- This topic has 1 reply, 2 voices, and was last updated 4 years, 8 months ago by .
Viewing 1 reply thread
Viewing 1 reply thread
- You must be logged in to reply to this topic.
› Forums › Speech Synthesis › Festival › Levels of stress
When I look in the unilex dictionary, in the utterance structure of the database sentences (in the utt-folder) and also when running sentences through the so-called Korin’s script, I see four levels of stress. E.g. (“grandnephew” nn (((g r a n d) 1) ((n e) 2) ((f y uu) 3))) and (“accentuated” vbd (((@ k) 0) ((s e n) 1) ((ch uw) 0) ((ei t) 2) ((i d) 0))) Does this mean that at synthesis time the unit selection in Festival distinguishes between these four levels (giving, in theory at least, eight possible stress combinations per diphone)? Or are some of these stress levels somehow collapsed? And if they are collapsed, is it possible to know if level 2 (and 3) is collapsed into level 0 or into level 1?
The target cost function collapses all levels of stress (1,2,3) into a single level (1 = “stressed”).
Some forums are only available if you are logged in. Searching will only return results from those forums if you log in.
Copyright © 2024 · Balance Child Theme on Genesis Framework · WordPress · Log in