Fugues

[ted]

Here are a few from a set of a thousand fugues I wrote twelve years ago. I couldn't be bothered learning them all.

[nicco]

lol this is a joke, right?

[piano121]

This is not a joke. Those are his compositions as midi files. By the way, composition wise it sounds not bad at all. Wish to hear it on a real piano.

BTW, did you realy composed a set of 1000?

[pianogeek_cz]

Sounds nice (better, would sound nice on a real piano/organ). BTW, what's that strange harmonic change happening in the first one around 0:55?

[ted]

Okay ! I'll tell you all about it.

They are composed and played using Amiga game sound samples by a Basic programme of a couple of dozen lines I wrote twelve years ago. It was my first attempt at algorithmic composition. Thousands ? Well, yes, of course there were, as many as could be done at the rate of one every few seconds in fact.

I was going through all my old recordings on the weekend when I found a tape with a few dozen of the fugues.

The aesthetic routines in that first effort were very crude. The motif cells need restricting in rhythmic type and the whole thing needs freeing up in harmonic options. Nonetheless, the code does seem to have its own distinctive personality, and such obvious defects as occur, I am confident I could repair very easily given time and thought.

My end object, unlike that of David Cope and other more serious algorists, is not to produce imitation classical using complicated artificial intelligence, but to write heuristic, simple routines which have personality and capacity to surprise and delight.

To tell you the truth I had forgotten about it and become very busy with my improvisation. Hearing these old things has given me a bit of a kick and I am wondering seriously if I should spend some more time on it. 

What do you people think ?  Is it worth my giving it another shot ?

[da jake]

Funky stuff, man! 

[Derek]

Definitely give it another shot. I wonder why more people haven't tried to make a computer compose music originally rather than endlessly trying to make them imitate human music perfectly?

[quantum]

Derek that's a wonderful perspective you have on computer music. 

Ted, this is a really interesting approach you took to writing fugues.  I'd say it's worth revisiting your code. 

[soliloquy]

BARLOW o.o

[michael_sayers]

Hi Ted,

Thanks for sharing these!

One thing which springs to mind is the pre-computer age equivalent of this, such as here with Conlon Nancarrow's Study for Player Piano No. 21:

t=00m33s


Mvh,
Michael

[michael_sayers]

p.s. - Study for Player Piano No. 21 seems like an inaccurate title; unless it is the 21st player piano, maybe the title needs to be Study No. 21 for Player Piano?

[ted]

One thing which springs to mind is the pre-computer age equivalent of this, such as here with Conlon Nancarrow's Study for Player Piano No. 21

You're right, I had forgotten about him. There don't seem to be many people developing heuristic composition algorithms, yet anybody with a home computer and a compiler has all the means to do so, and very quickly. The main requirement is that the results are capable of enough variety, beauty and surprise to excite human ears, To do that the program needs to be considerably bigger than the few dozen lines of Basic I used for the fugues.   

[michael_sayers]

You're right, I had forgotten about him. There don't seem to be many people developing heuristic composition algorithms, yet anybody with a home computer and a compiler has all the means to do so, and very quickly. The main requirement is that the results are capable of enough variety, beauty and surprise to excite human ears, To do that the program needs to be considerably bigger than the few dozen lines of Basic I used for the fugues.   

Hi Ted,

I know how to programme in Basic, too.

Someday I might ask for tips about how to proceed with such a project, although probably all such information is freely available online.  This doesn't mean that the persons posting the information have necessarily given it a go, or that they know what mistakes to avoid . . . some things perhaps one has to learn by trial and error.


Mvh,
Michael

[emill]

From a non-pianist ... sounds very Bach-ish!
and not bad ... not bad at all!!  liked most the 482
... the alienish/ufoish 479 was most interesting.

Did you record each hand on a separate channel?
Do you have "newer" work?  Just interested to listen and
find out if time had an effect on your "fugue" compositions.

[ted]

Did you record each hand on a separate channel?
Do you have "newer" work?  Just interested to listen and
find out if time had an effect on your "fugue" compositions.

They are not my "compositions" at all, but the output of a very crude and simple computer program (algorithm) I wrote on the Amiga many years ago. Nothing at all was recorded, channels and hands being irrelevant. I just flicked through the Well Tempered Clavier (as good a place as any to start) like an alien looking for patterns. I found quite a number and just incorporated them into a few simple subroutines optimising harmony and phrase generation accordingly - not from a musical theory standpoint because I couldn't, I don't know any theory. What happens is entirely over to the code, and does not change unless I rewrite the program. If you mean have I developed fugue writing code since, the answer is no, but if I can tear myself away from my improvisation for a day or two I might be tempted.

The point of the exercise was to find out if it could be done. Having satisfied myself that, with a bit more effort, it could, I put the exercise aside. I suppose I should really try it again as it would be dead easy on my iMac using something like a Blitzmax compiler.