VideoHelp Forum


Try DVDFab Video Downloader and rip Netflix video! Or Try DVDFab and copy Blu-rays! or rip iTunes movies!
+ Reply to Thread
Results 1 to 4 of 4
Thread
  1. This one is aimed at major, I think.
    How can one tell what pixel aspect ratio a particular file is using other than just making assumptions based on display aspect ratio. I mean, am I just to assume that a file that is 480x480 uses the SVCD 15/11 (or more accurately 6480/4739) rectangular pixels or is there some way to find this out, other than opening the file in quicktime and VLC and eyeballing it (since QT seems to be blind to PAR?)
    If a file is 480x480 with rectangular pixels, one wants to encode it quite differently than a 480x480 file with square pixels (or 768/767) after all.
    Quote Quote  
  2. It seems that you liked the latest aspect ratio posts.

    Unfortunately, there is no method to determine exactly the pixel aspect ratio, other than to make assumptions based on both the source video and the encoded video and to look at the images, and there are lots of possible cases derived purely from standards.

    However a 480x480 video could even have square pixels (this is very unstandard, but it is possible to make one, who knows how insane is the person who authored a particular video?). And even if it is a SVCD NTSC, you still don't know if the source was itself square pixel, or ITU-R BT.601, or anything else, or if it was PAL (which has another aspect), and whether the author has correctly translated the source aspect ratio in the encoded video. If he has not, then the pixels will have even another aspect ratio.

    In other terms, you must master the whole chain from image acquisition to encoding and playback to be able to tell the exact pixel aspect ratio.

    About 15/11 and 6480/4739: 15/11 is the pixel aspect ratio of ITU-R BT.601 NTSC signals (based on their sampling rate) when scaled to SVCD. However, the "industry standard" NTSC square pixel is not exactly square, it is 4752/4739 (not 768/767, that is for PAL). So if you want to convert 15/11 to euclidean square pixels (but few people do that) then you should multiply 15/11 * 4752/4739 and you obtain 71280/52129 or about 6480/4739.
    Quote Quote  
  3. Actually, I didn't even know you had posted anything, or my adventure into the land of SMPTE foolishness might have been a bit more pleasant. Instead I read spec papers and other documents outlining the math involved, and marveled at the obvious flaws in the positions taken by the relevant standards bodies responsible.

    As for the rest of the math. I think I shall now rest my case!

    *sigh
    Quote Quote  
  4. The "latest posts" were the answers to your recent thread.

    It is true that standards are often flawed, but they try to answer the financially-driven need of maintaining compatibility with older standards.

    I read somewhere that the width of a particular piece of the Space Shuttle engines was depending on the width of a container used for shipping it by train. But this was in turn depending on the train wagon width, in turn depending on the train tracks width, in turn depending on old road tracks width, in turn depending on old Rome road width (where horses in double row and chariots circulated), in turn depending on the horses width. So it results that the true origin of part of the Space Shuttle engine standardization units is the old Roman horses' butt size.
    Quote Quote  



Similar Threads