Precise monophonic pitch detection
I need precise monophonic (voice) pitch detection and I have tried fzero and fiddle. The results I'm getting are very inaccurate.
So I'm wondering if I need to improve my settings or if the mistake is somewhere else. What would be a good setting for voice detection of single notes? Is there a better object for handling monophonic signals?
Any advice will be much appreciated.
Thanks!
Have a look at Antescofo:
http://forum.ircam.fr/product/antescofo-en/
Sigmund~ is also worth looking at, it's basically fiddle 2.0.
And there's the vanilla retune~. I had quite good results with voice input.
Try an eq and compressor before the pitch detector. Don't hesitate to use aggressive settings. The goal is to get rid of unwanted frequencies and make a clear distinction between the active vs. inactive state of the input. The fzero help file has some filter settings can be a good starting point.
Consider changing your I/O vector size in audio setting.
Also setting ranges for pitch detection makes triggering events with pitch detection easier. IE instead of looking for C4, look for a note between B4 and C#4, or a note above B4, or bellow C#4 etc.