Re: adaptive string quartet

sethares@e... wrote:
> >
> > Ive uploaded an adaptively tuned "string quartet".

paul e wrote:
>
> I wonder if you are aware that John deLaubenfels has been hard at
> work on an adaptive tuning program for several years now:
>
> http://www.adaptune.com/
>
> He defines "pain" as a sum of contributions from vertical dissonance,
> horizontal pitch shifts, overall pitch drift, and optionally other
> factors. Then his program minimizes the total pain.
>
> I'd be interested in comparing your approach with his . . . perhaps
> the main tuning list would be the right place for that . . .

Yes - Ive followed John's work - the methods are similar in many
ways, but also quite different in others. One of the major
differences seems to be that John's approach rests on the
use of familiar (just) intervals which form the basis of
the "pain" calculations (I think I sort of prefer John's
original "spring" terminology!) to define the tables used to adjust the pitches.

In the system used in the "string quartet" now at:

/tuning/files/sethares/Recalled_Opus.mp3

the basic idea is to calculate the sensory dissonance at each
time instant and to move all sounding notes in the direction of
smaller dissonance. By using the parameterization of the
Plomp-Levelt calculations that Ive used before, its possible
to write this as a "gradient descent" algorithm. Hence one of
the strengths of this system is that it can be applied directly
to inharmonic sounds (even though it was not used so here -
since the partials of the violins are hamronically related.)
One of the strengths of John's approach is that it also takes
into account successive notes - all of my algorithms have
been "instantaneous".

You can clearly hear the action of the algorithm in the piece.
The compositional method used here was to start each tone on
its 12-tet location and to adapt the pitches (in real time
using a MAX patch). When the pitches begin close to JI intervals
such as a 12-tet fifth, then the adjustment is only a few cents.
But when the pitches begin at far from JI intervals (such as a
12-tet minor 2), then the sweeps of pitch are quite dramatic.
All of the pitch bending in the piece is done by the algorithm
in real time (though the whole thing was not performed in real
time - I cut and pasted sections together).

I've written about the method, and you can find a copy of
my 1994 JASA paper at:

http://eceserv0.ece.wisc.edu/~sethares/paperspdf/adaptun.pdf

Bill Sethares

Could you share the MAX patch ?!

Vincent-Olivier Gagnon.

-----Message d'origine-----
De : SETHARES@ECESERV0.ECE.WISC.EDU <SETHARES@ECESERV0.ECE.WISC.EDU>
� : tuning@yahoogroups.com <tuning@yahoogroups.com>
Date : 29 novembre, 2001 23:06
Objet : [tuning] Re: adaptive string quartet

>
>sethares@e... wrote:
>> >
>> > Ive uploaded an adaptively tuned "string quartet".
>
>paul e wrote:
>>
>> I wonder if you are aware that John deLaubenfels has been hard at
>> work on an adaptive tuning program for several years now:
>>
>> http://www.adaptune.com/
>>
>> He defines "pain" as a sum of contributions from vertical dissonance,
>> horizontal pitch shifts, overall pitch drift, and optionally other
>> factors. Then his program minimizes the total pain.
>>
>> I'd be interested in comparing your approach with his . . . perhaps
>> the main tuning list would be the right place for that . . .
>
>Yes - Ive followed John's work - the methods are similar in many
>ways, but also quite different in others. One of the major
>differences seems to be that John's approach rests on the
>use of familiar (just) intervals which form the basis of
>the "pain" calculations (I think I sort of prefer John's
>original "spring" terminology!) to define the tables used to adjust the
pitches.
>
>In the system used in the "string quartet" now at:
>
>/tuning/files/sethares/Recalled_Opus.m
p3
>
>the basic idea is to calculate the sensory dissonance at each
>time instant and to move all sounding notes in the direction of
>smaller dissonance. By using the parameterization of the
>Plomp-Levelt calculations that Ive used before, its possible
>to write this as a "gradient descent" algorithm. Hence one of
>the strengths of this system is that it can be applied directly
>to inharmonic sounds (even though it was not used so here -
>since the partials of the violins are hamronically related.)
>One of the strengths of John's approach is that it also takes
>into account successive notes - all of my algorithms have
>been "instantaneous".
>
>You can clearly hear the action of the algorithm in the piece.
>The compositional method used here was to start each tone on
>its 12-tet location and to adapt the pitches (in real time
>using a MAX patch). When the pitches begin close to JI intervals
>such as a 12-tet fifth, then the adjustment is only a few cents.
>But when the pitches begin at far from JI intervals (such as a
>12-tet minor 2), then the sweeps of pitch are quite dramatic.
>All of the pitch bending in the piece is done by the algorithm
>in real time (though the whole thing was not performed in real
>time - I cut and pasted sections together).
>
>I've written about the method, and you can find a copy of
>my 1994 JASA paper at:
>
>http://eceserv0.ece.wisc.edu/~sethares/paperspdf/adaptun.pdf
>
>
>Bill Sethares
>
>
>You do not need web access to participate. You may subscribe through
>email. Send an empty email to one of these addresses:
> tuning-subscribe@yahoogroups.com - join the tuning group.
> tuning-unsubscribe@yahoogroups.com - unsubscribe from the tuning group.
> tuning-nomail@yahoogroups.com - put your email message delivery on hold
for the tuning group.
> tuning-digest@yahoogroups.com - change your subscription to daily digest
mode.
> tuning-normal@yahoogroups.com - change your subscription to individual
emails.
> tuning-help@yahoogroups.com - receive general help information.
>
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

