Experiment in Stacking

[Round Tuit]

These are miniature daisies in a bouquet of cut flowers. They are approximately one inch in diameter. I stacked 15 shots using the Dmax algorithm of Zerene and retouched the background and the edges of the petals using the result of the Pmax algorithm as input for the corrections.



At a first glance, the image looks OK but a close examination still shows many smudges and fuzzy areas around the white petals particularly in the top left quadrant. I looked through the entire input stack to find clean edges without success.

Any suggestion on how to control these smudges would be welcomed.

[JBW]

I have found that stacking requires absolute Rock of Gibraltar steadiness. Is there a possibility that you moved the camera when clicking the shutter. if you manually clicked how much of a delay did you program. Was there a fan going to create a breeze? I belong to a couple of macro clubs and the best of the shooters go to some serious extremes to negate camera movement and subject shudder. Hope this is a help

[DanK]

I only have a little laptop monitor with me, but those look like stacking halos rather than motion blur to me. The petals themselves are not blurry, as there would be if there were motion. Even if the stack is perfect, halos like this occur frequently on edges where the distance (front to back) along the edge is large. Basically, the software can't fully compensate for parallax. You often can't paint from an individual image onto the composite because the edge will be in the wrong place; you can't get both sides of the edge from one image.

However, you should have one image in which each edge is sharp, even if you can't successfully touch up from it. if you don't, you may have too large a focusing change between images.

PMax suffers much less from this than DMap, so when the stack is complete, I can often reduce this to tolerable levels by touching up from the PMax composite, as you did. If that fails, you can try to clone in Photoshop.

[Round Tuit]

Brian,
I tend to be paranoid about motion blur. So, when I use a tripod I always used my cable release and, if using a slow shutter, mirror lock up. I also stand very still for the duration of the shot when shooting indoors to prevent the transmission of movement via the floor. I hang a 5 Kg bag of sand from the tripod when shooting outdoors. All this to say that if my subject is still, which was the case, then I can be confident that the camera does not induce motion blur.
Thank you for taking the time to comment.
Dan,
Your analysis turned out to be right on. I used the focusing ring on the lens to step through the stack. My focusing changes between images were not quite even and I believe most were too large. I shot the stack at f/8 and might have been better at f/16. On closer examination, I also found that a couple of the fuzzies were actually part of the background
Thank you for your input and I'll apply the lesson learned to my next stack.

[Manfred M]

Andre - I'm not an expert here, but your focusing method might be causing some of the issue.

All lenses exhibit something called "focus breathing", where the image size changes slightly as you change focus on the lens. In most day-to-day photography, we never notice that, but it places where we are trying for in-shot focus changes, it can make a difference. My own experience relates to cinematography where focus is sometimes pulled in a scene; for instance focusing on an object in the scene and then, without changing camera position or focal length, focusing on something else in the scene.

In closueup work like you are doing, this can impact the image. My understanding is that is why advanced macro shooters use a focus rail to move the entire camera when doing focus stacking. This also seems to be the case where in landscape photography, focus stacking is becoming very much in fashion. I'm thinking of getting a focusing rail (rather than using my long rail) to avoid the issue you are seeing in my landscape work.

[xpatUSA]

All lenses exhibit something called "focus breathing", where the image size changes slightly as you change focus on the lens. In most day-to-day photography, we never notice that, but it places where we are trying for in-shot focus changes, it can make a difference.<>

My understanding is that is why advanced macro shooters use a focus rail to move the entire camera when doing focus stacking. <>

I have played with stacking for the so-called 'super-resolution' just our of interest. I wonder if André's app includes image alignment as part of it's process. I did a focus stack on a cicada some time back and the rail method certainly messed with the perspective a bit, so alignment did help with that one:



Orientation of the insects was deliberate - I was testing an app ...

[DanK]

Manfred,

You are of course right about focus breathing, but Andre's method of focusing is not the problem. A rail will not help with this.

When doing macro, the image size will change as you work through the stack regardless of whether you change focus with the lens barrel or with a focusing rail. Stacking software will compensate for this, except under certain circumstances, such as the one I mentioned in my earlier post. In that situation neither using the lens barrel nor using a focusing rail will help. The bottom line is that if your reason to buy a rail is to avoid the problem I mentioned in my post, I think you will be disappointed.

