Mike, thanks for sharing your experiences, which are interesting. If I were getting jpegs that showed banded color for given subject, I’d use another format, too. And a subject that resembles a flat, golden mirror intuitively seems like the sort of thing that might produce banded color.
This said, once you have the other issues ironed out, I wonder something: If you retest the jpeg format with your subjects—with quality level set to “large, fine,” color rendering set to “neutral,” and in-camera sharpening set to “none”—is the banding gone?
Nikon’s jpeg engine is quite good, and I wouldn’t be quick to bet against it. I use jpegs when I can get away with them, raw files when I need them. BTW, if you haven’t tried Nikon’s NX-D software for raw conversion, it produces excellent results and is free. (It’s also somewhat buggy and has a clunky interface—but has good batch capability, which comes in handy for deep stacks.)
At first I thought this was due to variations in illumination, and tried a bunch of things related to such, including waiting longer for the strobes to recharge. I even monitored the 120VAC line voltage. I would pre fire the strobes to warm them up having read somewhere that the strobe color temperature can shift with internal temperature, and warming up can help stabilize the strobe internal temperature.
Flash is notoriously variable, no matter how careful you are with it. Rik made an interesting post that demonstrated random shot-to shot color variation with a speed light. Charlie made another interesting post on shot to shot intensity variation. My own observation (which in my foggy memory, accords with Charlie’s observation, but may not), the most variable flash units can vary 1/3 stop between shots; the least variable, 1/10 stop. This is true even with care given to fully recharge the capacitors between shots. I once had a set of flash units that demonstrated almost no variance, but they were old, are now burned out, and I have no idea why they were more consistent. Zerene Stacker handles this sort of variation quite well, but I prefer continuous light, which can be very consistent.
I use manual mode for everything, and am using an objective in front of a reversed Raynox....so no lens selection for AF.
For others who may read this thread, let me repeat for emphasis: With Nikon camera bodies (at least), even if the lens assembly has absolutely no provision for autofocus, you may want to turn autofocus off at the camera body. Otherwise, the camera may sometimes fail to fire if its autofocus system does not detect crisp focus—which can easily happen when focus-stacking.