poly aftertouch retuning patent

http://tinyurl.com/jiqk?___aftertouch_retuning_patent

Comments?

-Carl

--- In tuning@yahoogroups.com, Carl Lumma <ekin@l...> wrote:
> http://tinyurl.com/jiqk?___aftertouch_retuning_patent
>
> Comments?
>
> -Carl

<<[0067] A third traditional approach to solving MIDI's pitch
limitations involves the use of Pitch Bending. Pitch Bending is a
MIDI function which allows a note to be "bent" (pitch shifted up or
down), in real time. Pitch Bend MIDI messages cause all the
notes on that channel to be shifted up or down a specified
amount. The pitch bend is specified as a 14-bit number, using
two 7-bit bytes. This results in 16,384 pitch steps being
available. The availability of 16,384 steps can produce pitch
bends which are completely smooth to the human ear (even
though they are carried out in discrete steps). It should also be
noted that Pitch Bend commands are standard within the MIDI
interface.
>>

14 bit pitch bend results in 98304/12 = 8192 steps per halfstep,
not 16384.

<<[0071] The proposed invention implements notes which do
not lie on the traditional Western semitone scale using the
following method: First, a standard note lying on the Western
scale is played. Then, a Polyphonic Aftertouch signal is
immediately sent to shift the pitch of the standard note to
transform it into a non-standard note. The Polyphonic Aftertouch
message follows immediately behind the Note On message
and pitch value creating the standard note. Thus, the standard
note only sounds for about 960 microseconds. The shortest
temporal resolution of the human ear is approximately 2-3
milliseconds. As a result, the human ear does not hear the first
standard note, since it does not sound for a sufficient length of
time. The human ear only perceives the pitch shifted note. >>

MIDI is serial and it's clunky; I'm afraid that the above claim is
indefensible in practice...

All other qualms aside, using aftertouch has been tried before
for various tuning purposes, but the question remains in this
patent : how, specifically, is the retuning accomplished? It
doesn't really say. It implies something like sysex because
apparently every pitch can be retuned. Is the suggestion really
that new hardware should interpret aftertouch messages as
tuning messages? How is this an improvement over sysex? Is
the suggestion to change the MIDI protocol? A tuning box would
apparently be built. What messages does it send to retune?
Sorry, I must have missed something.

Aaron

>http://tinyurl.com/jiqk?___aftertouch_retuning_patent
>
>Comments?

Just a note -- this is a patent application. So if anyone
knows of any prior art here, he should write to the patent
office.

-Carl