all, so that tells me the process is valid not just for reducing the amount of work to get a smooth strikeweight but also MAY contribute to smoothing the tone as well. The wild cards are as Ric B describes, real honkers that have no real pitch and these should be removed. Lord help you if you find more than two in a set. Wake up manufacturers!!
As David and Jon Page have noted unless you are measuring shank radius weight and zeroing out the flange we really can't compare results.
Hope this adds to the mix.
Chris Solliday rpt
----- Original Message -----
From: Albert Picknell
To: College and University Technicians
Sent: Thursday, May 15, 2008 12:19 AM
Subject: Re: [CAUT] strikeweight
Tim
I agree that there can be a wide variance in shank pitch, and that it amounts to more than a few shanks that go "plock". I should probably clarify that when I sort shanks, I find that there is not only a range of pitches, but a whole spectrum of tonal qualities as well, from the plinkiest plink to the plockiest plock. Often there will be two shanks with almost identical pitches, but one rings out clearly while the other requires much closer concentration to identify its pitch. In this case, I did not weed out the plockers.
I too am sort of sitting in the middle on the sorting by weight discussion, but with more research we may eventually get a little closer to an ideal (not that there will ever be anything resembling universal agreement on what that ideal is!). Perhaps the "strikeweight" people are sacrificing something in terms of voicing, while the "shank pitch" people are sacrificing something in terms of evenness of touch; both have valid reasons for choosing to give one characteristic precedence over the other. Perhaps one "ideal" would be to buy ninety sets of shanks and sort them all by pitch and by radius weight :>) Unfortunately, my R&D budget doesn't permit me the luxury of trying it out :>(
Pleased to be taking part in this respectful discussion,
Albert
Tim Coates <tcoates1 at sio.midco.net> wrote:
Albert,
It would be nice to see a study that correlates strike weight to pitch. I sorted the last set of shanks I installed first by strike weight and then checked to see if they were in pitch order. They weren't. I tried clipping off some excess shank to find a pitch change, I really couldn't. This particular set only had a variance of .2 of a gram throughout the entire set. I reordered them by pitch and installed them in that order. I will continue to sort by pitch because it makes my voicing much easier.
I respectfully disagree that shank pitch is insignificant. I use the word "respectfully" purposely. I know the "strikeweight" people have their reasons for insisting on using their methods to sort, but I know of others more learned than me who feel that type of sorting is inconsequential. I am sitting in the middle about the sorting by weight discussion.
I find there is a very, very wide variance in shank pitch and it amounts to more than just a few shanks that go "plock". It ends up being a wide variance with a very even progression of pitch.
I'm not trying to argue here just present experiences that I have and share them. I have not responded to much of the discussion since I first brought up the shank pitch. I have not been swayed by the information presented to change my ways. It isn't worth arguing about and I want to make sure no one thinks I am trying say my method is the correct method. I am just saying it is comfortable for me.
Tim Coates
On May 14, 2008, at 7:43 PM, Albert Picknell wrote:
Thank you, Ed
Your first sentence states directly what I was hinting at in my last point, namely that since it would be very difficult to predict what resonating qualities a shank/hammer assembly will have once the shank ends are trimmed off, it may be rather pointless to try to use shank pitch as a primary sorting criterion. And your second sentence reminds me of what Ted Sambell taught us many years ago (I was one of his students back in the '80's): always listen to the tone of the shanks before installing them. The ones that go "plink" can go in the piano; the ones that go "plock" can go somewhere else. There is no sorting by pitch, just a test that weeds out the shanks that are more likely either to break due to irregularities in the grain, or to adversely affect the tone by flexing too much, damping tone, etc.
Thank you, David, for your comments too. It sounds like there is more to be gained by sorting shanks according to what effect they will have on the touch rather than what pitch they produce before being coupled with hammers and mounted on rails. As long as they are good and stiff (they go "plink" rather than "plock") they should do the job.
Am I reading you correctly?
