Bug #1655
closed
PAT/PMT generation with passthrough muxer without libav is broken also
Added by Eric Valette about 12 years ago.
Updated about 12 years ago.
Found in version:
git today
Description
if you look at ts files in https://www.lonelycoder.com/redmine/boards/5/topics/7845 you will see that we do not have the PAT/PMT generated each 1000 packet as expected, that the TS recorded file does not start with the PAT packet.
My TV does not like the TS generated by new TVHEADEND either.
Simply record a file is passthrough mode and run TSINFO on it.
What was happening was that the passthrough muxer was injecting the PMT packet first, followed by the PAT.
I've just pushed a change to master which swaps the order. Does this make your TV happier?
Dave Chapman wrote:
What was happening was that the passthrough muxer was injecting the PMT packet first, followed by the PAT.
I've just pushed a change to master which swaps the order. Does this make your TV happier?
Yes the write order was swapped. But that does not explain the count?
Eric Valette wrote:
Dave Chapman wrote:
What was happening was that the passthrough muxer was injecting the PMT packet first, followed by the PAT.
I've just pushed a change to master which swaps the order. Does this make your TV happier?
Yes the write order was swapped. But that does not explain the count?
tvheadend outputs the PAT/PMT, then 1000 media packets, then the PAT/PMT, then 1000 media packets...
Dave Chapman wrote:
Eric Valette wrote:
Dave Chapman wrote:
What was happening was that the passthrough muxer was injecting the PMT packet first, followed by the PAT.
I've just pushed a change to master which swaps the order. Does this make your TV happier?
Yes the write order was swapped. But that does not explain the count?
tvheadend outputs the PAT/PMT, then 1000 media packets, then the PAT/PMT, then 1000 media packets...
Problem is that tsinfo output says something else. https://www.lonelycoder.com/redmine/boards/5/topics/7845 "Found 10 PAT packets and 4 PMT packets in 5000 TS packets." Or the output of tsinfo is wrong?
Yes, as discussed in the forum, the tsinfo output is wrong.
Can we now close this issue?
Dave Chapman wrote:
Yes, as discussed in the forum, the tsinfo output is wrong.
Can we now close this issue?
You can close it but I haven't finished testing with the TV. I would like to record channels with EAC3 5.1 audio and make sure it works also. I will probably manage to finish testing over the week-end.
Eric Valette wrote:
Dave Chapman wrote:
Can we now close this issue?
I will probably manage to finish testing over the week-end.
Its working so you can close the bug. Thanks for the support.
- Status changed from New to Fixed
Also available in: Atom
PDF