message from Antje Schweitzer <antje.schweit...@ims.uni-stuttgart.de> to
festival-talk
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Dear all,
we are trying to build a clustergen voice for German. We have
successfully built one before, but this time, we are observing that
labels are doubled in the alignment process, i.e., we have correct
sequences of segments in prompt-utt/*.utt and prompt-lab/*.lab but
occasionally segments are doubled in lab/*.lab after labeling.
Interestingly, it's always "S" fricatives that are doubled...
This is problematic because it causes alignment mismatches in building
the utts afterwards.
Can anyone help me here? Any idea what could be causing this??
Thanks very much!
Antje
--
Antje Schweitzer
IMS Uni Stuttgart
0711-685 81376
http://www.ims.uni-stuttgart.de/~schweitz
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
= University of Edinburgh's Festival Speech Synthesis System =
= http://festvox.org/festival Sent Via festival-t...@festvox.org =
= To unsubscribe mail majord...@festvox.org =
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
_______________________________________________
Festlang-talk mailing list
Festlang-talk@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/festlang-talk