Thanks for the reply all. My setup was a) for the coin, a black foamcore background with copy stand side lighting, 120mm lens, and #32 ext. tube, f22 at 1.5 sec., and b) for the camera, it was placed on the copy stand's built-in lightbox, with side lighting still on, and ext. tube removed (had to raise the camera height), same exposure reading. I used an incident meter reading from both the coin and camera positions to the camera lens, so the exposure should not have changed drastically. It's a mystery. I could understand a reflective reading being off because of the backlighting, but not an incident reading from the same positions.