Haven’t seen these results posted elsewhere.
Here is the output of the cyclictest
program running on the Organelle M with Wifi on and Additive Synth running.
WARN: Running on unknown kernel version...YMMV
policy: other/other: loadavg: 1.09 0.73 0.32 2/129 10985
T: 0 (10168) P: 0 I:1000 C: 54594 Min: 10 Act: 13 Avg: 28 Max: 3276
Units are in microseconds. Not sure what the impact of turning off Wifi would be…
I don’t know what to make out of that numbers…
In My OG and in pd, I could go down to 7 Ms (Which is pretty good). but not much less if the patch is actually doing something. I would be really amazed if an organelle M could do under 5 Ms without dropouts…
This is more of a test of context switching latency in the real time linux kernel on the Organelle M (avg = 28us), which has an impact on the latency, but isn’t a direct measure of audio latency. How did you measure your audio latency?
In my case, I trust PD and my ears… but If I had to measure I/o latency, I would record an impulse, directly to soundcard in one channel and processed by organelle in other channel. then in the DAW, I would see the difference. (soundcard latency wouldn’t affect this measure.)
1 Like