In the graph editor you can select the intensity channel for your lights, go to "modifiers" and add "audio channel." This lets you add noise to an channel based on an audio clip and might give you something like what you're after.
How well it works depends on the quality of the sound file, and the dialogue being spoke, as well as the setup you're using.
I'd be inclined to try something where the inner lights were set to control the outer lights via "channel follower," again found under modifiers in the graph editor. Something like as the lights move further out from the center, they "follow" the next light in's intensity by like 90, or 75% or something. Doing this would give you a fall off as the light moves out from the middle so the corners of the mouth aren't as intesense as the center.
Alos you don't have to render it out full everytime, there's a number of options there. Quickshade being one method to greatly reduce render times if all you're doing is rendering out something for timing tests.
The best method tho, since renders don't have audio, is to generate a preview in Layout itself, this is usually pretty quick and plays back your audio with the preview.
|