Dear forum,
I am wondering what the differences are between the different yin objects: yin~, pipo.yin and gbr.yin. The first one has no way to determine the slice size and hop size and there is no autocorrelation output but other then that they seem similar. If I look at the output of these objects though then I see that the quality factor in yin~ and gbr.yin are the same while the quality output (also called periodicity) of pipo.yin is about a third of this and there is no constant factor between the two. Is this a bug or is there a difference in what this quality factor tells us?
How are the slice and hopsize of the yin~ object determined? Are they determined by the the value of the attribute lowest estimated frequency?
I did not see a possibility to ‘train’ yin~ or determine an input instrument. Would it not help for pitch estimation when the instrument is known? (I thought yin uses a spectral method for pitch estimation).
Thanks in advance, Hans.