First of all, the --voice all feature is really cool! Helped me check out the various voices which is the first thing I wanted to try after hearing about the new TTS API 😁
One minor issue I noticed is that the part where the voice is mentioned by name first works inconsistently.
For example, I ran this:
ospeak -v all "This is me!"
And it sounded like:
Alloy this is me!
Echo this is me!
This is me!
This is me!
Nova This is me!
Shimmer This is me!
Running again, and I get different variations of the same issue.
I wonder if the formatting of the voice name with new lines isn't picked up correctly (maybe it looks like dialogues in the training set, where the character names aren't mentioned?)
Or maybe the audio played back to back interfere with each other?
First of all, the
--voice allfeature is really cool! Helped me check out the various voices which is the first thing I wanted to try after hearing about the new TTS API 😁One minor issue I noticed is that the part where the voice is mentioned by name first works inconsistently.
For example, I ran this:
And it sounded like:
Running again, and I get different variations of the same issue.
I wonder if the formatting of the voice name with new lines isn't picked up correctly (maybe it looks like dialogues in the training set, where the character names aren't mentioned?)
Or maybe the audio played back to back interfere with each other?
FYI this is on version 0.2.