why the clicks?????


    Jan 31 2009 | 6:52 am
    Hi all
    Here is a granular sampler type patch I created after seeing the video; http://ca.youtube.com/watch?v=m1GElnwAXn8. (a little long start watching @ around 3min)
    My patch is not exactly the same but has the same idea behing it. The problem I get w. my version IS THOSE CLICKS. (how come he gets no click?)
    Yes I know this is an old topic and god knows I tried many times solving it (w. rampsmooth etc.) but was never completely succesfull.
    I have tried incorporating my patch to the poly~ clickless voice as well as the grooveduck patches, in the example folder. But with no success. THe clicks remain!
    Here is the patch;
    If any of you have any suggestions, please send them my way.
    Thanks a lot to all of you
    phil

    • Jan 31 2009 | 7:12 am
      oops !!
      I need to get some sleep. there was critical connection in the previous patch (line~ to [*~ ] ) the clicks still remain nonetheless!
      here it is again
      thanks
      phil
    • Jan 31 2009 | 5:46 pm
      Hi again
      Wouldn't it be a logical thing for a trapezoid or triangular shaped [BPfunction] amp env. with it's [setdomain ] lenght being the same size as the as the selected sample in [waveform~], get rid of any clicks???
      THanks for any advice
      phil
    • Jan 31 2009 | 5:51 pm
      Quote: phil hughes wrote on Fri, 30 January 2009 22:52 ---------------------------------------------------- > Hi all > > Here is a granular sampler type patch I created after seeing the video; http://ca.youtube.com/watch?v=m1GElnwAXn8. > (a little long start watching @ around 3min) > > My patch is not exactly the same but has the same idea behing it. > The problem I get w. my version IS THOSE CLICKS. (how come he gets no click?) > >
      I'm worried that the timing between your line and your metro isn't tight. Line has a default granularity of 20 ms. I think you'd be better of using the same metro to bang a counter. Also, be very careful of order of operations.
      Good luck,
      mz
    • Feb 02 2009 | 12:30 am
      Hi and thanks for the reply
      >I'm worried that the timing between your line and your metro isn't tight.
      How come,it's the same metro that is banging the bpf? Is the that bpf is a UI? what would be a work around?
      >Line has a default granularity of 20 ms. So i just have to put a min of 20 on metro.right?
      >I think you'd be better of using the same metro to bang a counter. What function would counter have int this case? How would it help?
      Also, be very careful of order of operations? would you mind telling me what i'm doing wrong as to order of operations. I would really appreciate it.
      Thanks a lot
      phil
    • Feb 03 2009 | 6:47 pm
      Quote: phil hughes wrote on Sun, 01 February 2009 16:30 ----------------------------------------------------
      > >I'm worried that the timing between your line and your metro isn't tight. > > How come,it's the same metro that is banging the bpf? Is the that bpf is a UI? what would be a work around? > > >Line has a default granularity of 20 ms. > So i just have to put a min of 20 on metro.right?
      It some point the line gets the go message, then it outputs a new value every 20 ms. The metro is changing all over the place. Even if metro never went below 20, these two events aren't synchronized.
      Hence:
      > > >I think you'd be better of using the same metro to bang a counter. > What function would counter have int this case? How would it help?
      Don't rely on line's timing, drive it yourself with the metro you have. > > Also, be very careful of order of operations? > would you mind telling me what i'm doing wrong as to order of operations. I would really appreciate it. >
      You're starting the envelope before you start to play. The loop points are being set completely asynchronously. The subtract object coming out of waveform is comparing the endpoint of one loop with the start of the previous loop. It's a systemic problem.
      Look at the trigger object.
      I was trying to give you a patch when Max crashed. I think what you've got sounds ok at longer grain lengths, like 50 ms. When the grains get down to 15, your envelope is very sharp. (3ms ramp time is probably not enough to prevent clicks)
      also, there are some patches in /MaxMSP 4.6/examples/sampling/granular/ that might help.
      mz
    • Feb 03 2009 | 6:51 pm
      Quote: phil hughes wrote on Sun, 01 February 2009 16:30 ---------------------------------------------------- > Hi and thanks for the reply > > >I'm worried that the timing between your line and your metro isn't tight. >
      One more thing. Put your Scheduler in Audio Interrupt and set the vector sizes low -- 64, for example.
      mz
    • Feb 03 2009 | 7:47 pm
      OK. Try this:
    • Feb 04 2009 | 4:13 pm
      Wow THanks so much!!!
      At the moment I cannot try your patch as I am out of computers. My Ac adaptor caught on fire. I am waiting for a new one. The second I get it , I will try your patch and get back to you asap.
      I will 1 or 2 questions.
      THanks again.
      phil