I’ve done this multiple times. I generate around 6 seconds of silence at the beginning of a file which has a few seconds of audio after the silence, then nothing. I then select all and then use the menu - Effect - Special - Repeat. Whether I export the audio or select all - export - selected audio, the new file only has around 5.5 seconds of silence at the beginning, as well as only around 5.5 seconds of silence after each audio section.
This is remedied by adding the half-second to the beginning silence, but why the heck do I have to do this? Is this a known bug? Is there some way to have to avoid the extra work?
My initial thought is you should check to see if you somehow turned snapping on.
Disregard. The system works. This was user error. The problem was that once I did a large repeat to create a 3-minute and 30-second file from a roughly 7.5 second file, there was, of course, much less space between the seconds on the graph for the larger file. When I put the cursor where I thought was just before the audio, I thought I saw around .5 seconds missing, but I was wrong, as I was unable to control the cursor to put it on more dead time before the audio. When I played the file, I saw that the timing had been preserved.
I was going to ask mods to delete the entire thread, but perhaps my mistake can help others in the future.