SAR , PAR , DAR - some practical questions

[Ennio]

I only watched the first video. Other than looking for circles, at the wider shot where the cowboys are all standing together (at like 0:00:14.000), the people appear to me a bit "thin". I'm no vet, but also I wouldn't be surprised if to some people the horses look like as they've hit a tree during gallop.
As such, the image could use a wee horizontal stretching. So, when having to choose between the two given values, here my bet would be on 12/11.


As a funny side note, imagine I almost copy/pasted this post into the reply-box while not noticing I forgot the "r" in the "first video" phrase. Which surely would have upset some people, or at least made them wonder what the hell this was all about

[Skiller]

Frankly I don't know as I didn't find a nice circle for 4-3 (1).m2v, but I would vote for 16:15 for this clip.

And a second example 4-3 (2).m2v with circles where the winner is 12:11 as I see it.

I would agree for both.

I find that the year of the production (the media, not the movie) and the transfer can be good hints.

In case of a DVD version of an HD transfer of a movie, chances are it's 16:15.
If it's an older DVD, before there was any HD transfer of that movie, 12:11 is much more likely.

(2) Looks like an older transfer, whereas (1) looks like a DVD version of an HD transfer to me (very detailed, perfectly clean frame borders).

[Sharc]

Noteworthy perhaps that ffmpeg and some other tools which report a PAR (Pixel Aspect Ratio) aka SAR (Sampling Aspect Ratio) tell for both files that it is 16:15, which is obviously wrong for the second clip 4-2(2).m2v.
I think it's just based on the known facts that the files are mpeg2, 720x576, DAR flag=4:3 which then defaults to 16:15, or calculated as 4/3/720/576 = 16:15.

I am not aware of any tool which can really evaluate ("measure") the true PAR (aka SAR) of the video and report it correctly, unless it is correctly written in the stream (metadata) or container. The tools have no eyes.

It's not terribly important anyway. As one can see from the first clip 4-3(1).m2v it's not eyecatchingly easy to decide on a possible 2.4% error.