Struggling with fzero~...


    Oct 11 2013 | 6:09 pm
    Hi, I'm trying to immigrate out of gbr.yin~, my trusted pitchtracker and move on with my new life with fzero~. (since now fiddle~ seems defunct)
    I put all kinds of filters to fit me (violin) but still can't get it to work well as gbr.yin~. I wouldn't think of using fzero~ if it wasn't for the gabor objects (IRCAM) problem in M4L; it won't work, and FTM is also a pain for standalone.
    If anyone has any suggestion I greatly appreciate... Someone said something about using zl stream, zl median but I don't know how. It sounds like those maybe what I need...
    ps. All the stuff I have under fzero~ - change, round, speedlim etc. etc. is for the violin...I use it also with gbr.yin~

    • Oct 11 2013 | 6:41 pm
      ps. BTW this is my gbr.yin~ patch made for me at IRCAM... (by Norbert Schnell) I have been using this for a few years now in all my concerts.
    • Oct 11 2013 | 11:00 pm
      What is happening with fzero~ that isn't working for you? I'd be glad to help, but I'm not sure where to start.
    • Oct 11 2013 | 11:19 pm
      Thanks so much, going to private convo. It's the general behavior, erratic output and mistakes...
    • Oct 12 2013 | 12:15 am
      Why go to private convo? One of the advantages of an online community is that you might find someone who has already answered your question. You could save a future user the trouble by keeping the conversation, along with any solutions, public.
    • Oct 12 2013 | 12:36 am
      Yes sorry---Michael kindly emailed me to help---I just sent him this, side by side comparison. I made a simple violin phrase to load in so you can see exactly what the problem is...
    • Oct 12 2013 | 12:37 am
      OK here it is---please load this violin soundfile to see the result. Here it is in the dropbox, soundfile named 10-11.aif (I can't get rid of the error message. The file was too big to upload in this forum)
    • Oct 12 2013 | 2:00 am
      I'm still working on this, but taking the median helps. Also, normalizing the input audio made a noticeable difference. The levels on the file above are really low.
    • Oct 12 2013 | 2:18 am
      Thanks so much! I understand my level is low, but using gbr I didn't have a problem...
    • Oct 12 2013 | 2:33 am
      ps. And those who use FTM you can see the new comparison:
    • Oct 14 2013 | 1:13 pm
      Since I'm pressed with projects I decided to use yin~ to replace gbr.yin~ since yin~ works in M4L and Standalone without hassle. Still cannot use fzero~ for performance...