Dear all,
After collecting and analyzing more scenes (saved PPI scenes + images) we found that the problem was switching of images.
For PPI version you need to use the same set of images for REGISTRATION (align) and for creating MESH. If this is not the case then system will do this problem.
A work around is to add all images - both the set of images used to align cameras and the set for creating mesh - and license both sets. Must be all in the same project. If file names are the same, then just put them in different folders.
Reason: RC works as block-chain to implement PPI and dependency chain contains then both images used for alignment as well as for images for meshing. If you provide just one set of images then mesh (or camera poses) cannot be decrypted. The app did not solve this case in UI so far. I’m adding a dialog which will inform you about this issue so that system will notify you about the inconsistent state.
How to avoid this state? .9696 release introduced image layers where you can define image which is used for creating camera poses (mesh geometry) and image used for creating texture. PPI price is calculated using the geometry image.
Please if your issue does not meet what is described here, please contact us.
thank you