What would be the smartest process method to detect and create a layer of breaths only, from very clean dialogue?
This method may help: https://youtu.be/78UsfeW-MKY?si=n-pH6FStr8U_ssPH