I'm learning how to use PS. How does this sound for my workflow so far?
Its a test of concept for a vfx set extension interior of a Edwardian market in london.
1) imported 650 photos of an interior, added lens
2) aligned
3) added 40 markers - markers on non-key photos are all in the matching position - so assume all is good.
4) 'align selected cameras' - for all cameras.
5) build dense cloud
6) tried to optimize but it failed so skipped this
7) tried to build mesh - after 4 days i was getting render times of two weeks at medium so i stopped it. (2.9 i5 proc / 24gb ram)
7) duplicated the main chunk (250,000 points), keeping all markers, and broke it up into overlapping 7 chunks of between 50,000 & 70,000 points. (took 6hours)
I am then going to:
8 ) align chunks (marker) & merge chunks (no scale fix)
9) texture.
When running the build mesh i assume whether a camera is there or enabled makes no difference as its just building a mesh from the point cloud?
ta .rafal
www.movingimagearts.com