Vortex 3D Fulldome Stills

Posted by on Jun 6, 2012 in Fulldome, Stereoscopy, Vortex | 2 Comments

A few stereoscopic test stills from Vortex. These are low res and quality but you’ll still need to click on the images to see them properly. On a standard monitor these will look very subtle in terms of depth, but when stretched over a 20 meter dome the effect will be very dramatic. Keeping the positive parallax in safe non-divergent values can be tricky. I don’t currently have a proper mathmatical way of doing this… Something I need to work on.

I’m using 360 degree panoramic stereoscopy so the stereo effect can work for any part of the dome. The pinch-point is quite high in the dome for most of the film so most of the 3d is easy to see on a flat screen. The stereo tilt takes into account the focal point of the vortex and uses this as an approximate line of parallex. I guess the words used to describe these things don’t really exist so it’s hard to explain what’s going on easily. I’ve been using the term ‘pinch-point’ to relate to the point around which the stereoscopic parallax rotates. This could also be called the pole, up-node or zenith… All of which could be confused with other terms. The term pinch-point is really describing the spiral or vortex like pinching that happens when the stereo separation is strong at the zenith of the stereo effect. Really I guess this effect should be removed thus leaving no ‘pinch-point’, but as this is the point that in theory no-one should ever be looking at, it doesn’t really matter so much.

4k tests coming soon.

2 Comments

  1. JDMulti
    February 13, 2015

    Interesting. Did you manage to remove the “pinch-point” in your stereo panorama renders? I’ve got the same issue and can’t get around how to fix this. Do you have some ideas?

  2. Aaron
    February 16, 2015

    Hi,

    It is possible to remove the pinch-point using maps of the separation parameter and the head-turn parameter. Actually the ones that come with the Domemaster 3D install do a pretty good job for zero tilt footage. However, it can still be tricky to remove any stereoscopic violations in the area around the panoramic zenith point (changes based on tilt) depending on how strong the negative parallax is in that area and the shape of the objects.

    The aim is to have no vertical disparities for the most likely horizontal position of the viewers head, so it cases where the viewer can explore the frame at all angles, this can be very hard to control. In these situations it’s probably best to flatten the overall separation and use a map to flatten the panoramic zenith completely. Or possibly consider repositioning objects around the point so they are close to the zero-parallax sphere.

    If you have a targeted scene, where the viewers attention is heavily controlled you can usually push the tilt angle to a position out of sight and achieve very good stereo depth in the area of attention.

    Hope that helps.

Leave a Reply