These are probably already known, but I will post them anyway. I noticed all of these in the current dev stream, by the way.
For both the V1 and V2 voices:
- Sometimes, when Neuro says a message, the subtitles are appended to the end of the previous message's subtitles.
- Sometimes, the first sentence is cleared from the subtitles after Neuro starts the second sentence.
For the V2 voice:
- Subtitles are not formed dynamically (i.e., they are not formed word-by-word as they are spoken).