0

negative frames are played out of order

I believe that RV is not handling negative frames appropriately. It is playing the negative frames in reverse order.

 

I have sequence on disk from frame -10 to 10. The images are named a.#.exr with a 4 padded frame number. For example:

a.-0010.exr, a.-0001.exr, a.0000.exr, a.0001.exr, a.0010.exr

 

In a shell I type "rv a.#.exr". The RV timeline goes from -10 to 10, but the images are actually displayed -1 to -10, 0, 1 to 10 (this can be easily seen by showing the info strip). If I instead specify a range as in rv_manaul.pdf ("a.-10-10#.exr"  or  "a.-0010-0010#.exr"), then RV claims that all of the negative frames are missing files and loads frame 0 in their place.

 

note: this was tested with RV 3.8.4 on CentOS 5

2 comments

  • 0
    Avatar
    Seth Rosenthal

    Hey Andrew,
    I think RV will work properly if the negative sign is considered to be part of the padding. In otherwords for 4 padded frames the negative 10th frame is: foo.-010.exr not foo.-0010.exr.

    You can generate a test sequence easily with RVIO like this:

    >> rvio solid,start=-10,end=10.movieproc -overlay frameburn 1 1 50 -o foo.#.jpg -v

    That will procedurally generate a sequence of black frames with the frame number burned in (you don't need a .movieproc file on disk, that just tells RV or RVIO to use the filename to generate the image or sequence).

    Then rv foo.* or rv foo.-10-10#.jpg works fine.

    However it looks like using the foo.#.jpg is failing to get the frames in the right order. That may be a bug. Thanks for the report. We'll look into it.


    Cheers,

    Seth

     

  • 0
    Avatar
    Andrew Kaufman

    As it turns out, the sequences we typically generate do consider the negative sign part of the padding. My test sequence didn't, but that was user error. "rv a.*.exr" worked fine, although it would still be far preferable to use a.#.exr

     

    Thanks

    Andrew

Please sign in to leave a comment.