Forum Discussion

🚨 This forum is archived and read-only. To submit a forum post, please visit our new Developer Forum. 🚨

8 Replies

  • nosys70's avatar
    nosys70
    Expert Protege
    you can purchase a mobius camera for about 70$, and you will need about 15 to make the equivalent of the samsung.
    So the cost would be $1k just for the cameras. add a bit more for the cable between the sensor and the electronic, plus a bunch of SD card and you will be around 1500$.
    The big difference is the samsung is stitching realtime, and that is invaluable, because with a DIY one, you will need to sync and stitch 7 video files for one eye and do it again for the other eye.

    another way would be to go with less camera (gopro are ok for this, but more expensive) so you reduce the number of files to stitch.
    another way (the easiest) is to get a parabolic mirror and shoot 4k, so you got no stitching and no seams to deal with.
    you loose a bit on vertical FOV, but that's a way cheaper.
    http://www.pano-pro.com/index.php?route=product/product&product_id=50
    http://www.gopano.com/products/gopano-plus#page=technology

    the only problem i got with most of these setup is that usually the mirror is enclosed in a transparent tube and you got a lot of reflection on the tube.
    the pano-pro or gopano is great because it got nothing around the mirror.
    few years ago this technique was limited to pictures because no camera had enough resolution for movie mode.
    Since 4K is available, the deal is different.
  • j1vvy's avatar
    j1vvy
    Honored Guest
    The parabolic mirrors give too narrow VFoV.

    The Mobius cameras suffer from too much rolling shutter to more the camera around. I created a rig
    photocreations.ca/3D/index.html

    I think the iZugar solution is the best because it reduces the number of seams down to 3 and you only have to deal with 3 cameras. You get the plastic shell and lenses for the price.
  • mrowell's avatar
    mrowell
    Honored Guest
    I haven't seen a 3D sample from a GoPro based 3D camera solution that didn't give me a terrible headache. IMO none of them are worth buying, you'd just as well spend your money researching something better to build yourself.
  • nosys70's avatar
    nosys70
    Expert Protege
    yep, if you anyway need to add a lot of money for software and hardware, plus the fact you need huge time to sync and stitch the video, i do not see the point of spendig almost 2k$ to get....almost nothing.
    I am not sure the limited vFov is a problem for most shots, since looking straight up or down is nice, but most of time you just get blue sky (outdoor) or white ceiling (indoor).
    definitely if we can get realtime stitching with calibrated cameras, it would be the solution.
  • j1vvy's avatar
    j1vvy
    Honored Guest
    Every time you see an edge you look immersion. Most panographers went though this process of thinking shooting cylindrical was ok for the reasons you spoke of but the difference between cylindrical and full spherical is a huge jump in quality of the experience. I really hate the mirror balls used to hide the fact that a panographer was too lazy to take a shot to replace where the tripod was. That might only be a 5° hole, but it still distracts from the immersion.

    If you don't want stereo then the Theta is an easy to use inexpensive camera the can record the entire sphere with very little parallax error because the pair of lenses are very close together. But the limitations of 15fps and low res does not make it very appareling for HMD viewing.
    http://theta360.com/

    To get higher resolution at a high frame rate, currently you need to use more cameras. In the future when there are 10k square sensor cameras shooting 90fps you can put a 240° super fisheye lens and capture everything from 60° below the horizon up seamlessly. But you wont like the cost.

    @mrowell have you watch my Fireworks video in a HMD especially after I restitched the footage last week?
    http://photocreations.ca/3D/index.html