Links
- Dave Krooshof
- Ernst van der Loo
- Aslaug Holgersen
- Thijs Scheele
- Juan Parra
- Leon Spek
p2p weblog
our weblog about filesharing
and legal issues concerning
the publishing of music.
home:
www.dendriet.nl- linkpage
new subject
Archives
Dendriet is a network of musicians in The Hague (.nl) that perform with networked electronic and acoustic instruments. The word dendriet, or dendrite, is also the name for the receptors of a nervecell. Dendrites are typically branching branches. You'll find these fractal shapes in ice, in metal chrystals in limestone, and likewise, in our performance setups. Please do check out our homepage by clicking the stone. You too are invited to join the discussions.
12.9.04
Bit by bit
I guess I have to learn to program in C, as I want to alter sounds on the bitlevel, or at least timeframes that are shorter then a 20KHz period. Here's a long term wish: Markov chaining with soundbytes!
Here's Markov on the wordlevel. (Nick Didkovsky is an old friend of Joel Ryan).
For a Markov chain, you pick a random word from a text, look randomly in the text for that word. The word in the text next to the one you were looking for is added to the chain, and it is the next word to look for in the text. repeat.
This produces sheer comprehencible text, which is statistically similar to the input. I imagine a sound that, when markoved, is statistically familiar, but pitchwise messy.
Here's Markov on the wordlevel. (Nick Didkovsky is an old friend of Joel Ryan).
For a Markov chain, you pick a random word from a text, look randomly in the text for that word. The word in the text next to the one you were looking for is added to the chain, and it is the next word to look for in the text. repeat.
This produces sheer comprehencible text, which is statistically similar to the input. I imagine a sound that, when markoved, is statistically familiar, but pitchwise messy.
