February 18, 2006

Chaos And Music...

by Chris Randall

I've been thinking a lot about chaos lately, as it relates to music creation. Watching the Olympics nightly, I've noticed one thing in particular about the speed events, such as downhill, combined, luge, skeleton: the guy that wins is the one that is willing to go right to the edge, and maybe stick a toe over. In the men's downhill in particular, the fastest times were turned in not by the guys that skied a perfect line, but by the ones that exhibited more of a (barely) controlled fall down the side of the mountain.

This, of course, results in a lot of wrecks, but being careful and controlled aren't necessarily desireable traits in someone that wants to go faster than everyone else. The same is true of music. Making safe music, or music that adheres to strict rules, is akin to making boring music. The most exiting music ever made is by musicians that stuck a toe over the edge. The Beatles may sound banal today, but that's only due to the fact that their best records were oldies before most of us were born. Strawberry Fields is simply out of hand, and much more "alternative" and experimental than anything Coldplay has ever done. (Then again, most anything is, so maybe that's a bad example.)

Anyways, to get back to the subject, I make no secret of the fact that I often use aleatoric generators to come up with melodies and rhythms. My favorite programs are M and UpBeat, both written by David Zicarelli before he founded Cycling '74. I've released several songs that are almost entirely generated by those two programs. What I like about them, and aleatoric generators in general, is the fact that I can set the initial conditions, but the timing and order of notes is left up to the ghost in the machine, and while it takes patience to get something that people will find listenable, once I do, the result is never something I would have thought of on my own.

I've been pondering how to take pieces of these programs I like and put them in the (somewhat stricter) framework of the VST/AU environment. My thinking is that the more control one has over the source material and initial conditions, the more likely something interesting is going to come out. So imagine this: a synth/sequencer combination (think 303) except it has Buchla's Arbitrary Function Generator (pictured above) instead of the typical Roland step-time fiasco, and a somewhat more sophisticated synth section. What would this need? What would it not need? 2 VCOs, for sure, and a multi-mode filter, all of which could be controlled by the sequencer. A randomizing section for the various bits. I like Buchla's circular design above rather than the more linear x0x style. Something to think about, anyways.



Page 1 of 3

Feb.18.2006 @ 6:42 PM
If you put it in VST format, why not as a midi-only plugin?

Not only should it send out notes, but you could let the aleatoric
process loose on cc values. It's then up to the user what (s)he
chooses to drive with it, instead of being bound to limitation set by
the audio engine of the plug...


Feb.18.2006 @ 10:00 PM
Something to think about indeed! That, sir, would be very cool.

I agree, I do think generating MIDI out (which I believe the VST spec does accomodate for) would transform this from "very cool" to "essential sequencing tool". Even if the internal sound engine was top notch (and I have no doubt that it would be) if one is using it for melodic elements you do want to be able to shift around the result to really make full use of it.

I like your mind.


Feb.18.2006 @ 11:17 PM
it'd be nice to see the bars and pipes framework of sequitur {for beos} and, erm, bars and pipes placed into a vst plugin. it galls me that no one has done it, yet.

Feb.19.2006 @ 12:28 AM
d wagenbach
i've used M for a long time and played around with symbolic composer. I would love the sort of control that leads to unexpected places that I got in symbolic composer in a VST. I agree with the rest, midi out is essential. maybe even just the sequencing engine without a synth, maybe.
if you build a synth around it, i'd love to bring the randomness in as modulators for just about all of the synth's parameters and maybe a clock divider/multiplier with you route from the randomator to the destination

great idea


Feb.19.2006 @ 2:18 AM
Chris Randall
I agree with the comment about Sequiter. I'll boot that bitch up and have another look at it; it's been a while. I've also been fooling with an OpenGL interface along the lines of the 3D mixer from BeOS. Not sure what (or if) we'll use it for, but I like to explore all possible UI routes.

As for the other comments, I also agree that a synth might be superfluous. We don't have any experience at all with making a synth, and the market is very crowded. (It amazes me how many three-OSC monosynths this business can support.) I was speaking with Adam last week about a way to harness the ability of most all of our products to respond to MIDI learn by creating a master modulation source that could also drive other people's products that had MIDI learn features, as you can never have too many LFOs. This idea kind of grew out of that, plus my long-standing fascination with aleatoric and algorithmic generation.

So, persuant to the above, I'm thinking about a 16-step sequencer in a circular pattern, like the Arbitrary Function Generator, with a couple extra mod sources, some sophisticated randomization functions, and assignable controller out. As long as we stick to simple note-on and -off, pitch, and the normal [0,127] MIDI CC numbers, it actually seems like not that complicated an idea. It would be a VSTi, rather than a MIDI plug, so it could be used in hosts other than Cubase/Nuendo, of course. Quite frankly, I don't think we could pull it off in AU. AU instruments can shit MIDI, but the hosts generally don't agree on how to deal with it like VST ones do. With Logic, you'd have to get in to the environment and do some patching. Live would be easy enough, of course, but it also runs VST, so it doesn't matter. DP? I have no fucking idea. That DAW is a complete mystery to me. How something so nice looking could be so oblique is beyond my understanding.



Feb.19.2006 @ 10:50 AM
have you seen <a href="link [www.sseyo.com]">link [www.sseyo.com]">koan pro</a>? eno did a record with it a few years back....

Feb.19.2006 @ 10:50 AM
barf, sorry for the formatting on that!

Feb.19.2006 @ 11:10 AM
Bjorn Vayner
Hey Chris,

Have you checked out Audiorealism's new synth 'abl pro'?
It has a great randomizer for the sequencer.
I'm not up to speed on above mentioned apps or plugins.
But what you are talking about, sounds like the advanced randomizer in that plugin.

That randomizer is based on an app somebody wrote to generate patterns for the 303 that are fixed to a scale.
MIDI out is being added when the AU version comes out.
Unfortunately it will only send the notes.

Anyway. I would love to see a sequencer that can generate notes, intervals, length,.... randomly but controlled
And also generates some cc's to control other stuff.
It would come in handy to have an envelope generator ;-)

There are plenty of synths. But not nearly enough decent sequencers.
Apps like plogue bidule do the trick though.
But it sucks without a proper GUI for it.


Feb.19.2006 @ 12:30 PM
I guess you know xoxos' vsti plugins? link [www.xoxos.net]">link [www.xoxos.net] and link [www.atomictraveer.co...]">link [www.atomictraveer.co...]

Talent I know is specifically a midi out plugin, and I think the others have it too. I've never used 'em but they sound similar to what your talking about.

I know the market is crowded, but an AD synth with lots of midi out functions still sounds pretty cool. I'd like to see what y'all could do. Of course, It would have to look as cool as the Buchla.


Feb.19.2006 @ 1:26 PM
Chris Randall
Well, the xoxos plugs are made with Synthedit, which automatically takes them out of the running for anything but "curious oddity." While he certainly pwns Synthedit, it's a whole different ball of wax making a real VST or VSTi.

I should mention that this is not, as we say at AD, "in the queue." I'm just thinking about how one would go about something of this nature. We haven't made any plans to make anything like what we're talking about; I'm just gauging the level of interest at this point, and getting some ideas together to see if it's feasible.



Page 1 of 3



Sorry, commenting is closed for this blog entry.