<!doctype html public "-//W3C//DTD W3 HTML//EN">
<html><head><style type="text/css"><!--
blockquote, dl, ul, ol, li { padding-top: 0 ; padding-bottom: 0 }
--></style><title>Re: SIREN sound synthesis vs.
CSound</title></head><body>
<div>At 09:56 AM 11/22/99 -0800, you wrote:<br>
</div>
<blockquote type="cite" cite>This is an interesting thread.<br>
<br>
First, Kyma is still alive and well, see
http://www.symbolicsound.com/;<br>
they use VisualWorks Smalltalk and do require the Capybara<br>
synthesizer/DSP hardware. It's a pretty wonderful system.<br>
<br>
Second, I feel quite frustrated that I haven't gotten around to
working<br>
more on sound synthesis within Siren (I got all caught up in a<br>
multimedia database project, and am now working on sound
spatialization<br>
for another project [all in Squeak]). John Maloney's synthesis
framework<br>
and the Siren representation would make a perfect marriage for a<br>
powerful and flexible real-time system that would put Csound to
shame<br>
(its language model is IBM 360 assembler).<br>
<br>
The two most important references to look into are (1)
SuperCollider<br>
(James McCartney's synthesis language that can be considered a
dialect<br>
of Smalltalk) -- see http://www.audiosynth.com, and (2) the CNMAT
SDIF<br>
sound description interchange format -- see<br>
http://cnmat.CNMAT.Berkeley.EDU/SDIF/. At the recent Int'l
Computer<br>
Music Conference in Beijing, I promised to make a Smalltalk reader
for<br>
SDIF format so that I can import SDIF into Siren. The next obvious
step<br>
would be to implement synthesis methods that correspond to the SDIF
base<br>
types.<br>
<br>
Who wants to help? Maybe we should start a Squeak/Sound mailing
list...<br>
<br>
--<br>
<br>
stp<br>
Stephen Travis Pope<br>
stp@create.ucsb.edu -- http://www.create.ucsb.edu/~stp<br>
</blockquote>
<div><br>
<br>
<br>
<font face="Courier New">I vote that if there is an interest in music
projects we start a new Sequeak mailing list. It may make sense for
this thread to be wider than music and include multimedia, graphics,
animation, video, etc., as ultimately all these should be working with
each other anyway.<br>
<br>
I agree that sound synthesis is very important and that Squeak can
provide a framework in which significant improvements could be made.
You mentioned Super Collider and SDIF, one may also want to look at
Kyma (<font color="#0000FF"><u>www.symbolicSound.com</u></font>),
Synth-Builder (<font
color="#0000FF"
>http://www-ccrma.stanford.edu/CCRMA/Software/SynthBuilder/SynthBuild<span
></span>er.html</font>), and Max / MSP (<font
color="#0000FF"><u>http://www.cycling74.com/index.html</u></font>).<br
>
<br>
In addition to Sound Synthesis I think that high-level composition is
another major area in which Squeak could play a winning role. This is
what my own DMIX environment is all about. For those who don’t know
about it, here is a short overview.<br>
<br>
DMIX (<font
color="#0000FF"><u>http://www.research.ibm.com/mathsci/cmc/dmix.htm</u
></font>) is a novel environment for musical composition, interaction,
and performance. Some simple examples of cool things Dmix can do:<br>
<br>
</font>v<x-tab> </x-tab><font face="Courier New">The
graphic piano roll editor can interact with real-time controls
(real-time editing), be programmed on the fly’ by Slapping Smalltalk
statements, or perform more complex edits by Slapping Dmix modifiers
[see refs. below].<br>
<br>
</font>v<x-tab>
</x-tab><font face="Courier New">Music processing tools (Dmix
Modifiers) such as functions, filters, or interpolators, can be both
created and configured directly from musical content, or via graphic
editors.<br>
<br>
</font>v<x-tab>
</x-tab><font face="Courier New">Quill, the textual algorithmic
music-programming environment, can link with graphic objects, such as
Modifiers, that integrate with your program code.<br>
<br>
</font>v<x-tab> </x-tab><font face="Courier New">Echo, is a class of
real time algorithms that interactively process MIDI input. They can
be used for composition, in performance, or improvisation.<br>
<br>
</font>v<x-tab> </x-tab><font
face="Courier New">Shadow is a framework for building interactive
performances, includes score tracking and tempo matching. External
