pYIN implementation #809
pYIN implementation #809
Conversation
| import time | ||
| import os | ||
|
|
||
| filename = os.path.join(os.path.dirname(__file__) ,'../../audio/recorded', 'long_voice.wav') |
There was a problem hiding this comment.
This audio should be added to the essentia-audio repository (do a pull request). What is the origin of the audio (license)?
There was a problem hiding this comment.
We can use the existing audios in essentia-audio. How can I use the audios there? How we add the link in this test .py?
There was a problem hiding this comment.
Create a pull request with new audio files to add to essentia-audio.
There was a problem hiding this comment.
pull request sent.
| # check if estimations actually have some sense | ||
| # it is better to have real case using a monophonic audio | ||
|
|
||
| # def testARealCase(self): |
There was a problem hiding this comment.
Add a real regression test (given an audio input check if the resulting pitch values are as expected).
There was a problem hiding this comment.
Great! I have changed waf for a flac file (to save some space).
|
@dbogdanov Hey, man, I did some comments on your comments. Could you check them when you are free? |
|
@ronggong Replied to your comments. |
|
@dbogdanov please resolve the conflicts. constantq.cpp can't be compiled in my c++, because it can't convert double complex to double. check MTG/homebrew-essentia#9 added math functions in essentiamath.h |
| pitch += [f] | ||
| voicedProbs += [vp] | ||
| self.assertAlmostEqual(mean(f), freq, pitch_precision) | ||
| # self.assertAlmostEqual(mean(voicedProbs), 0.6, vp_precision) |
There was a problem hiding this comment.
Merging now. Can I remove this line as it is commented? Or if that's useful to keep, then we can give additional comments.
There was a problem hiding this comment.
You can remove it, it is not important.
|
Everything merged, thank you for this great contribution!!! |
Implemented Pyin algorithm,
Wrote the unitest and pythontests
Wrote the documentation