[LAD] Music, Undecidability, and the tiling problem (was Re: update: OT-ish: realtime 2d placement algorithms :-/)

Jens M Andreasen jens.andreasen at comhem.se
Fri May 28 10:24:49 UTC 2010


On Thu, 2010-05-27 at 13:54 -0500, Charles Henry wrote:
> I would like a computer to be able to say, "This would sound good, if
> I were a human".  Better yet, I'd like the computer to describe it to
> me in numbers that I myself could not calculate.

The question then becomes: Do androids really dream of electric sheep,
and if so, why would we care?

       - Computer, explain: "Like tears in rain" ?!


Ravi Shankar in one his masterclasses (televised here recently) talked
about hitting an inner melancholic string. If you can find that and get
it to resonate so you can feel it, then others will feel it too. This is
to a certain degree bound to be dependent on the cultural background of
the involved parties, what kind of musical vocabulary they possess - not
all elements of music can easily be explained by heartbeats or simple
Pythagorean relationships, and just like with any other language you may
hit an initial language barrier where it doesn't matter how beautiful a
poem or well written a scientific article you present; if the audience
do not recognize the words it will be in vain.

As an aside, this is not in any way to say that you need to understand
the /lyrics/ of a song to get the underlying musical structure -
certainly not if the structure is very direct and efficiently exposed. 

I believe that just about anybody on this planet would be able to
understand that these tongue wrestling Karelian chicks easily could eat
up Pythagoras in a heartbeat:

Värttinä - Käppee [*]
 http://www.youtube.com/watch?v=t4YRT-MZR_U



Keith Emerson, while working on the Trilogy album, also talked about
searching for a kind of resonance with inner emotions - doing it by
actually living them out on the piano - and then how he would use formal
composition only at a later stage, as a touch-up to make ends meet. So
here again we have an expert in the field in favor of using your own
mind and mental abilities as the primary tool. It makes perfect sense to
me - because, hopefully, the intended audience will be a lot more like
people like yourself rather than, say, people from Mars.

Emerson - Composing The Endless Enigma
 http://www.youtube.com/watch?v=abVhSCzByw8

Judging by the number of ambitious ELP covers on YouTube - not all of
them equally interesting - he did manage to hit a note within quite a
few, and still do. As an example, here a really cool French acoustic
jazz rendering with strings. No Hammond organs was stabbed to death
during this performance:

Jad & Den - Trilogy
  http://www.youtube.com/watch?v=xNxoIDPoURY


So there you have it, at the end of the day, when the last note has been
written down and  gone to print, as a composer - human or inhuman - you
would still end up being at the complete mercy of the musicians who are
eventually going to play your music. It is their sound judgment while
being in the process of making what was yours into their own, that
decides whether the music will finally reach out to a real human
audience.

If they then in that process are not listening for that inner resonance
Shankar talked about, then that just won't happen. You will instead end
up with something only a refrigerator could enjoy. And even if computers
one day, somehow, might be programmed to elegantly dream of electric
sheep, it will still be a poor substitute for us.

Because we are not.

/j


* For an alternative rendering of "Käppee", try: Бони НЕМ :)
  http://www.youtube.com/watch?v=to12RY_p2Rs





More information about the Linux-audio-dev mailing list