Work at SourceForge, help us to make it a better place! We have an immediate need for a Support Technician in our San Francisco or Denver office.

Close

ScanOrder / ScanOrder_Original

Help
2013-09-29
2013-09-29
  • Mark Himsley
    Mark Himsley
    2013-09-29

    Can you confirm the difference between ScanOrder and ScanOrder_Original.

    I have a .mov file containing 720x576 i/25 DV essence. I expect this DV to be BFF, but the .mov container appears to be forcing it to TFF.

    I hypothesise that ScanOrder is the setting of the container and ScanOrder_Original is the setting of the essence.

    $ mediainfo --Full --Language=raw DOGS.mov
    ...
    ScanType                         : Interlaced
    ScanType/String                  : Interlaced
    ScanOrder                        : TFF
    ScanOrder/String                 : TFF
    ScanOrder_Original               : BFF
    ScanOrder_Original/String        : BFF
    Interlacement                    : Interlaced
    Interlacement/String             : Interlaced
    ...
    

    Thanks in advance.

     
  • I hypothesise that ScanOrder is the setting of the container and ScanOrder_Original is the setting of the essence.

    Naming scheme is definitely not the best one, the hypothesis is right.
    Lot of MOV muxers have a problem with scan order and DV (DV specs don't help :( )

     
    • Mark Himsley
      Mark Himsley
      2013-09-29

      Thanks for the confirming.
      :-)

       
    • Mark Himsley
      Mark Himsley
      2013-09-29

      I have an additional question.

      The .mov file in question has a 'fiel' atom with values '2' and '9' (seen in Atom instector).

      page 158 of qtff.pdf states that this file is therefore interlaced with bottom field displayed earliest and top field stored first in the file.

      Surely MediaInfo ScanOrder should be reporting how the fields should be displayed, so why is MediaInfo stating that the container ScanOrder is TFF?

      Thanks in advance again :-)

       
  • The goal of "ScanOrder" is definitely to provide the display order.
    I am reviewing the code, and maybe there was a wrong implementation (mixing "displayed" and "stored"), I check with all my files.
    The MOV muxer you have may actually be right and I may be wrong.

     
  • Actually, none on my files has 'fiel" atom with '2'-'9', and I did not focus enough about the "displayed" / "stored" difference.
    I changed the behavior. patch.

    A sample file would be appreciated, so I could check that all is OK and I could put it in my regression tests.