I'm baack :) for a while anyhow

Well, the natural rotation of hobbies being what it is, I haven't had
my fingers in microtonality since early November of 2001, at the El
Paso MicroHoot. But ... my fingers are still itching for this, and I
have a brief period -- about a month -- before I rotate hobbies again
and spend three weeks in Neuro-Semantics land
(http://www.neurosemantics.com/Training/2002_International_Conference.
htm if you're interested in that sort of thing).

Sooo ... what am I up to?

1. I'm expanding the code that produced "When Harry Met Iannis"
(www.borasky-research.net/HarryIannis.htm)in an attempt to make the
resulting music more interesting. I'm going to add a random volume
level to the random walk and I'm going to expand the chord space to
the 13-limit tonality diamond.

2. I'm trying to get the rights to port David Cope's LISP-based
algorithmic analysis and composition software to Linux. As far as I
know it currently only works on Macs, and the CDs that come with his
books containing all that wonderful source code are unreadable on my
Windows box.

3. I think there's finally an open-source driver for my sound card
for Linux, so I may finally be able to get rid of the Windows 2000
partition on my big system.

4. I'm looking to code up ancient Hebrew, Sumerian and Greek musical
ideas with the Cope software (that's why I want the port).

--
Ed Borasky
znmeb@aracnet.com
http://www.borasky-research.net/

Hello Ed,

>2. I'm trying to get the rights to port David Cope's LISP-based
>algorithmic analysis and composition software to Linux. As far as I
>know it currently only works on Macs, and the CDs that come with his
>books containing all that wonderful source code are unreadable on my
>Windows box.

That would be great. I had just read a column of Douglas Hofstadter
about it:
http://www.acm.org/membernet/stories/invisible_future.html

Manuel

On Tue, 21 May 2002 manuel.op.de.coul@eon-benelux.com wrote:

> Hello Ed,
>
> >2. I'm trying to get the rights to port David Cope's LISP-based
> >algorithmic analysis and composition software to Linux. As far as I
> >know it currently only works on Macs, and the CDs that come with his
> >books containing all that wonderful source code are unreadable on my
> >Windows box.
>
> That would be great. I had just read a column of Douglas Hofstadter
> about it:
> http://www.acm.org/membernet/stories/invisible_future.html
>
> Manuel

1. David Cope thinks this is a great idea. :-) I'm going to port the
LISP stuff to an Open Source Linux LISP, probably CMUCL. The GUI stuff I
will leave for later, and I'm going to use "sfront" as the synthesis
engine, so what I do will be portable to Windows assuming availability
of a Common Lisp / CLOS package for Windows.

2. I just ordered the Cope / Hofstadter book, so I will have to wait
until it arrives before porting any of that code. I do have a utility
now to read Mac CDs on my Windows box so I can now extract the source
from previous Cope CDs.

3. Speaking of "sfront", the latest version is 0.80. They put some more
of it under GPL, including, I think, the wonderful textbook on MPEG 4
Structured Audio. The "sfront" manual has always been GPL, IIRC.

Since you did a portable GUI for "scala", if you want to take on the GUI
piece of this project, an area where I have very little expertise and
even less desire to gain some, I'd be glad to hand it off to you.
Whatever we end up with will have to run on Macs, I think :).
--
M. Edward Borasky

znmeb@borasky-research.net
http://www.borasky-research.net/HarryIannis.htm

Stand-up comedy -- because man does not live by dread alone.

>>2. I'm trying to get the rights to port David Cope's LISP-based
>>algorithmic analysis and composition software to Linux. As far as I
>>know it currently only works on Macs, and the CDs that come with his
>>books containing all that wonderful source code are unreadable on my
>>Windows box.
>
>That would be great. I had just read a column of Douglas Hofstadter
>about it:
>http://www.acm.org/membernet/stories/invisible_future.html

I saw Hofstadter give a talk/demo on EMI in 1996. Most of what he
said was either obvious or wrong.

"I was truly shaken. How could emotional music be coming out of a program
that had never heard a note, never lived a moment of life, never had any
emotions whatsoever?"

Yep, this kind of wrong thing. Surprising coming from Hofstadter, a man
who for any other aspect of thought, including Turing tests, has no
problem admitting that computers will do it, and has even done important
work in getting them to!

EMI is just an expert system. Nothing interesting about it, as far as
I know. Rather I suspect it is like most expert systems; it took 25
years to build and is capable only of a relatively fixed set of behaviors.

Not to be the nay-sayer. I'd still like to see the LISP! It could be
very useful for translating music into other tunings, depending on how
notes are internally represented. . .

-Carl

On Tue, 21 May 2002, Carl Lumma wrote:

> I saw Hofstadter give a talk/demo on EMI in 1996. Most of what he
> said was either obvious or wrong.
>
> "I was truly shaken. How could emotional music be coming out of a
> program that had never heard a note, never lived a moment of life,
> never had any emotions whatsoever?"

Well ... I have Cope's CD of pseudo-Beethoven/Mozart etc. I am
singularly unimpressed, but then, I'm not all that interested in
generating pseudo-anyone. I'm looking for tools to assist musical
analysis.

