It is a pity that

[Rolf-100R]

It is a pity that:

1. There are no tools to prompt for the correct stress placement when pronouncing a word. In my language, stress plays a primary role. Sometimes it can take an entire hour to get a decent result, even though the video itself is generated quite quickly (on Distilled GGUF Q6_K).

2. It's impossible to influence the voice timbre.

3. There are no markers to indicate people. When two characters are involved, they often intercept each other's lines, or even both simultaneously voice the same text.

4. The program's developers maintain a proud silence and rarely respond to user questions.

[Rolf-100R]

It's been two hours, I've created over twenty videos, and every single one of them has some pronunciation error—word stress is incorrect. So, everything's gone in the trash. Isn't it possible to come up with some kind of symbol—*, #, !—that can be placed directly in a word so LTX knows which syllable to stress when generating it?