Artisan New wrote:
Sorry Klaus that is incorrect....even Kolor says it has parralax issues: this is from the Kolor site:
Parallax errors can appear on close objects as the transitions between the videos are slanted in the middle of the scene. When inclined at 45°, the lenses are oriented horizontally. This REDUCES the parallax errors.http://www.kolor.com/360-video/record
Closed objects can rise issues with 360° VIDEO. That was what Kolor said.
And thatÂ´s obvious: shooting movie/vido bases on different parameters than shooting single pictures - for example you canÂ´t correct CPs in each one of the hundreds of images a movie/video contains . .
Correcting CPs in 6 still-images from a GoPro set definitely is no big deal. . .
Artisan New wrote:
Our Russian freinds know (as in work with) the paralax issues....they can calculate them away(ish) using complex eulerian/carthesian mathematics. Russians are brilliant mathematics and great coders....I've worked with quite a few of them for different companies but they can't bend the rules of physics (only work around them). They only true way of having no parralax would be to take the lenses and shove them all in the NNP but that is physically impossible since they would eat up each others space.
I donÂ´t know whether each one of some hundred of millions of russians are "brilliant mathematics and great coders" - but i know our russian friends are professionals who very well know what they do.
Fact is: it works very fine with their rig of 4 fullframe DSLRs with fisheyes - as photography as well as movie.
Parallax with a good aligned rig only occur in a very close-up range. I tested it mysef - lower than 20cm can cause issues. It CAN - but not necessarily does.
Depends on several items - and is editable in still-photography. In video itÂ´s not editable in the same way. Obviously.