> Yep, this kind of wrong thing. Surprising coming from Hofstadter, a
> man who for any other aspect of thought, including Turing tests, has
> no problem admitting that computers will do it, and has even done
> important work in getting them to!

Considering the kind of compute horsepower involved in, say, today's
chess engines, I'm not at all convinced that *I* will live to see a
computer pass the Turing test.

> EMI is just an expert system. Nothing interesting about it, as far as
> I know. Rather I suspect it is like most expert systems; it took 25
> years to build and is capable only of a relatively fixed set of
> behaviors.

EMI is an approach to music analysis and synthesis. What I am looking
for is mostly analysis; if there's something better than EMI for
analysis I would use it instead of EMI.

> Not to be the nay-sayer. I'd still like to see the LISP! It could be
> very useful for translating music into other tunings, depending on how
> notes are internally represented. . .

Well, there are four books, three of which come with CDs. I have the
first three (two CDs) and just got a utility to read Mac CDs the other
day. There are several variants of what's in the books: "EMI", "SARA"
and "ALICE" at least. I don't know if the full EMI code or all of Cope's
databases even exist. Given that the output is MIDI, I suspect there are
better tools if all you're interested in is re-tuning existing music.
There's probably millions of lines of MIDI out there and only thousands
of lines of Cope's LISP-coded music.

I'll let everyone know what I come up with; it will probably be this
weekend before I attempt running some examples with the LISP code on my
Linux system. I want to get the "HarryIannis" upgrades done first and
release the whole enchilada under GPL and send it off to the "sfront"
people to use as an example. This includes redoing the Power Point piece
in a Linux-readable format; the background theme I used only works with
Internet Exploder.
--
M. Edward Borasky

znmeb@borasky-research.net
http://www.borasky-research.net/HarryIannis.htm

Stand-up comedy -- because man does not live by dread alone.

--- In tuning@y..., Carl Lumma <carl@l...> wrote:

> EMI is just an expert system. Nothing interesting about it, as far as
> I know. Rather I suspect it is like most expert systems; it took 25
> years to build and is capable only of a relatively fixed set of behaviors.

You don't anything too elaborate to produce the Hofstadter Effect, since similar comments have been made about outout from high order Markov chains.

Ed wrote:

>Since you did a portable GUI for "scala", if you want to take on the GUI
>piece of this project, an area where I have very little expertise and
>even less desire to gain some, I'd be glad to hand it off to you.
>Whatever we end up with will have to run on Macs, I think :).

If it's not a lot of work, I will consider it. For Macs I
might only be able to help once Gimp runs on OS X. Success,

Manuel

>Well ... I have Cope's CD of pseudo-Beethoven/Mozart etc. I am
>singularly unimpressed, but then, I'm not all that interested in
>generating pseudo-anyone. I'm looking for tools to assist musical
>analysis.

Hofstadter played samples for the audience in several trials for
Bach, Mozart, and Chopin. The audience was to vote to identify
the originals among several EMI compositions in each trial. It
was a bit of a laugh because it was completely obvious to everyone
in the audience in all cases. Many people felt the Bach came out
the best. I felt the Chopin. In all cases the endings were a
problem. Then we were informed that the examples given were selected
by Cope from hundreds of EMI runs. Not very impressive.

I haven't heard the CD. I'm not interested in copying either, per
se. EMI also has modes that turn out "original" works. Hofstadter
played one, but I don't remember it.

Yes, something like EMI could be very useful in musical analysis.

>Considering the kind of compute horsepower involved in, say, today's
>chess engines, I'm not at all convinced that *I* will live to see a
>computer pass the Turing test.

I don't know how old you are, but I'm 25 and have little doubt that
I'll see it before I retire.

Chess is a hard problem when approached in the simplest way. The
narrowness of the current approach is alarming. The fact that humans
can still compete with the state of the art brute-force stuff tells me
that the game is easier. I expect an algorithm based on an analysis
of the game to be produced before long and just beat everything, with
minimal searching. We can take this off-line if you'd like to discuss
it further.

>Well, there are four books, three of which come with CDs. I have the
>first three (two CDs) and just got a utility to read Mac CDs the other
>day. There are several variants of what's in the books: "EMI", "SARA"
>and "ALICE" at least. I don't know if the full EMI code or all of Cope's
>databases even exist. Given that the output is MIDI, I suspect there are
>better tools if all you're interested in is re-tuning existing music.

Maybe so, but the problem with just re-tuning music is that different
tunings require different structures. But if EMI creates any kind of
coherent higher-level abstractions of structures like you'd expect it
to, it might be possible to switch the tuning there above the point
where it would break anything... just a thought.

>There's probably millions of lines of MIDI out there and only thousands
>of lines of Cope's LISP-coded music.

That's why I'd rather see the LISP!

Actually, if it's thousands of lines of LISP, that's too much for it to
be much good.

-Carl

Gene wrote...

>>EMI is just an expert system. Nothing interesting about it, as far as
>>I know. Rather I suspect it is like most expert systems; it took 25
>>years to build and is capable only of a relatively fixed set of behaviors.
>
>You don't anything too elaborate to produce the Hofstadter Effect, since
>similar comments have been made about outout from high order Markov chains.

I think that's a different Hofstadter. Maybe Douglas' father?

Markov chains keep coming up. I should probably learn what they are.

-Carl