[hts-users:00063] HTS voice output problems

["Nicholas Volk" <nvolk@xxxxxxxxxx>]

Hi,

I've managed to get my Training_bl_fi_ms_clunits.pl script to work
without errors and generate the files described in
labels/fullcontext/gen.

I put a sample file in http://www.bitlips.fi/HTS/ms_272_16.raw
(sample rate 16000, Lin16, Mono, Little Endian).
As you can see (probably not hear ;) there are some pitch period
like areas in the file. (The reason they have nothing above 5500KHz
is that I upsamples my training data from 11K to 16K.)
Most of the file is however problematic: very low noise and clicks,
no speech. Any ideas what causes this would be appreciated.

I used only 90 sentences as the training material while experimenting
(due to speed). Is the small training data the reason for this?
Can anyone suggest a minimum number of instances of each phone that are
required for training?

BTW: There seems to be an error in HTS_ARCTIC's lab_format.pdf.
The PDF says that there's an underscore '_' between i1 and i2.
It probably should be an equal-sign '='?
At least that what utt2lab puts there.

regards,
  Nicholas Volk