Ticket #1598 (closed enhancement: duplicate)
HLS: investigate Zencoder support
|Reported by:||jeroen||Owned by:||jeroen|
|Cc:||sanil, pablo||Forum thread:|
Description (last modified by jeroen) (diff)
Got an email from Justin @ Zencoder. We should check out the test stream again and see if we can provide more info:
I'm not sure who got back to you on this, or if anyone actually did. I'd like to make our outputs (in optimized mode) compatible with JW Player, and was looking at the source of the adaptive branch a bit...
Off hand, I'm not seeing anything that should cause issue in our streams. Looking at this ticket: http://developer.longtailvideo.com/trac/ticket/1224 (which 1545 was marked as a dupe of), it talks about PES data crossing ts packet boundaries, and that being signaled by CC not being zero.
A couple notes on that: First, I don't think we spread the PES data between packets... Each stream in a new segment starts with a packet that has the payload_unit_start_indicator set. (You call it "stt" in the adaptive playback code, if I'm reading things right.)
Also, Apple's segmenter doesn't begin segments with packets that have the CC set to zero, so it doesn't sound like this is the exact issue? If you could provide some more specific details on the errors you're getting during playback, and/or how our streams differ from Apple's, I'll look into whether there's anything we can tweak to help get it working (either on our side or in the player).
We have the following streams in our adaptive testset that use Zencoder:
- Status changed from new to closed
- Resolution set to worksforme
- Status changed from closed to reopened
- Resolution worksforme deleted
- Component changed from flash to streaming