My use case with this program might be a bit more niche but I think there is room for wider functionality for many people regarding this feature idea.  I'd like to see an option in which the user can select a second (and maybe third?) audio source for the model to be able to respond to.

In my use case I am using PNGTuber+ to act as the avatar for an AI assistant that listens and responds to the user through TTS and STT. So the avatar's main audio source is a VB audio cable in which is TTS audio is run through to move the mouth and any other movement. What I'd like it to be able to do is select a second audio source it can reference independently so the modal can act as a bit of an audio visualizer, bouncing to music or animating select parts based on that input or whatever, without moving the mouth. Having a third source would be 'nice' as I could set it to react when detecting someone speaking to it or whatever, but that would just be extra.

Anyway, for anyone this would allow independent movement of the avatar with audio without having it be tied to random movement or speech as it is currently.

if this were implemented, a bit more fine control over where in the audio profile the movement is coming from, probably more generally the bass rather then just the volume, would be needed for the best effect.