I’m having a hard time aligning set of stills taken of a person on a turntable at several angles, captured from a fixed rig comprised of three cameras on a vertical axis.
Basically instead of shooting around the subject the cameras are fixed and the subject is rotating.
I have distance and measurements at hand and I feel I may need to tell Reality Capture that the subject is rotating, I just don’t know how.
Currently the alignment fails, as expected. How should I set Reality Capture up for this?
Capturing human being as well as any alive object that breath, have hearts that push blood 60-140 times per minute through whole body, etc. are not recommended using one camera. Just because when you will capture such object you will spend more than one minute, but probably even more, and no one, even specially trained actors can freeze their body especially all this micromovements that all are exist.
But all this micro movements are can highly affect alignment precision and as result give errors in mesh.
That’s why are expansive multi camera rigs are exist. 54-124-244 DSLR or mirrorless, Red or industrial cams. That can trigger shutter in same moment for all this cams in less than 1/250sec to freeze moment.
But with some practice you can try to capture head or even body iwoth only one camera, just not expect multicamera rig quality.
If you sure your capture on turntable made fast enough, you should try mask all background from image, because this can be a reason why you can’t align.