This is nothing serious, but maybe the value reset could be avoided if I accidentally launch it when an incorrect layer is accidentally selected? When I initialize the script when one of these folders is chosen (instead of the main switch layer), the script somehow 'crashes' (it doesn't popup and the values it contains are reset the next time I start it). I have a switch layer where each switchable mouth is a folder (containing an image and a bone structure). Here are some issues I've found so far (besides the amazing speed in workflow that this script provides!!! ) Maybe that's more than you wanted to know, but I hope it answered your question. It turned out that start and end frame was easier and more accurate for a wide range of audio. The idea was that you only needed to put in the starting frame. I originally put it a control to set the length for vowels. So 'noooo waaaaay' would extend the 'o' and 'a' sounds (I'd actually spell it 'nOOOOOO wAAAAA' for long 'o' and 'a' and no real 'y' sound and use the phonetic approach) you can use '-' for the rest mouth position.you can repeat letters (typically you'd do vowels and character like s and z) to extend them.stops (ie consonants that are not like s, z, v, f etc) are assigned only one frame.it knows the difference between characters that can have extended sound and ones that do not.It does space space out the phonemes based on the number of frames with the following exceptions It worked ok, but for a number of reasons correlating to specific phonemes is more complex.ДЇor this approach I took a different tack and focused on getting the appropriate phonemes in order. It even had tolerance controls, to help identify local maxima and minima. I had written a script that located the maxima and minima of the audio. Breinmeester wrote:Does it analyse the sound file to determine the timing of the phonemes?

Lip sync in anime studio pro 10