?

Log in

RMG Development
 
[Most Recent Entries] [Calendar View] [Friends]

Below are the 7 most recent journal entries recorded in Random Music Generator (RMG) Development's LiveJournal:

Thursday, July 21st, 2005
11:14 am
[kyeli_unlikely]
ringtones!
Make RMG ringtones!! :)

Wheeeee!! I would *love* that!! :D

Current Mood: silly
Tuesday, May 3rd, 2005
12:19 am
[ex_askesis860]
cherry picking
rmg1.midi and rmg2.midi, range-constrained, thinned, quantized, and slapped onto a house beat.

This is something like how I envision musicians using RMG - finding sections they respond to and doing their own looping and edits.

And, as long as we're here, another loop that has nothing to do with RMG.
Friday, April 22nd, 2005
7:57 am
[ex_askesis860]
some MP3s, with notes
The following are the sorts of things I would be doing to RMG output:

Larry Fast sounds
Formanty sounds
Chimes and bells
All analogue

This is a simple four-track recording of rmg1.midi and rmg2.midi. In the first three tracks, the first file was sent to a K5000s, and the second to a Q. On the last, both files were sent to an FR-777 and overdubbed. The results are kind of aimless, and work best (I think) with the bell sounds.

Notes:

- Name the output files *.mid. Most applications are expecting the three-letter extension.
- Widely varying velocity (amplitude) isn't necessarily a bad thing. In the analogue track, for example, you'll hear it controlling filter cutoff and resonance instead of volume. It's just a number - people will do weird things with it.
- The music in the files is insanely fast. The files were played back at 30bpm for these recordings. Try getting in the neighborhood of 120bpm, or 30 bars per minute.
- The program changes in rmg6.mid are fun, but I won't want them. I suspect most other users would want to insert their own patch changes manually - with the exception of GM instruments, there's no telling what a program change will get you.
- Can we start doing range constraints soon? Even this meandering, noodley thing would been more interesting if each part had been constrained to half the piano range.
- Can you solve all of my compositional problems so I never have to write another note of music? Can you? Huh? Huh? When?

Current Mood: why am I up so damn early?
Sunday, April 10th, 2005
6:37 pm
[ex_askesis860]
General MIDI and you
Today is a good day I think for talking about channels and patches and ranges!

Word and numbers and tables OH MY!Collapse )
Saturday, April 2nd, 2005
9:11 pm
[ubiquity]
RMG infrastructure progress

  • rmg1.midi: strawman datastructures, varying amplitude
  • rmg2.midi: strawman datastructures, fixed amplitude (the varying amplitude just sounded too unpredictable and wacky)
  • rmg3.midi: happy, well-factored, clean datastructures
  • rmg4.midi: unstubbed the note duration calculation, now the notes don't all have to be the same length
  • rmg5.midi: figured out how to make it stay in the major scale
  • rmg6.midi: figured out how to change instruments (and went a little overboard)

There's still no actual algorithm generating the music, I'm still working on the infrastructure. But pretty soon I should have enough infrastructure to actually reimplement the algorithm used by RMG v2.0.
Thursday, March 31st, 2005
11:17 pm
[ubiquity]
it's so stupid! it's wonderful!!!
I've been so frustrated with RMG infrastructure lately that I had a conniption of joy over getting it to play a simple scale with note delays. (: For your amusement...

TEST TEST TEST!Collapse )

Cool. Now I've gotten most of the intro examples from Common Music to work, and I have a vague idea of how to write the RMG infrastructure on top of it. The Sierpinski example was kinda cool.
Friday, March 11th, 2005
1:06 am
[ubiquity]
Candy in 27 seconds! (:
Woot, I actually made some "progress" tonight!

I took a very nice MIDI file, loaded it into Allegro Common Music, and wrote it back out with all the note delays removed. Then I wrote it out again with a 10-second delay and then after that, all the note delays removed.

Never mind that I did all this by accident, at least it's a proof of concept. (:
RMG Homepage   About LiveJournal.com