Albert
Ed Sutton <ed440 at mindspring.com> wrote:
David-
Once the hammers are hung, the "pitch" of the shank/hammer will be altered, so I don't see how the "shank tone" as such is significant.
However, when all other factors are the same, it may be an indicator of the stiffness of the wood, which may influence the response of the action.
For example, my sense in a short trial of Bruce Clark's action with carbon fiber shanks was that it was fast and even in response and delivered easy power for the effort. But that was a short trial by a low-skilled performer, and there are many other creative adaptations in his design that make it work so well.
Nevertheless, those carbon fiber tubes should be able to deliver a very perfect and even "plinck" line. not to mention even weight and stiffness.
Ed Sutton
----- Original Message -----
From: David C. Stanwood
To: College and University Technicians
Sent: Wednesday, May 14, 2008 6:03 PM
Subject: Re: [CAUT] strikeweight
Dear Albert,
Great work and very interesting and important ideas you are working with! My comment: Most of the dead weight is concentrated in the flange and flange/knuckle end of the shank and I would imagine that for that reason the dead weight value might relate so much to it's effect on tone...
I would be very interested to see additional data using Shank Strike Weight (SS) instead of the dead weight of the Flange/Shank assembly. This value measures the weight of the shank tipped on a roller bearing with the flange oriented vertically so that it's weight is not measured. The end of the shank rests on the scale. Values are usually aroun 1.4g for narrow shanks and 1.8g for regular shanks. We routinely sort shanks, within each type, by weight, then hang the hammers, then measure Strikeweights, then add or subtract hammer weight to smooth the strikeweights to a curve of our choosing.
The "thinking" is as follows: Shank Strike Weights can very within a shank type within a set by as much as 0.6g. These variations don't show up in the StrikeWeight measure but when we measure the Strikeweight and make changes in hammer weight to smooth the curve we may be changing hammer weight to compensate for a variation in SS. .6g of SS will not have the same inertial moment as .6g of hammer weight because the center of weight is different. (a physicist could explaing this more eloquantly than me). So by sorting the SS by weight we theoretically make the inertial moments of the shank/hammer more even as related to smooth Strike Weights.
Here is a drawing of the setup:
http://www.stanwoodpiano.com/ss.jpg
Hope this helps.
David Stanwood
Hello List
Chris Solliday <csolliday at rcn.com> wrote ('way back on Feb 20):
Alot of good ideas and ways for producing some very refined work are being floated regarding shank radius weight and hammerweight which combine to produce strikeweight and the action's main contribution to overall tone. ...
...I pre-sort the shanks heavy to light bass to treble before I channel them and then again after channeling them. I too find that this reduces the quantity of the variation if not the relative variation. I do not make a spreadsheet until that point after the second sorting. ...
...I may be going over the shanks twice but I have much less work in the end.
I am intrigued at the possibility of working shank tone into the equation and will be first looking for a correlation between pitch and weight.
Thanks,
Chris Solliday
This is my first posting to this list, so I hope at least some of you find what I have to say interesting and/or useful. Back around mid-February a series of threads ran on this list entitled "Shank to Hammer weight spreadsheet", "strikeweight", and "Shank Pitch". The comments at the very end of Chris Solliday's post (see above) particularly caught my attention, so I thought I'd do a little "tinking" and weighing to generate some data which Chris (or anyone else) might find useful.
My data-gathering proceeded as follows:
Taking a box of new Renner shanks with flanges for Steinway, I first separated the "regular" from the "thinned" shanks; the set contained 59 and 31 shanks respectively. Then I listened to the pitch of the shanks and arranged them in order from lowest to highest. Interestingly, both groups of shanks fell into the same overall pitch range, i.e. the major third A#5 to D6. The thinned shanks covered a slightly narrower range, but that is probably due to the fact that there were fewer of them.
Next, I weighed each shank/flange assembly and recorded its weight, to the nearest tenth of a gram. This was just the dead weight of each assembly on the scale.
Next, using a Correx gauge, I measured centre pin friction, also to the nearest tenth of a gram. This involved some estimating and averaging, but I used a consistent technique, so I think the numbers are pretty good.
