[CAUT] F..riction

[David Love]

I don't know, I have heard the claim about .1 grams and have erroneously
made that claim myself in the past but I suppose it depends on what you
mean.  You can easily set strike weight and front weight curves to .1 grams
since that's what the scales are calibrated to but whether that translates
to .1 gram tolerances in the BW (or UW and DW) is problematic as you've
outlined.  I have on many occasions surveyed actions using UW and DW,
calculated the BW and modified the FWs to achieve a uniform BW.  That will
not necessarily yield, as you know, a perfectly smooth front weight curve
even if you start with the most accurately set SW curve because there are
other variables that will influence both your ability to take accurate
measurements (such as friction in various parts of the key stroke and from
various sources) as well as variations in the overall ratio itself from note
to note.  Since the other method of setting a smooth front weight curve
using an "average" ratio produces some unevenness as well due to small
variations in the ratio from note to note it's not clear to me which is
better, or maybe more appropriately, which is worse.  Both methods, if done
carefully, will produce a very uniform feeling action.  Even using the
survey and set method, the general shape of the FW curve will be still be a
fairly smooth curve and small differences in inertia are not likely felt any
more than letting the BW float.  Even when the sharps come in at a
completely different action ratio and therefore two different front weight
curves are produced (whichever way you do it--unless you modify the ratios
between the two), that difference between sharps and naturals is masked by
the differences in the lever lengths themselves (between sharps and naturals
that is).  When you combine that with the fact that pianists don't play the
piano at the very end of the keys where these measurements are taken from
but rather from anywhere from the tip of the key to the fallboard where the
action ratio will change, and very dramatically, it makes one wonder just
exactly what pianists are feeling when they say the piano feels very "even".


David Love
www.davidlovepianos.com

-----Original Message-----
From: caut-bounces at ptg.org [mailto:caut-bounces at ptg.org] On Behalf Of David
Stanwood
Sent: Saturday, December 04, 2010 1:20 PM
To: caut at ptg.org
Subject: Re: [CAUT] F..riction

Hi David,

I've been an avowed balance weighter since 1990 I can offer the 
opinion that it is a squirrely value because it is derived from two 
squirrely measurements.. down and up weight.. both prone to 
misinterpretation, error, and variation of method.  Balance weight 
itself assumes equal friction in either direction.  Emphasis on 
ASSUME.  This adds to the nutty nature of it all.  There is no 
question that using balance weight to separate out friction is a 
useful tool in touch designing.

I've long felt that it is really inaccurate to set lead weights using 
such unreliable measures as down weight, up weight, and balance 
weight.  Using my patented method of calculating front weights that 
assume average levels of ratio eliminates any reliance in 
interpretation of touch weight.  This allows for more accurate and 
consistent ways of setting key leads, namely using a digital scale.
Down Weight and Up Weight...... let'm float.. they always will.

Accuracy to 0.1 gram is negated by the inability to repeat 
measurements to that accuracy.. it just doesn't happen.  Yes we use 
scales accurate to 0.1g and this is appropriate.. I've seen techs who 
use scales accurate to 0.01g... it's a waste of time.

I know there is a comment in the wonderful documentary film "Miracle 
in a Box" which mentions that the Callahan Shop in Alameda California 
builds everything to a tolerance of 0.1g... I've never heard John 
Callahan make the claim and I think the film maker was just 
misinterpreting the use of scales that are accurate to 0.1g.  That's 
Hollywood.

I'm all for spending time refining strike weight and front weights to 
reasonable tolerances.  Fixating on making accurate down weight 
results in diminishing returns very quickly simply because down 
weight is never felt by the pianist because it never goes through let 
off.  It is a much less relevant value than many of us might believe. 
Getting it within acceptable ranges is good enough for me....

Hope this helps.

Regards,

David Stanwood

>One thing it does suggest (assuming no measurement error) is that 
>setting balance weight by taking up and down weight measurements and 
>adjusting lead weights accordingly has its limitations.  The claims 
>of some touch designers that their actions are balanced to the .1 
>grams is obviously limited by the ability to determine accurately 
>the actual UW and DW in a given system (when does the key really 
>start to go down and when does it actually come up) and the 
>potential for variations in friction to skew the data and thus the 
>resulting action taken to achieve those uniform balance weights. 
>Working from the other direction, a calculated front weight curve 
>based on a given strike weight curve and a constant action ratio, is 
>also limited by the actual accurate executions of the contributing 
>levers themselves to achieve a precise action ratio through the 
>scale.  While attempting to set a very precise standard will likely 
>yield better results than targeting a less precise standard, it is 
>somewhat unclear as to where the point of diminishing returns falls. 
>Just a thought.
>
>David Love