I don't have to follow a strict script. When a new idea comes to me, I try to make it happen. However, I wonder about the sounds in a KP episode. I don't only mean music playing over the dialogue, which can't always be separated using an AI tool to isolate the speech. No, the sound itself is often distorted. It can even be seen sometimes when using a tool like OcenAudio. This is strange because the picture quality is excellent. However, it seems that compressing the sound (A52 codec, also known as AC3) drastically reduces the quality. Noises are also difficult to work with in dialogues, so AI is overstrained (currently still). Whether there were unwanted noises in the original sound is unknown to me. This means that I can't manipulate the original dialogue, such as making the text spoken slower or faster; the sound quality is too poor for that. Only in exceptional cases does this work.
In short, I can't use the original dialogues the way I'd like, and AI text-to-speech applications are still far from useful. They mostly read the given text as if someone were reading a letter. Professional applications are too expensive for me.
I have many great ideas for the dialogue in my short animation, but I can't use them the way I want to. I think it would be better to have no dialogue than subpar dialogue. I'll try my best.





![Happy Mother's Day [by PhillipThe2]](https://gagala.org/htmly/content/images/20250506034241-KimAnnK001e_cutout.jpg)
