4 ms·

"Stochastic" made me hopeful - thinking hey Xenakis, sonic exploration of data, etc - and "background music" made me skeptical - thinking here comes another gar

"Stochastic" made me hopeful - thinking hey Xenakis, sonic exploration of data, etc - and "background music" made me skeptical - thinking here comes another garbage piece of general midi mapping that passes for generative music. It's hard for me to be so negative, but this is a bit like posting your new bootstrap site on a design forum. Not that HN is all computer musicians and generative composers, but some of us are. This might be your first step into generative music, or music in general, and please keep sharing! Some of the best advice I ever got when starting to write music was: you're going to write junk for quite a while, don't let it discourage you. I'd love to hear more about the specifics of what you're working with algorithmically. After listening to more of the "deep ocean" mp3, it sounds like you have a set of 10 or 15 short gestures, and the algorithm is determining the order they'll play and how they'll be stacked. Good place to start! If you aren't interested in applying systems to the material itself, a next step might be to just play with amplitude over time. You can prolly increase the overall density of samples if you apply stochastic envelopes to each, without affecting the still mood -- it might help reduce a feeling of monotony, too.

lurzo 12y ago
Hi erik! i'm the music/audio guy of the site. First thank you for the precious feedback and encouragement :). Your description at the end of the reply is accurate. There are 10 to 15 instances of synths/sample based stuff, which take midi imputs based on the following criteria: A list of frequencies. A list of note durations. A list of amplitudes. All this process is triggered by a set time (in seconds) which, going in tandem with a selected chance rate of generating the note, makes a music event. Having this running on a server would be hard for us, so the decision to make a long export of these "sessions". I agree on the fact that Stochastic music has a certain academic connotation and standards for a composer, standards that are far from being achieved from the tracks on the site. I will try my best to make new soundscapes even more dynamic in the future :).
- erikschoster 12y ago
  Hey thanks for being so understanding! I just want to say that I feel badly for being this negative, it's out of character for me, but this is a realm that's close to my heart and I'm just critical because I care about it and want to see projects like these exceed my wildest expectations. You should also take my comments with a grain of salt because my hopes for things like this are for lack of a better way of putting it musically bound. I can't help but listen to this as art, and so I'm responding to it that way. I'm taking these pieces as things for close listening. I think making long renders for static streaming is a fine approach by the way - have you thought about automating a re-render more often for each piece? A more expensive alternative would be to run a dedicated render server to generate music in realtime, and then multiplex the stream on demand with something like wowza. Makes me miss the days of r4nd.org: http://rhizome.org/discuss/15607/#29894 http://rhizome.org/discuss/15607/#29894 Good luck with the project & thanks again for sharing!
  - lurzo 12y ago
    Constructive criticism is always appreciated! A nice happy solution that we want to implement in the coming days is to have even longer exports start at a random file time (having it looping off course). Also i didn't know about r4nd.org! Is there a link where to listen some snippets of it?

lurzo 12y ago

Hi erik! i'm the music/audio guy of the site. First thank you for the precious feedback and encouragement :). Your description at the end of the reply is accurate. There are 10 to 15 instances of synths/sample based stuff, which take midi imputs based on the following criteria: A list of frequencies. A list of note durations. A list of amplitudes. All this process is triggered by a set time (in seconds) which, going in tandem with a selected chance rate of generating the note, makes a music event. Having this running on a server would be hard for us, so the decision to make a long export of these "sessions". I agree on the fact that Stochastic music has a certain academic connotation and standards for a composer, standards that are far from being achieved from the tracks on the site. I will try my best to make new soundscapes even more dynamic in the future :).

erikschoster 12y ago
Hey thanks for being so understanding! I just want to say that I feel badly for being this negative, it's out of character for me, but this is a realm that's close to my heart and I'm just critical because I care about it and want to see projects like these exceed my wildest expectations. You should also take my comments with a grain of salt because my hopes for things like this are for lack of a better way of putting it musically bound. I can't help but listen to this as art, and so I'm responding to it that way. I'm taking these pieces as things for close listening. I think making long renders for static streaming is a fine approach by the way - have you thought about automating a re-render more often for each piece? A more expensive alternative would be to run a dedicated render server to generate music in realtime, and then multiplex the stream on demand with something like wowza. Makes me miss the days of r4nd.org: http://rhizome.org/discuss/15607/#29894 http://rhizome.org/discuss/15607/#29894 Good luck with the project & thanks again for sharing!
- lurzo 12y ago
  Constructive criticism is always appreciated! A nice happy solution that we want to implement in the coming days is to have even longer exports start at a random file time (having it looping off course). Also i didn't know about r4nd.org! Is there a link where to listen some snippets of it?