In camera focus stacking - slightly erratic behavior of E-M1 Mk II

Got myself a used E-M1 Mk II for macro shooting together with the Oly 60mm/2.8 macro lens. The good IS and fast readout speed of the M1 II let me hope for some easier hand held stacking compared to the results I got with my E-M10 Mk II (which is absolutely great for tripod based stacking work IMO).

One thing that differentiates the M1-II from the M10-II is the ability to do internal focus stacking - the M10 "only" can do focus bracketing.

I found the internal stacking on my Olympus Tough TG5 always fun to use - and of course knew that better results could be obtained doing the stacking later with specialized software on a computer. For one I got considerably less artifacts that way.

Compared to that the internal stacking of the E-M1 Mk II now seems to be a bit more erratic to me. I'm on the newest firmware 3.6, and it's sure is nice to have the frame indicating where the camera will crop the resulting stacked picture. But comparing the in-camera stacked picture with one - using the same single exposures - stacked in Zerene I noticed some odd things.

Very often the in-camera stacking seems to refuse to use pictures that are at the end of the stack. E.g. in this example: On the left the Zerene stack, on the right the in-camera one. You can clearly see that some of the leaves further back are OOF in the in-camera stack, but sharp (at least in part) in the Zerene version. That the Zerene version also covers a wider region (see lower edge) is a nice bonus, but of course no fault of the in-camera version which cropped as advertised:

And sometimes (a rarer thing) I find frames within the stacks seemingly omitted, as can be seen here in this stack of a clematis. This time the in-camera stack is on the left, and on the lower left you can see some regions that are sharp in front (circled in red at the bottom), but behind them a region of unsharpness shows that the Zerene version doesn't have. Also lot of the pistils are much sharper in the Zerene versions (one also circled in red in the middle of the pic) which was generated flawlessly without any stacking retouching necessary:

My conclusion: It's nice to have the internal stacking, but you get better results doing the stacking later - not only to avoid artifacts and to get better NR (especially from RAW files), but also to ensure that all available sharpness on all pictures is actually being used.

Apart from that internal stacking has one advantage for hand held work: It sort of indicates if the stack produced is stackable at all - if the camera succeeds (even slightly faulty or not completely as in the examples above) Zerene will sure will too. If the camera is not able to compute the stack and displays an error (e.g. because of too much camera shake or object movement) one should do another one to be on the safe side.

One note: Using the internal stacking the M1-II with 15 exposures at f5.6 with the 60/2.8 the camera takes the first one at the point the camera is focused one and afterwards 4 in front of this point and 10 behind it. To get a correct stack I place the first one after exposures #2 to #5 of the bracketing series before I let Zerene do its work.

Phil

Thanks for the write up, even if I don’t have an om or in camera stacking, it’s useful to know for the future

i did a stack today with dpp, and didn’t like the result much, I had to adjust many edges by hand

I did a quick review of zerene, helicon, and dpp (results posted somewhere in the macro forum) but I found that results vary depending on the subject, each s/w occasionally gave the better result, although helicon always got there faster and 8 times out of ten I preferred its result, but I still use dpp as I don’t do enough stacking to justify the spend

Jim