adc~ inside a poly~


    Dec 08 2012 | 2:45 pm
    Hello,
    It is possible to specify logical channels received by adc~ by sending it a list.
    Though, used inside a poly~, it won't work correctly & causes crashes for various reasons (Max 6.0.8 on OSX.6.8).
    Since these problems are not mentionned in the reference, can we consider that as a bug ?
    Below are the poly~ and the patch for your tests.
    Nicolas

    • Dec 20 2012 | 12:31 pm
      Anyone would like to try this ?
    • Jan 03 2013 | 6:11 pm
      Yes, you have to put your adc~/dac~ out of the poly~. Using in~/out~ to communicate with the main patch is the way to go.
    • Jan 13 2013 | 10:04 am
      Thank you.
      Is there anywhere a set of recommendations about specific programming rules inside poly~ ?
      N.
    • Jan 13 2013 | 10:41 pm
      E.J. is there a particular reason for this? Bit of a bummer on that one.
    • Jan 14 2013 | 9:46 am
      Poly~ has many options to do something like up sampling/downsampling, vector size that can be different from the main patcher, threading. All those things are handled by the poly~ object itself.
      FWIW, the patch doesn't crash here.
    • Jan 19 2013 | 5:43 pm
      @Nicolas, yes just use in~/out~ ;-) The use of send~/receive~ fromto poly~ (within it's fine) is not supported either for the same reasons.
    • Jan 19 2013 | 7:16 pm
      Not supported as in "parallel optimizations don't work" or not support as in "will not be allowed in the future"? I strongly hope it's the former and not the latter.
      There are situations that don't justify a full matrix mixer, but having the ability to add dynamic sends/receives with effects is helpful, and poly~ fits that bill perfectly. I've also been doing some stuff with running poly~ as a serial process (vox 1's output into vox 2's input and so forth), which makes a great testing rig for developing things like allpass spring reverbs and cascading feedback delays.
    • Jan 20 2013 | 8:25 am
      From as long as I can remember, send~/receive was not fully supported in/out of poly~. You might have some unexpected latency. Multi-threading just make things even trickier.