Hi,
It depends on the subject, but in general:
- I shoot hundreds of photos per object using FF DSLR Cameras
- I light the objects in a controlled, studio lighting, often with cross polarization
- Process the scan data in Reality Capture
- Clean up usually in ZBrush
- Retopologize, unwrap etc in Blender
But it really depends on the subject… For the shotgun for example I needed to have two sets of images. One to capture the texture and one to capture the geometry (using a anti-reflective coating on the metal surfaces).