I entered these data into an Excel file, and generated charts from them in order to visually illustrate whatever correlations might exist. The file is attached, including charts - have a look. The data series with the connected blue dots represent the regular shanks; the unconnected pink dots represent the thinned shanks. The lowest- and highest-pitched thinned shanks are numbered to correspond with the regular shanks which had the most closely matching pitches; the rest of the thinned shanks are distributed as evenly as possible between those two extremes. Distributing them this way enabled me to plot them all on the same graphs in a somewhat meaningful way.
Finally, to further explore the relationships of shank thickness and shank length to shank pitch, I altered three regular shanks as follows. The first one, which had an initial weight of 7.0 g (including flange), I thinned substantially, removing 0.5 g of material. The pitch of this shank dropped by about a minor 2nd. The second one, which had an initial weight of 6.9 g (including flange), I shortened by approximately 24-25 mm, equivalent to 0.4 g of material; the pitch of this shank rose by about a perfect 4th. The third one, which had an initial weight of 8.5 g (it had a larger flange attached), I first thinned by 0.5 g, which lowered the pitch by a little less than a major 2nd. Then I cut off shorter segments of approximately 7 mm each (each weighing a little under 0.2 g); each of these cuts raised the pitch about a major 2nd; the cumulative effect of these three cuts was a pitch rise of about a tritone. Altogether, this last shank ended up thinner, shorter, and about a major third higher in pitch than where it was at the beginning.
Some observations/conclusions:
1. As I mentioned above, both the regular and thinned shanks fell into the same overall pitch range, i.e. the major third A#5 to D6. Hence, if one is going to sort shanks strictly on the basis of pitch, the regular and thinned shanks will end up being interspersed.
2. There is a significant amount of overlap in the weight ranges of the regular and thinned shanks. So if one is going to sort shanks strictly on the basis of dead weight, again the regular and thinned shanks will end up being interspersed.
3. The trendlines in the "Pitch vs. Weight" chart seem to indicate that, as a general rule, heavier shanks have a higher pitch. For two reasons, I suspect that the variations in pitch are primarily a result of differences in wood density from shank to shank. First, because the substantial thinning I did on two of the shanks I altered resulted in pitch changes of less than a major 2nd, I doubt that the slight dimensional variations which may exist after Renner's precise manufacturing process are likely to result in pitch differences amounting to a major 3rd. Second, the fact that the regular and thinned shanks produce pitches that fall within the same range suggests that something other than dimensional variations are responsible for the pitch variations. Another obviously potential source of variation in the weighing process is differences in the weights of the flanges. But I suspect that if one took the trouble to weigh the flanges separately, although there would be some variation, the data would generate a flat trendline. Anyone wishing to test this hypothesis is welcome to do so; right now I don't have time.
4. The random distribution of tighter and looser flanges throughout the entire range of pitches, and the flat trendlines in the "Pitch vs. Friction" chart seem to indicate that the pitch of the shanks is not affected by the pinning (although I do believe the pinning does affect the tone in the piano). To test this conclusion a little further, I took a relatively tight assembly, treated it with CLP to reduce the centre pin friction, and listened to the pitch again; there was no change in pitch.
5. Removing material from the end of a shank has a significantly greater effect on the shank's pitch than does removing an equivalent amount from the sides. Whether this is something that needs to be taken into account when sorting shanks may be worth considering, because when the shank ends are trimmed after the hammers are installed, they aren't all necessarily shortened by the same amount.
The really tough question now is, what am I going to do with these things?
Albert (Bert) Picknell
Head Piano Technician
The Banff Centre
--------------------------------------------------------------------------
Yahoo! Canada Toolbar : Search from anywhere on the web and bookmark your favourite sites. Download it now!
------------------------------------------------------------------------------
All new Yahoo! Mail - Get a sneak peak at messages with a handy reading pane.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/caut.php/attachments/20080515/4122a409/attachment-0001.html
This PTG archive page provided courtesy of Moy Piano Service, LLC