acceleration question: using DIY delta and snapshot -type abstractions


    Oct 27 2011 | 1:05 pm
    Hi I am using several [pipe] and [delay] objects to capture the rate of change in one slider, as acceleration data for a second slider (via message to [line~]); it works, but is occasionally unreliable, and I suspect my implementation is clunky due to my DIY versions of delta/snapshot. Are there any glaring optimzations possible that I've missed here:
    Thanks Brendan

    • Oct 27 2011 | 8:37 pm
      Hi brendan,
      I haven't look at your patch in details, and it may not help, but I usually write a delta this way :
      As well, to smooth things with [line], you may consider using its second argument (the time grain) maybe.
      hope this helps
    • Oct 27 2011 | 9:14 pm
      Hi Ch yeah this is the method I'm using - the issue is that sometimes (rarely) the delta is incorrect; low when it should have been high. But only rarely.
      Brendan
    • Oct 28 2011 | 11:00 am
      yeah this is the method I'm using
      you probably posted a wrong patch then, cause it's not the one used in what you posted
    • Oct 28 2011 | 12:39 pm
      Look inside the [myDelta] subpatch
    • Oct 28 2011 | 10:58 pm
      I did :)
    • Oct 29 2011 | 1:28 am
      Hi Ch the patch you posted as a suggestion uses [t f f] -> [-] in exactly the same way I do.
    • Oct 29 2011 | 8:57 am
      hi brendan,
      we're not going to continue this discussion for ever please. have a look to the patch you posted, and see that you're using a [pipe 20] in it. I dont that like solution, cause it strongly relies on the line and pipe timings.
      It's up to you, but this :
      is not this :