From: Patent Tactics, George Brock-Nannestad
it is a healthy sign of the all-pervading curiosity of musical man that all
avenues of research are explored, and modern and in particular fast computer
technology permits us to create definitions for filters (in the most general
sense) that will distinguish even individual piano tones in a chord. This
done in many places now, and I happen to know a bit about the work by Gerhard
Widmer and his group in Austria. Basically, it is a three-pronged approach,
where the individual sounds and their significance as regards timing and
strength are determined hand-in-hand with discovering how the musical mind
hears such phenomena and how performance should be to sound "nice" or good
The only sad side effect of such research is that the magic disappears, and
in particular if precise goals are set for a good-sounding and selling manner
of performance, and if training in conservatories aim for just that by
technical feedback. On the other hand, perhaps we can then rid ourselves of
the need for tightly fitting dresses or strange haircuts.
It is entirely feasible to create a file that will control a piano out of the
recording of a sonic event - Jonathan Berger has done that with the sound
from the famous Brahms cylinder. Obviously he has not proposed that we are
hearing Brahms playing - it is a modelling, and we may learn something else
than by merely listening to a noise-reduced rendering or slightly more
complex, which I have done - the sound after counteracting the influence of
the recording system.
The only unethical about such approaches would be to claim that we now have
the truth, that this is the way that a particular artist played and sounded.
It is a reconstruction, we may learn something, and the sound may well
compete with modern recordings, because there will be some out there who
prefer the reconstruction. Alas, if it also comptetes with modern live
performances, then concert life as we know it will die out ("Death of a
Music"). Because it will mean that just like singers who cannot perform any
more without a microphone at the ear or cheek, we shall have pianists playing
through a processor. The grand piano with piezoelectric fine-tuning of each
string has already been patented, and I see no reason why there cannot be an
electronic Zwischensetzer placed between the keys and the piano action to
fine-adjust the actual timing of the hammer blow in accordance with the newly
developed rules for good performance. The sound would still be live from a
piano, but the pianist's ability will be extended.
The crucial question is very well put in the NY Times article:
So is he [Dr. Goebl] saying that Dr. Walker's track isn't authentic?
> "There you have to go into the philosophical domain," Dr. Goebl replied. "A
> recording is just an acoustic document of what took place."
> In other words, a recording isn't authentic, either. It is also at a
> remove, or two or three, from the original performer, and it is also
> affected by the decisions of the engineers who helped create it.
----- here we may say that the simpler and more transparent a
recording/reproduction method is, the easier it is to compensate for some of
the technical shortcomings and to prove that this is all that has been done.
And so, the replay of original recordings is still closer to the source than
the present experiments.
----- however it does not help at all to misuse the Turing test, as in:
> The final criterion for any such reproduction is the rather
> imprecise "Turing test" of artificial intelligence: that is, whether it can
> make the listener think he or she is hearing a person rather than a
----- because here we do not need "a" person, but "the" person. And if we
only think that because we are told, then that is unethical.
Thanks to Dick Spottswood for having distributed the original text.
P.S. The hums and grunts of Glenn Gould or Sergiu Celebidache can obviously
be distinguished and added in suitable proportion.