cycle~ phase analysis accuracy to float

    May 06 2011 | 12:26 am
    I'm wondering if anyone has any ideas why cycle~ phase info to a float is so innacurate on my computer? I've lowered the vector size and tried the with and without overdrive/audio interupt, but at any signal higher than 0.5hz it's really inaccurate.
    See the test patch below, sorry if i'm missing something obvious.

    • May 06 2011 | 8:23 pm
      In your case the timing accuracy depends on two things:
      1. You use number~ to convert the audio signal to messages. number~ uses an internal clock to sample the audio signal values. You can set this sampling interval with the attribute "interval" (Or "Update Interval in Milliseconds" in the Object Inspector). You can also use [snapshot~] which does basically the same without GUI. The sampling interval can be set as first argument
      Edit: Ok, forget the stuff about audio vector size. I had already forgotten your exact post, when writing this ;)