I have a 3D model generated from photographs and the structure-from-motion algorithm in PhotoScan. A screenshot of that is here:
PhotoScan screenshot.
photoscan.png (115.08 KiB) Viewed 12902 times
When I export that model (PLY + JPEG) and import it into qCC, it looks like this:
qCC screenshot
cc.png (92.78 KiB) Viewed 12902 times
Why is qCC so fuzzy? The JPEG overlay appears sharp. Is there something I am doing wrong, or is this an implementation issue, in which case, should I file a feature to have sharper photo-realistic overlays as an option?
it's just a guess, but the result in PhotoScan seems to be a textured mesh. The mesh displayed in CC seems to only have an RGB field (and it can't be otherwise as PLY files doesn't provide a standard way to exchange textures, only texture coordinates - see http://www.agisoft.ru/forum/index.php?topic=16.0). If you can work with OBJ format it would be preferable.
However I never encountered a PLY file with texture-related information. I would be greatly interested in it.
Yes the OBJ file makes it look just like it does in PhotoScan.
The PLY + overlay is exported from PhotoScan as a PLY file and a JPEG. Just like the OBJ is exported as an OBJ + JPEG. I assumed CC was loading the JPEG to give it the fuzzy texture, although maybe that information is encoded in the PLY? I can send an example if you like.
From what I read, it seems that it's the responsibility of the user to load the JPG and manually 'associate' it with the mesh (there's no way to encode this in the PLY file).