Tips for macro photogrammetry that nobody talks about

I tried the stacking route last year with some coin-sized stuff and ran into a wierd problem. My stacker would shift the framing a tiny bit between frames, probably a pixel or two, and that was enough to throw off the alignment in RealityScan. Every model came out with fuzzy edges where the software couldn't decide if the surface had actually moved. Took me a week of head scratching before I figured out it wasn't the scanner, it was the stacking software re aligning each frame slightly differently. These days I just stop the lens way down and put up with softer images. I'm shooting with an old 60mm macro on a tripod, f/22 if I remember right, and I let the shutter sit for a couple seconds with a delay timer so nothing shakes. The photos aren't as crisp as a stacked frame but the geometry comes out cleaner and the whole capture takes a third of the time. For lighting I ended up doing almost what you describe but with two cheap gooseneck desk lamps and baking paper taped over them. Moving them between bursts made a huge difference on a polished brass piece I was fighting with, the reflections used to look like dents in the mesh.

The turntable masking tip lines up with what killed my first few attempts too. I scanned a grey plastic lid right along with the object and the base of everything came out domed until I started painting the lid matte black.

This thread takes me back. Before stacking software got good I used to do macro scans with a focus rail, nudging the camera forward a hair at a time and averaging the sharpest frames by hand in an editor. Slow as molasses but the geometry held up. One trick nobody talks about anymore: shoot a paper target with printed markers around the turntable and let the solver pick up those points for alignment. Helped me plenty on shiny little parts where the object itself barely had any texture to lock onto. Does RealityScan still choke on featureless surfaces the way the old tools did, or is it better now?

For really tiny stuff I quit using a turntable altogether. I hang the object from a thin elastic thread off a bent coat hanger and shoot it from the side while it settles. No base to mask, no spinning surface for the solver to grab, and you can get low angles you never get with a flat lid. Weights or anything dense enought just pulls the thread taut and it holds fine. Downside is the thread leaves a little dimple where it attaches, which I patch in cleanup anyway.

Quick update - I finally gave the hanging method a shot. Using that elastic thread idea you mentioned completley eliminated the base warp issue I was struggling with. It let me get much closer to the subject and capture angles that my turntable setup just couldn't reach. It's definitly a game changer for these small pieces.