cleaning a .ts with multiple video pids

[ryle]

Hi

I'm recording some 4k footage that is broadcasted on Eutelsat 10°E.
In fact, there is no 4k video pid but there are 4 video pids wich each represent one quarter of the 4k video.

I try to use avisynth (with ffms2 to access the diffferent videos) to stack them all in one video but I run into these problems:
-the pids for the 4 videos seem to change between each record I make (my avisynth script has to be modified between each try)
-the 4 videos don't seem to start exactly start at the same time, making the final video unwatchable

When I try to use TS-Doctor, it only wants to work on one video PID.

Will a future version of TS-Doctor manage to clean and keep all the videos in synch  (without asking me to export only one video) ?

[Cypheros]

Sorry, 4k is brand new. If I find the time, I try to investigate this new format. I guess there is no real transport stream standard yet.

[Djfe]

who gets such crazy ideas like putting one video into 4pids just because of the high resolution
especially if the pids change from video to video how is the receiver supposed to decide which one to put where?

and why can't the broadcast station just use HEVC aka h.265 for it?
you have to use a new receiver anyway so they could just implement the new standard/wait for it to become supported/popular instead of making their own unsupported, complex type of a stream
divx for example has build the first, well working encoder already, including decoder for Windows and a browser plugin (highest resolution is still Full HD and it is more like a test build but still)
you can build VLC from sources with an additional parameter to add the functionality to encode in HEVC
and MultiCoreWarex is working on x265 right now

[Cypheros]

I guess it's a workaround to get enough bandwidth until a new transmission standard is established.

[Djfe]

do they use several transponders to make this possible?
I don't think that 4 h.264 streams together have less bandwidth than one h.264 stream with a 4k resolution and h.264 level 5.1 or 5.2, right?

[Mam]

do they use several transponders to make this possible?
I don't think that 4 h.264 streams together have less bandwidth than one h.264 stream with a 4k resolution and h.264 level 5.1 or 5.2, right?

Wrong 

[ryle]

who gets such crazy ideas like putting one video into 4pids just because of the high resolution
especially if the pids change from video to video how is the receiver supposed to decide which one to put where?

and why can't the broadcast station just use HEVC aka h.265 for it?
...

There are different tests on different satellites, some in 4k / h264, others in 4k/hevc and they seem to work ok on pc when you've got the good setup (codecs, player ...).

For this particular test (Eutelsat 10°E, 11345 / H / 27500), there are 4 video pids, they are labelled (TOP_LEFT, BOTTOM_LEFT...) so I think that a receiver or a special software can identify them easily, this is not the problem itself.

I want to be able to see those 4 quarter of picture on one screen in only one software.

The only technique I thaught of for the moment is using avisynth (needed to use the StackVertical/StackHorizontal functions) and the ffms2 plugin (needed to be able to import different video PIDs for each quarter).
It seems that the ffms2 plugin may remap video pids in a different way than the original one or what other softwares could read, hence the "different captures, different pid order".

Potplayer manages to read the .avs file and display the four quarters but I usually get 2 out of 4 quarters that aren't synched with the others, so I'd like to use ts-doctor on the captured ".ts" so it may resynch the 4 video pids.

[Mam]

I think you should not pay too much attention and waste too many efforts on these tests.
They are just a promo for the new displays, take 4 normal HD receivers, tune them to the pids each, connect their HDMI outputs to the correspondent input of the display (which also features 4 of them now because the current HDMI bandwith is much too slow for 4k) and just WATCH.

None of these will stay for long. The inputs will merge into one faster HDMI type soon, the transponders will become rearranged (or, more likely, new sats with different standard transponders will be launched, where THERE IS a new standard someday) and the receivers will change their format too.

Everything you will do until then is most likely to be wasted time/money/both. Who buys a 4k TV set now should know that he has lost already. If not, he is even more stupid and beyond help (for those guys I could supply a bank account number that greatly would accept their money that they want to waste. Not that they will ever get more than NOTHING, but at least, its spend for a good purpose (for ME  )

[ryle]

Hdmi 1.4 is ok for 3840*2160@30Hz so many of the 4k content presently available can be displayed on the first "4k" models that are available.
I know that the future is with hdmi 2.0 and refresh rates higher than 30Hz.

I just wanted to try and display this current test (which has good quality and is @ 50Hz by the way) on any screen and on 4k displays I have access to.

[Cypheros]

There is no 4K receiver available right now, as far as I know.

[Djfe]

the only time it would make sense to use several video pids is if you are streaming in 3D, I think (instead of using just one video pid which is split up)