Hallo everybody, first of all I have found small bug on 64-bit platform: in src/hts_build/data/mkdata.pl (written by Heiga Zen), line 114, the format of first two items stored into tmp.head file are formatted by -al in command x2x. It will write two long numbers - in 32-bit platform they are 4B long, as expected by HTK. However, when SPTK (I use 3.1 version) is compiled at 64bit platform, the numbers are 8B long and HTK (also 64bit version) cannot of course read it. Quick workaround is to use -ai switch; it should work on both 32 and 64bit. What troubles me, however, is that Training.pl script from src/hts_build/ does not build HTS voice correctly (checked on CMU ARCTIC AWB 0.90 and CMU US SLT ARCTIC 0.95, on both the same type of failure occurs). No changes were made both in scripts and voice! The failure is manifested by the missing duration.pdf and trees-dur.inf files in the .tar.gz file built, although the building process claims correct finishing. I have looked into the messages, and the failure is in "HTS training: Make Unseen Models" (see line 5152 in build_hts_awb.log.shortened files attached). It is because of strange looking file tree_v_qst001_dur.inf, which is the only content: pau[2].stream[1] "dur_s2_1" for 'awb' voice. Nothing more. The build script, however, continues. The tree_v_qst001_dur.inf files are created (I believe) by HHEd in section 'HTS training: Duration Model Clustering', where the HHEd finishes without any error reported, only with the strange output in the file. What may be wrong? I do not expect a problem be a 64bit platform. All scripts and voice files were also used without any change (except -x switch in the build script and -ai switch for x2x, which, I think, are unlikely to cause the error). And the awb and slt HTS voices are available for Festival (and working), so they were build somehow - and I suppose that by build_hts script. Is the script up-to-date (the last date there is May 2003)? Or is there another preferred way how to build HTS voice for Festival? Any advice will be really appreciated. May I give you any additional information? Thank you very much, Dan Utils used: - Festov-2.1 - Festival-1.96-beta (Festival Speech Synthesis System: 1.96:beta July 2004) - HTK-3.4 (HHEd 3.4 CUED 25/04/06 : $Id: HHEd.c,v 1.67 2008/06/23 09:21:09 zen Exp $); patched by HTS 2.1 - SPTK-3.1 (SPTK: version 3.1) Attachments: - build_hts_awb.log.shortened - zipped outputs of build_hts script for awb voice, lists of utterances print during processing were shortened (not interesting, I suppose) Other files are not attached due to their large size. However, I will send you any other file you wish.
Attachment:
build_hts_awb.log.shortened.zip
Description: Zip archive
Attachment:
signature.asc
Description: This is a digitally signed message part.