[hts-users:00732] Re: Locating the boundaries of phonemes from the .cmp files

["Heiga ZEN (Byung Ha CHUN)" <zen@xxxxxxxxxxxxxxx>]

Hi,

Zuko Fani wrote (2007/06/27 19:06):

> Trying to do an error analysis of why my phonemes are breaking. 
> Now i need to know how to locate the phonemes (SegLab) in the .cmp files. 
> HInit says 50 Observation Sequences of the phoneme are seen.

It means that 50 segments for phoneme 'dh' were found in your .mlf file.

> I assume that the boundaries of these phonemes can be located in .cmp files.

No.

> In the .cmp files are frame indexes so how do i locate my specific
> phoneme between frame indexes.    

Phoneme boundaries should be specified in .mlf file.
Please refer to HTKBook section 6.2.1 for details.

Regards,

Heiga ZEN (Byung Ha CHUN)

--
------------------------------------------------
Heiga ZEN     (in Japanese pronunciation)
Byung Ha CHUN (in Korean pronunciation)

Department of Computer Science and Engineering
Nagoya Institute of Technology
Gokiso-cho, Showa-ku, Nagoya 466-8555 Japan

http://www.sp.nitech.ac.jp/~zen
------------------------------------------------