Closed gogins closed 2 months ago
I could not get Swami to run without the -p
option and when I did get it to run, I could not get it to play SoundFonts, with the following error:
(swami:12598): libswami-CRITICAL **: 19:13:56.462: swami_wavetbl_close: assertion 'SWAMI_IS_WAVETBL (wavetbl)' failed
(swami:12598): libswami-CRITICAL **: 19:13:56.462: swami_wavetbl_open: assertion 'SWAMI_IS_WAVETBL (wavetbl)' failed
I believe this is because Swami depends on libfluidsynth version whereas I build the Fluidsynth opcodes with version 1 which I have therefore installed.
However Polyphone works. I downloaded the bionic package and installed it and it played right away.
Using sounds from the wild can be done in several ways:
I need to collect sounds. They need to be sounds that draw one in, that is all. They can be long or short, pitched or not pitched, verbal or non-verbal, and natural or artificial (but not direct synthesis).
Possibilities:
Is there a way to load samples from the Web server for use by Csound? See how Strudel loads samples.
I am no longer working on issues that I do need for working on a piece.
That said, I have created a Python script fo embed samples and other files in <CsFileB>
elements in .csds that I can use for this purpose in future pieces.
Issue #117 was a success, and demonstrated that I can incorporate both non-pitched and note-generating instrument definitions into my standard patching system, and use them in my usual style of algorithmic composition.
Following this idea further, create some new patches that incorporate sampled and granulated sounds. This may involve using:
diskin2
orGEN01
.cmask
.The same principles of transparently normalizing and fitting instrument outputs into the maximum ranges of the score space should be used. Use normalized outputs if possible, otherwise reflect them into a usable range that lacks artifacts if possible, otherwise transparently omit the outputs.
Design alternatives include one patch for each sample, or one patch for each type of input (diskin2, SoundFont, etc.) with parameters for audio input and good ranges.