danijel3 / pyhtk Goto Github PK

The example uses the following parameters in configuration:
TARGETKIND = MFCC_D_A_0 LOFREQ = 80 HIFREQ = 7500 ENORMALISE = FALSE ESCALE = 1.0
The lofreq, hifreq values are also hardcoded in gen_list.c.

I would like to use MFCC_E_D_A_Z and the default HTK values for LOFREQ (??), HIFREQ (??), ENORMALISE (T), ESCALE (0.1). Can you provide some guidelines?

I find this project very useful, thanks a lot for sharing.

Hi! Congrats for your work, it is superb.
I am trying to load a .mfc file generated by HTK Tools, and when I do this:

from HTK import HTKFile
htk = HTKFile()
htk.load('devel_0001.mfc')

The following error shows up: Traceback (most recent call last):
File "/Users/PycharmProjects/speech_mfcc/test2.py", line 5, in
htk.load('j_htk/devel_0001.mfc')
File "/Users/PycharmProjects/speech_mfcc/HTK.py", line 142, in load
s = f.read(self.nFeatures * 4)
ValueError: read length must be positive or -1

Am I missing something?
Thanks!

Hello,

Thank you for publishing your work. I have a question question in regard to gen_filt.c
so if i change frame size and stride to
int frameSize=44100;
long sampPeriod=22050;
which is equivalent to 1000 ms frame size, 500ms off set, for 44100 sample per seconds.
The the gen_filt.c failed to generate the correct triangular weights. Can you be so kind of point me to the right direction?

Thank you so much

Loc

how to store features to 123jf.mfcc file in htk format