[Top] [All Lists]

Re: [OM] WTFocus Stacking`

Subject: Re: [OM] WTFocus Stacking`
From: Martin Walters <mwalters1440@xxxxxxxxx>
Date: Sat, 16 Jul 2022 10:44:03 -0400
I'm late to the focus stacking (FS) discussion. I finally decided to try FS with my E-M5-2 and the 12-100. Works fine, at least on a tripod.

First: my dumb question.

As FS moves both towards and away from the camera, where should I set the initial focus point? I chose a point somewhere near the middle of the subject (an echinacea flower, which has at least 2 cm depth). The eight frames did cover the full flower. So, while this worked fine, would it be better to start at the nearest point in focus?

Here's a DPReview post offering an explanation of how FS works, particularly as it relates to lens aperture/DOF and the focus differential setting. Quite interesting. In my case, I did not see any lack of focus anywhere in the stacked version, with using F/4. I did not try F/8 or something like that, though a single shot at F/8 was not sufficient to have the entire flower in focus.


"Ran an experiment last night to find out what is really going on with the focus stacking in my EM-1 ii.  I was under the mistaken belief that the focus differential setting was unaffected by the aperture setting. WRONG. I was under the impression that one would not get soft focus in the composite image when using a high focus differential setting as long as one also used a large f stop. WRONG.

So, what I was actually looking for was the minimum aperture I should use at each focus differential setting within the Focus Stacking feature.  IOW, I expected that maybe at a focus differential of 1, you could use f/2.8, but at a focus differential of 5 or 7 or 9...somewhere f/2.8 would fail to cover the gaps between the 8 shots the focus stacking feature takes. I wanted to know at what point f/2.8 fails to cover the gaps. I wanted to know the same for f/4 and f/5.6 and f/8.  Turns out I was asking the wrong question.

Here's what I found. At f/2.8, the highest focus differential one can use and not have soft spots between the frames and in the final jpeg is 3.  A f/4, the highest focus differential one can use is 3. At f/5.6, the highest focus differential is 3.  At f/8, 3. At f/11, 3.  Noticing a trend here?  But wait, higher f stops mean greater depth of field, right?  Why aren't the higher f stops covering the gaps?  Because as you increase your f stop number, the camera increases the space between the shots!!!!!

So, the difference between using f/2.8 and f/11 at a focus differential of 3 is the total depth of field covered.  If you go to a focus differential of 4, you will start to see some minor soft spots in the final product. Go to a focus differential of 5 and the soft spots interspersed become more apparent. The higher you go on the focus differential scale, the larger and more apparent the soft spots get...regardless of aperture. What the aperture does is increase or decrease the total depth of field.

So, it finally dawned on me that what Olympus seems to be doing here is relating the coverage as much to the number of shots stacked, as with aperture. It may well be that stacking 8 shots works well with a focus differential of 3 or less. If you want to use a focus differential of 4, you'd better stack at least 9 frames . If you want to use a focus differential of 5, better stack 10 shots. If you want to use a focus differential of 2, you may only need to stack 7 shots.  As I don't have a good focus stacking software, I was not able to test this theory. The EM-1 does do focus bracketing up to 999 frames.  I'm guessing that is overkill even when you set the focus differential to 10.

Bottom line...I did the testing and now I know. For in-camera focus stacking, the magic focus differential is 3.  Use the aperture to adjust the total depth of field. If you can't get enough depth of field even with a high aperture...then you need to switch to focus bracketing and take a lot more frames, which allows you to use a higher focus differential and still not get soft spots (I call them spots, but they are really more like strips).

One other note, the place to aim your focus when using the focus stacking feature is at 1/5th of the total distance from the close edge you want in focus.  So, you want your total depth of field to be 5 inches, focus 1 inch deep from the close edge of that 5 inches you've targeted (4 inches from the back edge)."


On 2022-07-01 8:14 a.m., Piers Hemy wrote:
There is a just released useful video presentation of both stacking and 
bracketing, with explanation of both modes, and clarification of the different 
presentation in the menus of OM-1 and OM-D


He goes on to cover S and F cameras too, as if anyone would want to use them, 
as well as merging stacks in PS with links to cover Helicon Focus.

Although it's on Gary Friedman's site, the video is Tony Phillips' work.


-----Original Message-----
From: olympus<olympus-bounces+piers.hemy=gmail.com@xxxxxxxxxxxxxxxxx>  On 
Behalf Of Moose
Sent: 27 June 2022 21:48
Subject: Re: [OM] WTFocus Stacking`

On 6/27/2022 11:07 AM, DZDub wrote:
On Mon, Jun 27, 2022 at 12:09 AM Moose<olymoose@xxxxxxxxx>  wrote:

On my E-M II:

Focus stacking          = Off (i.e. Focus Bracketing = On, as set on the
prior screen.)
Set number of shots  = 31
Set focus differential  =  1
Charge Time              = 0
Focus stacking          = On (as an option within Focus Bracketing)
Set number of shots  = 11
Set focus differential  =  1
Charge Time              = 0

This makes sense to me now.  Thanks for sticking with me.
Excellent! My pleasure, literally.

   I've even gotten
it to work, and it's extremely fast compared to focus stacking, which makes
me think the 15 seconds or so that it takes to do an in-camera focus stack
is mainly spent on processing rather than exposing.
Yup. If you watch the EVF/LCD during the process, you can see when shooting 
stops and processing begins. On my E-M1 II,
shots are quick,  writing out ORFs and JPEGs and then processing is almost all 
of the time.

You mention backgrounds more than once. That's not what I'm looking for
with stack/bracketing. I'm looking for deeper
DoF than one shot can provide. Yes, the background matters, but that's a
separate issue to me, dealt with differently.
Two sides to a coin to me, although I suppose I think about backgrounds
more because I'm generally unwilling to spend a lot of time on backgrounds
in post.  I generally bracket (the old fashioned way) from f5.6 to f11 to
see where I get the best compromise of DoF and good background.
Chuck and the other DoF mavens claim(ed) that f5.6 is optimal for a 4/3" 
sensor. My experience and testing disagree, but
f11 does get into visible loss of detail/sharpness.

I suspect that part of the disagreement 'tween the tables and practice may be 
that demosaicing the Bayer array data
loses detail. Ctein estimates that loss at 50%. I've certainly seen it. Shoot 
on a tripod, regular and HR. Downsample
the HR file to the size of the straight shot. The result has much better, 
clearer fine detail.

In my Bracketing practice, part of the process of leafing through the stack to 
determine which frames to merge is
looking at the background.

I guess I'm saying they are never really separated for me.

One of the jobs of the stacking soft/firmware is to align the frames.
The second is to adjust for focus breathing making
frames slightly different sizes. All before actually merging focus.

Focus breathing may be one of the criteria for choosing which lenses are
compatible with in-camera FS
Notice that the Stacked JPEG is larger than the individual shots. There's a 
frame when shooting FS to show the tighter
AoV. Dealing with focus breathing?

I assumed it was something more proprietorial than focus breathing per se.
I did say "may be" and "one of the criteria". 😁

Not all mZ lenses seem to work, which suggests design is a key.
It's all clear as mud to me. None of the non-macro primes are included, only 
zooms, the macros and 8mm fishy.

Do you happen to know if any Panny m43 lenses will focus bracket on an Oly 
Random grab - Panny 25/1.7 works fine. I suspect they all do. Bracketing is 
much less fussy than Stacking.

Testing, Testing Moose

Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz