[hts-users:00470] Re: How to generate speech in HTS?

[Omer Moav <smoavo@xxxxxxxxxxxxxxxxx>]

Hi Heiga,

We installed both "cmu_us_awb_arctic_hts" voice and "cstr_us_ked_timit_hts"
voice, and we got the same error with both:

festival> (voice_cstr_us_ked_timit_hts)
cstr_us_ked_timit_hts
festival> (SayText "Hello World")      
Segmentation fault (core dumped)
Cannot open file /tmp/est_05435_00000/utt.wav as tokenstream
Wave load: can't open file "/tmp/est_05435_00000/utt.wav"
Cannot load wavefile: /tmp/est_05435_00000/utt.wav
#<Utterance 0xb7254c48>
festival>  
festival> (voice_cmu_us_awb_arctic_hts)
cmu_us_awb_arctic_hts
festival> (SayText "Hello World")      
Segmentation fault (core dumped)
Cannot open file /tmp/est_05435_00001/utt.wav as tokenstream
Wave load: can't open file "/tmp/est_05435_00001/utt.wav"
Cannot load wavefile: /tmp/est_05435_00001/utt.wav
#<Utterance 0xb7347978>

Can you please help us fix it?


> Hi,
> 
> Omer Moav wrote:
> 
> > We saw in the mailing list that another option is to make .utt files by
> > festival, but how do we use them with HTS_engine to generate speech?
> > 
> > What is the simplest way for generating speech from text using HTS?
> 
> The most simple way is replacing pdf and inf files in festival voice by your
> own {mcp,lf0,dur}.pdf and tree-{mcp,lf0,dur}.inf.
> 
> 
> For example, if you install cmu_us_awb_arctic_hts voice into festival, you
> can find pdf and inf files under
> festival/lib/voices/us/cmu_us_awb_arctic_hts/hts/ directory.
> You synthesize speech with cmu_us_awb_arctic_hts voice by typing
> 
> $ festival
> 
> festival> (voice_cmu_us_awb_arctic_hts)
> festival> (SayText "Hello, world")
> 
> At that time, pdf and inf files under
> festival/lib/voices/us/cmu_us_awb_arctic_hts/hts/ is used to synthesize
> speech.
> So if you replace them by your own pdf and inf files, you can get speech from
> your pdf and inf files.
> 
> 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://kt-lab.ics.nitech.ac.jp/~zen
> ------------------------------------------------
>