Non-HD extras. PAR and fps detection

[ncahammer]

I have a Bluray that has some extras with SD resolutions.
Executing eac3to for those, I get :

M2TS, 1 video track, 1 audio track, 1 subtitle track, 0:02:47, 24p /1.001
1: VC-1, 480p24 /1.001 (3:2) with pulldown flags
2: AC3 Surround, English, 2.0 channels, 192kbps, 48khz, dialnorm: -27dB
3: Subtitle (PGS), English
[v01] The video bitstream framerate field doesn't seem to match the timestamps. <WARNING>


I have two questions regarding this

1) They seem that framerate differs when I use eac3to and DGIndexNV/DGDecodeNV
In the above video eac3to reports that the video is 23.976, but the .avs script (using DGSource) see it as 29.97 fps

2) What is the PAR of those ? I am not sure that's 1:1 as the HD videos, if not, how can I tell if it's ITU or not, 8:9 or anamorphic 32:27 ?

I did try to find any spec or Bluray authoring tool that has answer on how SD are stored in a Bluray disc, but found nothing so far

[nurbs]

1) is caused by the pulldown flags. eac3to removes them when it's soft pulldown (flagged) which is what most people would do. If you keep them it's 29.97.

[Inspector.Gadget]

That one IS 1:1, according to eac3to. Check it with MediaInfo, though. Also, MPC-HC doesn't display anamorphic VC1-in-M2TS properly, so that's not a reliable guide.

[ncahammer]

1) is caused by the pulldown flags. eac3to removes them when it's soft pulldown (flagged) which is what most people would do. If you keep them it's 29.97.Hmm you are correct, it seems that DGDecodeNV removes the pulldown, but leaves the framerate unchanged to 29.97 and the video has now a duration of 2:13 instead of 2:47 which is the correct. It's easy to correct this. Thanks!

That one IS 1:1, according to eac3to. Check it with MediaInfo, though. Also, MPC-HC doesn't display anamorphic VC1-in-M2TS properly, so that's not a reliable guide.
What did you read that made you certain that's 1:1 ?
The "(3:2)" ? What would that be, if it was anamorphic ?

About mediainfo, I am not sure,

Video
ID : 4113 (0x1011)
Menu ID : 1 (0x1)
Format : VC-1
Format profile : AP@L2
Duration : 2mn 46s
Bit rate : 4 163 Kbps
Width : 720 pixels
Height : 480 pixels
Display aspect ratio : 16:9
Frame rate : 23.976 fps
Standard : NTSC
Colorimetry : 4:2:0
Scan type : Progressive
Scan order : 2:3 Pulldown
Bits/(Pixel*Frame) : 0.502
Stream size : 82.7 MiB (91%)

Which IMO is wrong. It's not 16:9 for sure
Here they are two screenshots from this video

http://img509.imageshack.us/img509/2847/69129001.th.png (http://img509.imageshack.us/my.php?image=69129001.png) http://img80.imageshack.us/img80/4184/85192667.th.png (http://img80.imageshack.us/my.php?image=85192667.png)

They "look" that 1:1 is the correct answer

[Guest]

Hmm you are correct, it seems that DGDecodeNV removes the pulldown, but leaves the framerate unchanged to 29.97 and the video has now a duration of 2:13 instead of 2:47 which is the correct If so it is a bug. I will look into that and fix it ASAP.

May I please have a stream sample to test with? This is the first VC-1 stream with pulldown that I have ever heard about.

[nurbs]

IIRC the VC1 streams on HD-DVD generally had pulldown flags.

[ncahammer]

If so it is a bug. I will look into that and fix it ASAP.

May I please have a stream sample to test with? This is the first VC-1 stream with pulldown that I have ever heard about.It seems that tsMuxeR can't give proper output for this. I have to upload it all and it's 90,4 MB. It's ok, that with you ?