MIDI events can trigger events that manipulate the Dmix
environment.<br>
<br>
More generally, Dmix was designed to fill what I perceived as a void
left by existing tools, in particular, by better supporting the
creative workflow. Dmix simultaneously supports and integrates several
different approaches that have often been treated as contradictory,
such as direct graphic manipulation, improvisation / interaction, and
off-line algorithmic composition. For example, by allowing music
processing tools to be created from the music itself, musical ideas,
motifs, and concepts can easily be transformed into abstract tools
used for developing and structuring the work. I have tried to apply
this approach on all levels of the Dmix system. As a result, the
distinction between composition, improvisation, and performance gets
somewhat blurred in Dmix, as it should be.<br>
<br>
Much of the flexibility, object interoperability, and overall
multi-facted aspects of Dmix are facilitated through a technique
I call Slappability (see refs below). To the user this seems like an
extension of Drag and Drop. This is a natural for Morphic.<br>
<br>
* * *<br>
<br>
Returning to Stephen’s call for help. I am beginning a port of DMIX
to Squeak from VisualWorks. I am wondering (1) if there are people in
the squeak community that would be interested in DMIX and (2)
interested enough to get involved in some capacity from raw porting,
bug-finding, documentation updates, feedback both musical and
Code/Smalltalk/Design, and so on.<br>
<br>
<br>
References:<br>
<br>
</font>(1)<x-tab> </x-tab><font
face="Courier New">For an overview of Dmix see:</font> <font
face="Courier New"
color="#0000FF"><u>http://www.research.ibm.com/mathsci/cmc/dmix.htm</u
></font><u>.<br>
</u>(2)<x-tab> </x-tab><font
face="Courier New">Most complete publication: Oppenheim, D. (1996)
"DMIXA Multi Faceted Environment for Composing and Performing
Computer Music." Computers and Mathematics with Applications,
Volume 32, Issue 1, pages 117-135, 1996.<br>
</font>(3)<x-tab> </x-tab><font face="Courier New">Dmix documentation
can be browsed online at:</font> <font face="Courier New"
color="#0000FF"><u
>http://www.research.ibm.com/mathsci/cmc/dmixdocs/index.htm</u></font><font
face="Courier New">.<br>
</font>(4)<x-tab> </x-tab><font
face="Courier New"> For a PowerPoint presentation overview on
Slappability see:</font> <font face="Courier New"
color="#0000FF"><u
>http://www.research.ibm.com/mathsci/cmc/slappability/index.htm</u></font
><font face="Courier New">.<br>
</font>(5)<x-tab> </x-tab><font face="Courier New">The paper
“Slappability: A New Metaphor for Human Computer Interaction."
Can be found at:</font> <font face="Courier New"
color="#0000FF"><u
>http://www.research.ibm.com/mathsci/cmc/do_papers.htm<br>
</u></font>(6)<x-tab> </x-tab><font
face="Courier New">A more formal and general paper called “Direct
Combination” appeared in SIG-CHI 99.<br>
<br>
BTW -- a small note on my DMIX and Stephen’s SIREN may be in place
here. My work at CCRMA (Stanford) was greatly inspired by Stephen
Pope's seminal work (I think it was called Mode at the time). We had
great discussions, shared many ideas, and were both active in the
smOKe music representation group. Whereas Mode was geared more towards
synthesis and off-line composition, DMIX was geared more towards
real-time composition, interaction, and performance all through MIDI
(limitations of the time…). Although there is some overlap between
the two systems, I feel they play complementary roles, and thus their
combination is even more interesting.<br>
<br>
<br>
</font><x-tab> </x-tab>Danny
Oppenheim<br>
<br>
<br>
<br>
</div>
<div>---</div>
<div><br></div>
<div>Daniel V. Oppenheim </div>
<div
> <span
></span
> <span
></span
> <span
></span
> <span
></span
> <span
></span
> <span
></span
> <span
></span
> <span
></span
> <span
></span>
</div>
<div><x-tab>
</x-tab>Computer Music Center</div>
<div><x-tab> </x-tab>IBM
T.J. Watson Research Center<x-tab> </x-tab>
phone: (914) 945-1989</div>
<div><x-tab> </x-tab>P. O.
Box 218 (or Route 134)<x-tab>
</x-tab> fax: (914) 945-3434
</div>
<div><x-tab>
</x-tab>Yorktown Heights, NY 10598<x-tab>
</x-tab> www.research.ibm.com/music </div>
</body>
</html>