Focus Stack basics

For stacking focus, I've read that best results are obtained by moving the camera (e.g. rail system) as opposed to just refocusing from the same camera position.
In my tests, I wasn't able to discern a difference. The image size changed somewhat with each capture in both instances.
I deliberately chose a difficult subject and shot at f/4-- I used my pano set-up as a crude rail system, and auto stacked/aligned the shots (7 in each case) in CS5. If anything, the images seemed to change more in size with the rail shots. I can see where shooting minuscule subjects would be easier with a micro-adjusting rail, but for shooting something like florals, it doesn't seem it would be necessary with the new auto-stacking script in PS or similar software (I had used Helicon in the past but found it unsatisfactory). Sure would make my life easier (and less expensive) to just lens focus and forgo any rail system.

The attached pics are from the lens-ring focused test. The first pic shows the initial single exposure, the second shows the completed blend. (I know it's not perfect, but it was a test...)

Any thoughts/suggestions?

HOSTED PHOTO
please log in to view hosted photos in full size.

HOSTED PHOTO
please log in to view hosted photos in full size.

You can also put the subject on the rail and move it closer to the camera for each shot. In the tests I have seen the results seemed about equal, either way. The blending is in the software, and it seems to not care how the stack was made.

Agree there is no real difference between the methods of focusing ie fixed camera move focus ring or fixed focus move camera but I suspect with some lenses the lens focus may not be sensitive enough to allow the small focus increments you sometimes need.
The stack worked well !

Brian v.

Thanks guys!