Is this method to determine ITU-R standard DVDs correct?

[Aray]

It's a bit out of date to discuss the issue about ITU-R or non-ITU-R DVDs now.

I'v been read this thread (https://forum.doom9.org/showthread.php?t=81410), and this one (https://forum.doom9.org/showthread.php?t=174533), and Brother John's ITU-R BT.601 and PAR (https://forum.doom9.org/showthread.php?p=826896).

But I have always been confused about whether it was ITU-R or non-ITU-R standard DVDs, and didn't find any simple and accurate method to determine the ITU-R standard DVDs, so that I don't have to crop the pixels by the 704x method and solve the stretched resolution of DVD-Rip easily and completely.

I saw a parameter "Pixel aspect ratio" in MediaInfo by advanced mode, can this be used to determine whether it's ITU-R standard DVDs by comparing it with Brother John's PAR tables?

Based on Brother John's PAR tables, I added the calculated real number, for quick comparison with MediaInfo's "Pixel aspect ratio".


Exact PAR according to ITU-R BT.601
| PAL | NTSC
-----+----------------------+----------------------+
4:3 | 1150/1053 (1.092118) | 38800/42651 (0.909709)
-----+----------------------+----------------------+
16:9 | 4600/3159 (1.456157) | 155200/127953 (1.212945)


Calculated PAR according to ITU-R BT.601
Almost exact and commonly used ITU-PAR according to Jukka Aho’s well-known article.
However he did not account for the fact that the top and bottom half-lines of the analog image are only half-height.
| PAL | NTSC | NTSC (711)
-----+----------------------+----------------------+----------------------
4:3 | 128/117 (1.094017) | 4320/4739 (0.911585) | 72/79 (0.911392)
-----+----------------------+----------------------+----------------------
16:9 | 512/351 (1.458690) | 5760/4739 (1.215446) | 96/79 (1.215190)


MPEG-4 PAR taken directly from the MPEG-4 standard documents.
Very similar to exact ITU figures and usually used for MPEG-4 encodings.
| PAL | NTSC
-----+----------------------+----------------------+
4:3 | 12/11 (1.090909) | 10/11 (0.909091)
-----+----------------------+----------------------+
16:9 | 16/11 (1.454545) | 40/33 (1.212121)


Generic PAR (i.e. ignoring ITU)
This leads to an exact 16:9 DAR for an uncropped 16:9 PAL frame.
| PAL | NTSC
-----+----------------------+----------------------+
4:3 | 16/15 (1.066667) | 8/9 (0.888889)
-----+----------------------+----------------------+
16:9 | 64/45 (1.422222) | 32/27 (1.185185)






For example:

https://res.cloudinary.com/aray/image/upload/v1551573489/Screenshots/Chekist.1992.Rus.D5.PAR.jpg

MediaInfo PAR is 1.067, PAR tables to Generic PAR / PAL / 4:3, means this is non-ITU-R DVDs.


https://res.cloudinary.com/aray/image/upload/v1551573489/Screenshots/Les.Enfants.du.Marais.1999.FRE.D9.PAR.jpg

MediaInfo PAR is 1.422, PAR tables to Generic PAR / PAL / 16:9, means this is non-ITU-R DVDs.


https://res.cloudinary.com/aray/image/upload/v1551573489/Screenshots/So.Close.to.Paradise.1998.D5.PAR.jpg

MediaInfo PAR is 0.889, PAR tables to Generic PAR / NTSC / 4:3, means this is non-ITU-R DVDs.


https://res.cloudinary.com/aray/image/upload/v1551573489/Screenshots/The.Song.of.Sparrows.2008.D5.PAR.jpg

MediaInfo PAR is 1.185, PAR tables to Generic PAR / NTSC / 16:9, means this is non-ITU-R DVDs.



Is this method to determine ITU-R standard DVDs correct?



Also, I found that some DVD's MediaInfo PAR is 1.896, and the DAR 2.40:1, after crop the black bars with MeGUI, the DAR is close to 1.66:1.

