How audio in layers is automatically distributed to frames？

if the sound has its sync property set to event, you don't need any frame after the initial one.  

if it's sync property is stream, you do: minimize the size of the frames, scroll to the right, add a far-right frame. continue scrolling and adding.  after you have a minute or so of frames, copy those frames and paste them serveral times until you have enough frames for your sound.  finally, go back and clear all the keyframes generated by pasting.