Page 2 of 2

Re: Handling PAFF encoded interlaced h264 files

Posted: Tue Oct 25, 2016 9:17 pm
by admin
Thanks for sharing the clips :-) I will include them in my test bed.

-steinar

Re: Handling PAFF encoded interlaced h264 files

Posted: Wed Oct 26, 2016 12:35 am
by silicontrip
I'm chatting on the vegas forums, I believe that there is a problem because many programs report the FPS as 50 rather than 25. They are trying to convince me that 50fps is correct because it is encoded as separated fields but I think it should be 25fps because that's what went into the encoder regardless of how the file is encoded. Also PAFF encoded files from Adobe Media encoder report the FPS as 25.

I'm trying to use some h264 bitstream analysers on these files to identify the differences but don't know enough about h264 to know if the differences I am seeing are significant.

Mark

Re: Handling PAFF encoded interlaced h264 files

Posted: Wed Oct 26, 2016 6:36 am
by admin
This is close to just being semantics because both 50fps and 25fps can be correct, depending on how you use the interlacing information. Played back on a computer you would normally get 25fps playback but played back on a television set you should get 50fps playback at halved vertical resolution. So if the information you obtain is difficult to make sense of, or is somewhat conflicting, then doing a visual or software based analyze of your media is the only way to determine if it's interlaced or not. Because that's the only thing that really matters.

-steinar

Re: Handling PAFF encoded interlaced h264 files

Posted: Fri Nov 25, 2016 4:02 am
by silicontrip
If it were just semantics, then transcoders wouldn't interpret and process it incorrectly.
I have another file which is interpreted as progressive by FFAStrans yet it is PAFF interlaced and is being converted incorrectly.

https://ninenetwork-my.sharepoint.com/p ... 6f26&rev=1

Would you be able to investigate why this says it's progressive?
Thanks Mark

Re: Handling PAFF encoded interlaced h264 files

Posted: Sat Nov 26, 2016 10:12 am
by admin
Thanks for the file, mark. The next version (0.8.0) will have better mechanism for detecting interlacing and it correctly detects upper field on your files.

-steinar