[pianotech] Stretch in tuning

[David Love]

There is another important aspect to stretch that is worth considering.
While the piano itself may tell you something about the amount of stretch it
wants within a one or two octave range, one function of stretch is to
balance the disparity that can exist between the perception of harmonic
versus melodic intervals.  This is more of a psychoacoustic phenomenon but
real none the less.  We all are familiar with this phenomenon.  Play two
notes two (or three) octaves apart harmonically (together) and the
perception of their "intuneness" will be somewhat different than if you play
them melodically (separately).  Played melodically the upper note will want
to be sharper (or the lower note flatter depending on how you look at it)
than when played harmonically.  

Thus, choosing some amount of stretch for the piano can depend, to some
degree, on how you want the upper part of the piano, say, to sound when
notes are played melodically as well as harmonically.  Those more sensitive
to melodic relationships will tend to want a greater amount of stretch at
least in the upper (and lower) part of the piano.  Having the tuning curve
get a bit steeper in the extremes can prove a benefit in this case.  One
advantage of the ETD's is that the shape of stretch curve can be modified to
accommodate this and the stretch will be graduated uniformly to blend in
with the midrange parts of the piano so that there is smooth transition
between sections.  The RCT does this very easily by being able to
essentially click and drag the settings in any one area.  The Verituner
allows you do this by preprogramming the amount of stretch in each octave
via the weighting of a particular type of octave to be used as the reference
and guide.  The SAT has the DOB (Double octave beat setting) that allows you
to increase or decrease the amount of stretch on the fly.  I'm not familiar
with Tunelab but I assume it can perform a similar function. 

Ultimately, then, stretch *is* a matter of opinion though there are certain
practical limits.  Customizing a tuning style with respect to stretch can be
an important part of tuning whether that is done aurally or electronically.
However, there are certain benefits to doing this electronically both in
terms of control, predictability, repeatability, not to mention the ability
to set a particular tuning curve shape and then be able to test it by
sampling various octaves to see if it actually works to your liking.    

David Love
www.davidlovepianos.com


-----Original Message-----
From: pianotech-bounces at ptg.org [mailto:pianotech-bounces at ptg.org] On Behalf
Of Ron Koval
Sent: Tuesday, May 29, 2012 7:09 AM
To: pianotech at ptg.org
Subject: [pianotech] Stretch in tuning


"Just
the question of the stretch that has always been a bone of contention
amongst my peers. It has always been a 'matter of opinion', which is not
quite perfect tuning IMO.
Best regards,
David"                                                         

Ok, Let's take a little deeper look into this topic.  The common
misconception seems to be that by using one of the tests to set a particular
octave match (2:1, 4:2, 6:3 etc) we can set the "perfect octave".  We focus
our attention and our ears to rule out any  other information except one
partial from the lower note and one partial from the upper note. I did a
little experiment on two pianos, a Baldwin studio upright(243) and a
Steinway D.  I set 2:1, 4:2 and then 6:3 octaves between A4 and A3, then
measured the fundamental of A3.  The number represents in cents how far the
octave was stretched. Then I set the octave justby ear as a unison would be
tuned to determine the "sweet spot", or ideal placement.
Baldwin 2:1 -.6   4:2 -2.9  6:3 -6.7  "ideal" -1.3         Steinway 2:1 -.2
 4:2 -1.3  6:3 -3.1  "ideal" -2.0 You can see that the "ideal" octave is
some sort of combination of the "pure" single-partial octave types.  Lesser
pianos may have a wider range between the octave types, as well as a
difference  between which partial matches is the location of the "ideal
octave".  Likewise, throughout the scale  the partial choices change.  I
find that some pianos really like a 6:3 match for a section of the bass, or
a 4:1 double octave match for certain portions of the treble...  Most of my
tunings end  up using a mix of multiple partial matches to control the
stretch. Ron Koval  Chicagoland

 		 	   		  =