Hello,
I am now running in the dev branch (and this may have been in the other branch as well), but what is the option after the kHz of the DSP?
Thank you for your help with this question.
Brett
1 Like
Block size. So the number of samples per “bucket of audio”. This may be helpful.
FWIW I did some benchmarking of performance at different block sizes, and found that for most algorithms, dropping from 48 samples down to 16 samples has almost no effect (maybe +1% CPU), but below that it starts to ramp up. So I’m tempted to modify oopsy to set 16 as the default.
16 samples at 48kHz means the effective control rate is 3kHz, and latency is 1/3 of a ms. (At least, this should be true for audio and gate latency; other interface elements may be buffered elsewhere). Very responsive indeed, better than most desktop computer setups, and amenable to some feedback patching in the rack.
Some simpler patchers I could run at 96Khz, 1-sample block size, which is pretty intense in terms of possible feedback patching (IO latency around 0.01ms!)
Daisy is quite a powerful platform 
2 Likes
WOW, that is some very cool info!
Currently I am not doing a great deal of audio processing (I’ll be playing some audio files, but no processing … yes. 
, but I can see doing some future audio processing soon.
Very cool data and thank you for sharing.
Brett