etd's and ears

[David Love]

I think I understand what David Anderson was saying and to some degree I
think he's right.  Assuming we are not talking about institutional tuning or
other situations where the effort/reward ratio dips too low to concern
ourselves with 100% outcomes and where the ETD calculated tuning is an
appropriate use, if we're comparing aural tuning to a pure calculated
tuning, by definition, the aural tuning is more "artistic", it requires
judgment, the calculated tuning doesn't, at least not once the programming
is done.  Arguing which is "better" is not something I want to get into but
I will say that they are clearly different.  A calculated tuning will almost
certainly be better than a tuning by a non-skilled aural tuner.  However, a
skilled aural tuner will more consistently compensate for anomalies in the
piano that the ETD can't really make decisions about.  I say this as someone
who uses an ETD though more and more I use it only for direct interval
tuning, i.e. it's set to the coincident partial in question and I don't
bother to calculate anything and I check everything, I mean everything,
aurally as I go.   Of the thirty years I've been doing this I spent probably
20 tuning aurally and the last 10 using a device.  I've gone through several
different ones never really being that satisfied with any of them for
calculated tunings.  Some claim greater prowess than others.  I can say that
the complaints I have received about tuning (at least the ones I knew about)
were always associated with calculated tunings in which I did not tune with
aural checks.  I even lost a couple of regular customers during a period in
which I was experimenting with calculated tunings on a "state of the art"
machine.  They thought my tuning skills had deteriorated and (interestingly)
associated it with the change to using an ETD (so much for the theory that
people only hear solid unisons).  I gave up using the ETD for unchecked
tuning after that.  Since then I have gone back to using the ETD to
basically set the temperament-it's simply faster and overall more accurate
for that, and then I go to direct interval tuning and tune like I always did
going out from the temperament--bass first.  By doing that I can compensate
for areas in the piano that do not fall into the way the ETD would calculate
the tuning curve, there are always those areas, and more accurately choose
where in the piano to switch (or fudge) octave styles. In general, I find
that I do not stretch the tuning nearly as much as ETD's tend to.  My C88
tends to come in about 25 - 28 cents sharp whereas a calculated tuning would
likely come in somewhere around 40 cents.  Bass tunings tend to be less
stretched as well.  My own style of tuning, I find, leaves the piano with a
greater sense of sonority, i.e. unity and swell.  Those customers for whom
I've done AB comparisons tend to agree.

Unfortunately, I've come to the conclusion that the piano is simply too
irregular an instrument in terms of how it produces harmonics to hope for a
calculated tuning that is really spot on through the entire scale, though on
certain pianos it can get close.  The best choice, in my view, is to use a
combination as the ETD does some things very well:

1.  Serves well as a means of visual verification or as a guide which can be
verified aurally
2.  Provides stress reduction and speed on poor quality pianos or
institutional pianos
3.  Fast and accurate for temperament set up (all types)
4.  Good for tuning the extremes especially when pitch recognition is
compromised
5.  Performs quick and accurate pitch corrections

Beyond that, the devil is in the details and the art of tuning lies in
judgments made in areas or across sections of the piano that defy
calculation.  In that sense, I think David Anderson has a point.  

David Love
davidlovepianos at comcast.net
www.davidlovepianos.com