Using standard Photoscan license I am creating models of complex archaeological rock art cliffside panels. Current model using approximately 500 @ 10 megapixel DNG files. They are segregated into 10 chunks. Some future model areas will be a bit smaller but some will be substantially larger.
aligning each chunk at high quality level - results have been excellent
building point cloud to High level results are excellent but ultimate data set is huge
Have dropped down to medium quality point cloud
Create Mesh at medium quality
Texture mesh using all cameras or mosaic with 4 output files
In all cases the geometry of the 3D model is superlative.
When I merge models from multiple chunks I get superlative alignment and geometry.
The problem is that I really need higher resolution of the details on the cliff face.
I would appreciate any advice on what combination of factors may prove to be most beneficial
? 10 megapixel vs 24 megapixel images - The original 10mp photos are excellent for non model use resolution, we can easily acquire new photos in 24mp. Will increased image file size impact output model size. I don't care about processing time but I do need reasonable size output.
? Any reason not to stick with highest quality camera alignment
? What combination of quality levels/settings of Point Cloud, Mesh and Texture might result in improved texture resolution while maintaining a reasonable size output model.
At this point, smaller output models exported to PDF have been fine. Assumption is that large final models will be exported to OBJ.
Any suggestions would be most appreciated
Thanks and regards