Loading ...
Sorry, an error occurred while loading the content.

Action stereo spherical panos -- sync issues

Expand Messages
  • panovrx
    A while back I put two Canon5D/10.5mm together (one on top of the other inverted) at 180 apart to make 360*130 degrees timelapse sequences and this works
    Message 1 of 2 , Aug 3, 2008
    • 0 Attachment
      A while back I put two Canon5D/10.5mm together (one on top of the
      other inverted) at 180 apart to make 360*130 degrees timelapse
      sequences and this works pretty good sync wise in the overlap edge
      zone if the frame rate is
      slow enough -- eg. they will stay in sync for about 100 frames or so
      at 1 fps -- being fired by wired together releases. At 3 frames per
      sec (JPG) they will stay in sync for at least 25 frames. So I thought
      I would try spinning the 2 cameras in stereo camera mode ie pointing
      in the same direction -- at about 8 sec per revolution -- which
      would be short enough maybe I thought for one 360 in sync for a
      stereo action panorama.

      Here is about 180 degrees of a stereo action equi
      http://www.mediavr.com/chinatownanss.jpg
      (for crosseye viewing)
      -- the rest of the pano has crowd blending (mono) stitching issues

      The sync when it is working is good -- < 2 or 3 msec I think.

      I think this is a practical idea and stereo action panoramas will be
      possible with digital slrs (they are certainly possible with film-
      based stereo panorama cameras but they cant do spherical) if I can
      build a very solid rotator that will keep the cameras aligned with
      each other and the rotation axis accurately throughout the rotation.

      If you are not worried about sync (ie no action) standard
      pano methods with a single camera shifted for the 2 views work fine
      I think with say < 10 degrees rotation per shot if subjects arent
      closer than 2m say for standard eye separation camera separation.

      For really accurate stereo panoramas of close subjects with a
      digital slr I think the strip assembly method I described recently
      would be better and certainly easier to stitch though longer to shoot.

      Peter Murphy
      www.mediavr.com/blog
    • Roger D. Williams
      ... I am impressed. The stereo panorama looks fine, so the system is clearly viable (if a bit pricey from where I sit...). ... I agree; that accords with my
      Message 2 of 2 , Aug 3, 2008
      • 0 Attachment
        On Sun, 03 Aug 2008 19:57:02 +0900, panovrx <mediavr@...> wrote:

        > A while back I put two Canon5D/10.5mm together (one on top of the
        > other inverted) at 180 apart to make 360*130 degrees timelapse
        > sequences and this works pretty good sync wise in the overlap edge
        > zone if the frame rate is
        > slow enough -- eg. they will stay in sync for about 100 frames or so
        > at 1 fps -- being fired by wired together releases. At 3 frames per
        > sec (JPG) they will stay in sync for at least 25 frames. So I thought
        > I would try spinning the 2 cameras in stereo camera mode ie pointing
        > in the same direction -- at about 8 sec per revolution -- which
        > would be short enough maybe I thought for one 360 in sync for a
        > stereo action panorama.
        >
        > Here is about 180 degrees of a stereo action equi
        > http://www.mediavr.com/chinatownanss.jpg
        > (for crosseye viewing)
        > -- the rest of the pano has crowd blending (mono) stitching issues
        >
        > The sync when it is working is good -- < 2 or 3 msec I think.
        >
        > I think this is a practical idea and stereo action panoramas will be
        > possible with digital slrs (they are certainly possible with film-
        > based stereo panorama cameras but they cant do spherical) if I can
        > build a very solid rotator that will keep the cameras aligned with
        > each other and the rotation axis accurately throughout the rotation.

        I am impressed. The stereo panorama looks fine, so the system is clearly
        viable (if a bit pricey from where I sit...).

        > If you are not worried about sync (ie no action) standard
        > pano methods with a single camera shifted for the 2 views work fine
        > I think with say < 10 degrees rotation per shot if subjects arent
        > closer than 2m say for standard eye separation camera separation.

        I agree; that accords with my (limited) experience.

        > For really accurate stereo panoramas of close subjects with a
        > digital slr I think the strip assembly method I described recently
        > would be better and certainly easier to stitch though longer to shoot.

        Yes, fascinating! Thank you so much for this information. It is a real
        encouragement.

        Roger W.


        --
        Work: www.adex-japan.com
        Play: www.usefilm.com/member/roger
      Your message has been successfully submitted and would be delivered to recipients shortly.