halim.sahin at t-online.de
Tue May 29 14:54:58 CEST 2007
On Di, Mai 29, 2007 at 01:35:17 +0200, Jan Buchal wrote:
> >>>>> "HS" == Halim Sahin <halim.sahin at t-online.de> writes:
> HS> Hello all, I want post my problem again about the ibmtts module.
> HS> First the log: Sun May 27 09:08:25 2007 : Ibmtts:
> HS> Returned 17 bytes from get_part. Sun May 27 09:08:25 2007
> HS> : Ibmtts: Text to synthesize is |<speak> </speak>|
> HS> Sun May 27 09:08:25 2007 : Ibmtts: Sending text to
> HS> synthesizer. Sun May 27 09:08:25 2007 : Ibmtts: End of
> HS> data in synthesis thread. Sun May 27 09:08:25 2007 :
> HS> Ibmtts: Trying to synthesize text. Sun May 27 09:08:25 2007
> HS> : Ibmtts: Waiting for synthesis to complete. Sun May 27
> HS> 09:08:33 2007 : CMD: <STOP
> HS> Sun May 27 09:08:33 2007 : Ibmtts: module_stop().
> HS> You can reproduce this by typing spd-say " " or spd-say "-"
> I'm not sure if this depends with index marking but first in IBM TTS was
> it implemented and now is not included, so please try last cvs version
> of SD, maybe solve it this problem.
The - does not crash it any more but more important are the spaces.
I made a generic conf for mbrola and tested it in orca and speechd.
It works great so I think there must be other problems in ibmtts module.
viavoice is a great synthi but currently not very usable in speechd.
PS. Thanks for this great software (speechd).
I hope It will support more and more synths.
More information about the Speechd