How to determine this sort of?

https://res.cloudinary.com/aray/image/upload/v1551573489/Screenshots/Un.Coeur.en.Hiver.1992.D9.PAR.jpg

https://res.cloudinary.com/aray/image/upload/v1551573489/Screenshots/Keetje.Tippel.1975.D5.PAR.jpg

[Sharc]

….Is this method to determine ITU-R standard DVDs correct?

Also, I found that some DVD's MediaInfo PAR is 1.896, and the DAR 2.40:1, after crop the black bars with MeGUI, the DAR is close to 1.66:1.

How to determine this sort of?

To determine the Pixel Aspect Ratio you have to do the circle test. There is no other way.
Small black borders left an right are an indication - but no proof - that the PAR is according ITU. You find it typically on older (earlier) DVDs.

Apparently (as can be concluded from your examples) MediaInfo calculates the Pixel Aspect Ratio according to the formula
PAR = DAR/SAR.

PAR = Pixel Aspect Ratio (calculated by MediaInfo)
DAR = Display Aspect Ratio (e.g. 4:3)
SAR Storage Aspect Ratio = Sampled_Width/Sampled_Height (e.g. 720/576, including borders)

No real DVD has a PAR of 1.896. It's just a calculated figure by MediaInfo from a preprocessed source, I assume. NTSC 16:9 SVCD had a PAR of about 1.82 if I remember this correctly.

[Ghitulescu]

There is no other way than finding a circle (clocks, wheels), a square or another geometrical figure (heads of known persons) that would ease the identification.
The DVD-encoded material may originate from various sources, being subjected to various stages of pre- and post-processing//editing (various combinations of tape/film systems), so that a direct identification or a specific PAR/DAR-value is rarely found. To that one has to add the blanking interval, one finds especially on BBC movies, the 16:9 aspect that increases the visual error and so on.

[Emulgator]

When I was asked to encode DVDs according to ITU I found that Panasonic Players themselves do not respect ITU.
Panasonic players would pad the 704 content up with 16 black pillar bars (8 left, 8 right) and treat is as 720.
They applied only one horizontal resizer for both PARs.
So if studios and authoring houses would have been all aware of that
you should expect to find content matching playback with generic, non-ITU PARs almost everywhere.
But its always better to countercheck with known shapes like circles, as Sharc and Ghitulescu said.
I have to stick to that myself.

[Ghitulescu]

Maybe I am wrong but I think Panasonic is correct: the way I read the ITU, it mentions that the line has 720 samples, but only 702/711 should contain actual video data (not that the rest won't but it may be safely ignored - and anway hidden by the overscan).
IITC their recorders also work with 704 for PAL (I don't have a NTSC one but I have and had several PAL).

For 16:9 I see the numbers are 858/864, and the rest to 854/1024 or even 1050 (BBC) are various timing, blanking and overscan areas.

[SeeMoreDigital]

When I was asked to encode DVDs according to ITU I found that Panasonic Players themselves do not respect ITU.
I seem to remember that the early Panasonic players and televisions handled the aspect ratio signalling in MPEG-2 encoded video streams slightly differently to MPEG-4 encoded video streams with the same aspect ratio signalling.

I don't know if things have changed in more recent times...

[Ghitulescu]

Also part of the problem, and I mentioned it but not detailed, is that the true NTSC (taped, for instance) has no 480 lines but somewhere between 481 and 486, and I think ITU considers these 486 lines as the basis. Therefore, if the perfect 4:3 image is stored in 480, 481, 483 or 486 lines, the number of active pixels on the line has to differ: keeping the DAR must change the PAR and vice-versa. In the past, all what one needed to do is to adjust the vertical sweep of the monitor screen (in studios) and hope the grand public won't notice this small error.

Therefore, it plays a crucial role what exactly happened to the video before reaching the encoder, and how the encoder was set (for instance, one can set the encoder to ignore anything above 480 scanlines).