Glad to be here, very excited to help take this new technology to the next level from the inside of
the media industry
I am mostly in the testing phase right now and pushing for this tech at work, but its slow going, in terms of faces we need to sink 20-30k into a rig so Ive been looking at it for objects, props and the like. Its going pretty good but I have some questions:
1. Lets say I see a cool statues and take a pile of photos, is there a way to avoid any of these steps: point cloud generation for background objects (or is that part of the process, cant be avoided).
I have such images but PS generated a point cloud for the object and massive objects in the background, trees, buildings,etc which makes the mesh generation process intense...
Im at home today so I think masking is turned off in the demo (Registered at work) but how does the masking tool work and when do you use it? Can you use the square one to general isolation the object in the images to help PS optimize the project?
Even if not, once the point cloud is generated there should be a way for me to select the area of interest and only generating the mesh in that area...
2.Shine known down.. another issue with plastic and shiny objects, what are people using to knock down the spec on these objects? Anything food based like a spray that could be washed off so you dont ruin objects with some kind of dull spray paint?
3. Turn table for props... Is this viable or a better idea than physically moving the camera(s) around the object? Does it help for each background to be different?
4. That brings me to background... lately ive been thinking the more complex it is, the better... but to what extreme and what does that do to processing times?
5. My brain is going to explode
Thank you,