[Inspector.Gadget]

MediaInfo is right. 3:2 is of course a 1.5 DAR, which with 720x480 is a 1:1 PAR. This is what eac3to thinks, anyway. In reality, to display that content correctly (as MediaInfo tells you), you must apply a 16:9-appropriate PAR. In this case, the non-ITU display size would be ~853:480, while the ITU display size would be ~875:480.

Edit: To prove that it is in fact 16:9, resize the JJ Abrams head shot to be 853:480 in your favorite image editor. See how he looks human now? :cool:

[Inspector.Gadget]

@neuron2: If it would be helpful, you can check out vc1conv for example code that strips pulldown from both HDDVD and BD VC1 content. This link is the binary and source: http://www.w6rz.net/vc1conv.zip

[Guest]

I know what to do, I just don't have any samples.

[ncahammer]

Edit: To prove that it is in fact 16:9, resize the JJ Abrams head shot to be 853:480 in your favorite image editor. See how he looks human now? :cool:Yes, you are correct, non-ITU NTSC anamorphic (32:27) is the correct one. All players I did try, fail to realize that and play it incorrectly.

Is that info stored inside the .m2ts file ?

@neuron2: http://rapidshare.com/files/306025833/00065.m2ts.html

[Inspector.Gadget]

Is that info stored inside the .m2ts file ?

Yes. That's why MediaInfo can read it without the benefit of clip or playlist files.

[Guest]

@neuron2: http://rapidshare.com/files/306025833/00065.m2ts.html RapidShare???

"This file can only be downloaded by becoming a Premium member"

Thanks but no thanks. Can you please upload it to mediafire.com?

[ncahammer]

RapidShare???

"This file can only be downloaded by becoming a Premium member"
Be patient, try in a few mins, again...
Currently a lot of users are downloading files. Please try again in 2 minutes or become a Premium memberI did 5 tries, one was successful. I must leave in a few mins and it took me almost 30mins to upload it. I'll be back in a few hours

[Guest]

I think you need to read my message more clearly. I am not going to pay to become a "Premium Member".

Please upload it to mediafire.com.

Thank you.

[ncahammer]

Sorry about that. I do understand your frustration. I now logging from a different place I get the same message as you. RapidShare sucks big time !!
Anyway, here it is :
http://www.mediafire.com/?yz0mjyzvyrm
Hope that I helped !

On the AR part : Is any tool that will report accurately the AR flags of a video stream inside .m2ts. That 16:9 of Mediainfo, doesn't help much. Is at least anywhere any doc for the m2ts file format, that will help me, on how I can retrieve that info ?

[Guest]

There is no pulldown in the sample, which is consistent with the statement that EAC3TO removes it. But EAC3TO has not adjusted the frame rate down to film rate as the sample specifies a rate of 29.97. So this has nothing to do with DG tools.

Can you give me the untouched original M2TS, not passed through EAC3TO?

Can you tell EAC3TO to preserve soft pulldown.

[Inspector.Gadget]

-keepPulldown should work for that purpose.

[ncahammer]

There is no pulldown in the sample, which is consistent with the statement that EAC3TO removes it. But EAC3TO has not adjusted the frame rate down to film rate as the sample specifies a rate of 29.97. So this has nothing to do with DG tools.

Can you give me the untouched original M2TS, not passed through EAC3TO?

Can you tell EAC3TO to preserve soft pulldown.This is an unproccesed file, directly from the Bluray. When I run the .avs script (with the DGSource) it reports it's 29.97 and a wrong duration. Both EAC3TO and .avs report that is a ~4000 frames video. As I already quote the EAC3TO reports that's an 23.976fps video, just type "eac3to.exe 00065.m2ts" to verify it

[Guest]

OK, I haven't added pulldown detection for VC1 yet. I am working on correct pulldown handling for all three video types and this will be fixed when I release that in the next few days. Thanks for pointing it out.

EDIT: For now you can just do AssumeFPS(23.976) in your script.