I find this mp3 file interesting precisely because the adaptations
are so glaringly audible. From a performance viewpoint, it leaves
much be desired for precisely the same reason. Didacticly, however,
it is a clear demonstration of just how severe some of the
adaptations from 12-tET to JI are...a clear, non-verbal but visceral
refutation of the "12-tET compromises are insignificant" school of
thought.

--- In tuning@y..., SETHARES@E... wrote:
>
> sethares@e... wrote:
> > >
> > > Ive uploaded an adaptively tuned "string quartet".
>
> paul e wrote:
> >
> > I wonder if you are aware that John deLaubenfels has been hard at
> > work on an adaptive tuning program for several years now:
> >
> > http://www.adaptune.com/
> >
> > He defines "pain" as a sum of contributions from vertical
dissonance,
> > horizontal pitch shifts, overall pitch drift, and optionally
other
> > factors. Then his program minimizes the total pain.
> >
> > I'd be interested in comparing your approach with his . . .
perhaps
> > the main tuning list would be the right place for that . . .
>
> Yes - Ive followed John's work - the methods are similar in many
> ways, but also quite different in others. One of the major
> differences seems to be that John's approach rests on the
> use of familiar (just) intervals which form the basis of
> the "pain" calculations (I think I sort of prefer John's
> original "spring" terminology!) to define the tables used to adjust
the pitches.
>
> In the system used in the "string quartet" now at:
>
>
/tuning/files/sethares/Recalled_O
pus.mp3
>
> the basic idea is to calculate the sensory dissonance at each
> time instant and to move all sounding notes in the direction of
> smaller dissonance. By using the parameterization of the
> Plomp-Levelt calculations that Ive used before, its possible
> to write this as a "gradient descent" algorithm. Hence one of
> the strengths of this system is that it can be applied directly
> to inharmonic sounds (even though it was not used so here -
> since the partials of the violins are hamronically related.)
> One of the strengths of John's approach is that it also takes
> into account successive notes - all of my algorithms have
> been "instantaneous".
>
> You can clearly hear the action of the algorithm in the piece.
> The compositional method used here was to start each tone on
> its 12-tet location and to adapt the pitches (in real time
> using a MAX patch). When the pitches begin close to JI intervals
> such as a 12-tet fifth, then the adjustment is only a few cents.
> But when the pitches begin at far from JI intervals (such as a
> 12-tet minor 2), then the sweeps of pitch are quite dramatic.
> All of the pitch bending in the piece is done by the algorithm
> in real time (though the whole thing was not performed in real
> time - I cut and pasted sections together).
>
> I've written about the method, and you can find a copy of
> my 1994 JASA paper at:
>
> http://eceserv0.ece.wisc.edu/~sethares/paperspdf/adaptun.pdf
>
>
> Bill Sethares