I was a bit sleepy when I wrote that response. I guess I just assumed I was talking to someone who wanted to get the unique SSE audio tracks running with the 2.7 video, because that’s kind of something that makes sense. And my point was just that imperfectly matched run times of video and audio streams are fine, so no need to mess with 2.7’s video.

But it was indeed about using 2.7’s audio with SSE, which is of course more fitting with this thread but a more confusing idea. In that case, trimming the video is unnecessary for the reason I already mentioned, and the fact that you’re not interested in the video anyway. But you also wouldn’t have needed to trim the audio as it didn’t even extend to those 30 seconds.

It sounds like worked despite the extra trimming step, but it would have been fine using tsmuxer alone. I’m not sure how the trimming was done, but hopefully the audio wasn’t re-encoded or anything during the process.