I finally got around to throwing together a prototype of a real-time auto-orchestrator for Sonar. In theory, it should also work in other sequencers that support MFX. It's extremely simple, but amusing. A full-blown version would allow multiple Sonar MIDI tracks to be coordinated, but Orchid 1 only allows multiple channels within a single instance of a multi-timbral plugin like GPO.
Orchid has 16 "instruments", each of which has a MIDI output channel, a lowest possible note, a highest possible note, and a priority. As notes are played on the MIDI input, they are assigned to voices according to priority among those instruments whose range allows them to play the note. Each instrument can play only one note at a time.
The idea is to allow you to play a keyboard part live and have it played by an ensemble of monophonic instruments (violin, cello, horn, oboe), with each note going to a different instrument and without any spontaneous creation of musicians. The algorithm is simplistic: doing even passable orchestration in real time, when you can't know the duration of a note when you must assign it to a voice, is a very hard problem. But by choosing appropriate range and priority values, I've been able to do some interesting improvisations for chamber ensemble.
I haven't downloaded your example yet, but I did a fractal, stochastic composition project back in the 1980's, which ran on an Z80 microprocessor which drove an ARP 2600 via some 12-bit industrial D/A converters. Sometimes it would generate rather pleasant, very human-sounding music. But most of the time, it just noodled. ;o)
But I agree that running the output of your fractal sequence generator into Orchid, and Orchid into GPO, could be really interesting.
And I note that you give your location as Berkeley, CA. I was living on the 1300 block of Addison St. when I built that Z80 system...
...And I note that you give your location as Berkeley, CA. I was living on the 1300 block of Addison St. when I built that Z80 system...
Bay area resident since I came to UCB to do chem. Lived in Orinda for several years but returned to Berkeley 10 years ago. North-side flatlander now. I was doing strictly analog recording in the early 80's, polymoog, couple of Sequential Circuits synths...got pretty handy with the splicing block...I visited Xerox PARC in the late 70's and saw this thing called a "mouse"...look at us now.
Try googling "fractal music". There's a lot of stuff out there, including free and shareware programs.
Mathematical expressions that can generate series of numbers can be used to generate "music", or sequences at least. Most of the stuff I've heard is unremarkable, noodling like Kevin said, but occasionally there is something very interesting.
I visited Xerox PARC in the late 70's and saw this thing called a "mouse"
Could you tell us more about it?
We are the music makers, we are the dreamers of dreams …
24" 2.4 Ghz iMac, OSX 10.4.10, MOTU 828 MKII, 2 Glyph 250 Gig external drives, Logic 9, Finale 2008 GPO, JABB, Strad, Gro, Reason 4, EWQL Storm Drum, Adrenaline, Symphonic Choirs, SO Gold,All Arturia Synths, Many NI Synths, Spectrasonics Synths, KH Strings, VEPro on a Windows 7 4x 2.8 Ghz 12 gig of RAM
Btw, I've been fooling with 'composer' programs for 20 years or more. A surprising amount of my music incorporates elements from these programs. I use to be a heavy user of Dr. T's stuff for the Atari and still use Paul Whalley's "TANGENT" and "PNF" composers extensively.
I find that, while you can get surprisingly far into a composition with these things, in the end they are just good for generating 'snippets' or fragments that are incorporated into a larger composed work.
I recommend all interested musicians to check these programs out. Many are completely free and those that aren't are pretty inexpensive...
Billp - your fractal stuff sounds intriguing. While a student at UNC-Asheville, I got a chance to work with Dr. Wayne Kirby's "Serious Composer" software, which ran through a bunch of 1/f calculations based on user-defined seed and limitation values. The most interesting results usually came after it was edited and filtered a bit by the user - which means that the amount of 1/f and the amount of human composition was a big gray area. Now it seems that he's gone into music therapy, which is way cool.