I actually own a fairly expensive rail (a Kirk), but I use it solely to help me position the camera for the first shot in the stack. I have never used it for adjusting focus within stacks. I have used tethered shooting a few times (fewer than half a dozen), but all the stacked images you can see in my two flower galleries were created by adjusting focus with the lens barrel.

For several years, my advice to people was that there is no particular benefit to using a rail except at very high magnifications, where precision in the adjustment between images becomes more demanding. That turns out not to be correct: a rail is actually slightly inferior because it causes the entrance pupil of the lens to move. This is explained in a video that Odd posted here almost two years ago: https://www.youtube.com/watch?v=mzVD95-9YOU. This causes problems with stacking software.

Where a rail--or something other than manually adjusting the lens barrel--becomes important is at very high magnifications, where one is likely to be too imprecise with the lens barrel.

[Round Tuit]

Andre - I'm not an expert here, but your focusing method might be causing some of the issue.

All lenses exhibit something called "focus breathing", where the image size changes slightly as you change focus on the lens. In most day-to-day photography, we never notice that, but it places where we are trying for in-shot focus changes, it can make a difference. My own experience relates to cinematography where focus is sometimes pulled in a scene; for instance focusing on an object in the scene and then, without changing camera position or focal length, focusing on something else in the scene.

In closueup work like you are doing, this can impact the image. My understanding is that is why advanced macro shooters use a focus rail to move the entire camera when doing focus stacking. This also seems to be the case where in landscape photography, focus stacking is becoming very much in fashion. I'm thinking of getting a focusing rail (rather than using my long rail) to avoid the issue you are seeing in my landscape work.

I am not very good at this yet so definitely not an expert. The program that I used is called ZereneStacker. The recommendation of the author is to use the focusing ring rather than a focusing rail if you shoot "something bigger than a raisin". This tutorial explains why. The first step for the program is to align the shots and correct the perspective distortion cause by the focus variations. I am sure that Dan (DanK) can explain how and why this works better than I can. You may want to investigate further before buying a focusing rail for landscape photography. I really don't see how a rail would work in that application.

[Manfred M]

I am not very good at this yet so definitely not an expert. The program that I used is called ZereneStacker. The recommendation of the author is to use the focusing ring rather than a focusing rail if you shoot "something bigger than a raisin". This tutorial explains why. The first step for the program is to align the shots and correct the perspective distortion cause by the focus variations. I am sure that Dan (DanK) can explain how and why this works better than I can. You may want to investigate further before buying a focusing rail for landscape photography. I really don't see how a rail would work in that application.

Thanks.

There is a lot of information on this out on the internet and there is a lot of garbage made up by people who do not truly understand either the mechanics or physics of photography.

Right now I am at the "filter out the garbage" stage and am not sure if I even want to go there.

I use a rail when I shoot panoramas where there are things in the foreground that impact the image and I have to rotate the camera around the no-parallax point. I am pushing the limits of my rail (for one lens I rotate right at the limit it goes to) and if I spend money to buy an longer one, I was thinking of getting one that is less crude in making the adjustments than with my current one.

[DanK]

The Zerene tutorials are an excellent place to get solid information about stacking.


Sent from my iPad using Tapatalk

[Manfred M]

The Zerene tutorials are an excellent place to get solid information about stacking.


Sent from my iPad using Tapatalk

Which version of Zerene do you use?

[DanK]

Which version of Zerene do you use?

I have the prosumer version. You can see a description of the license types here

I have been highly satisfied with this software. It works very well, has some very handy features, and has detailed and clear tutorials. The author provides good customer support.

[billtils]

I have been highly satisfied with this software. It works very well, has some very handy features, and has detailed and clear tutorials. The author provides good customer support.

+1 to what Dan says. I tried Zerene and Helicon some time ago but didn't find either offered any significant advantage over the focus merge routine in Affinity Photo.

However, I've recently been exploring tethered capture, including using the tethering software to make the incremental focus steps. That in turn led to re-examining the stacking software options and this time up Zerene was the hands down winner.

There were many variables encountered in the above journey and they will be revisited until I am sure that the results are repeatable, at which point I'll come back and post some examples, but for now an unverified anecdotal comment is that I did sometimes see the effect mentioned by Andre in the OP when using AP and manual focus step adjustment, but have not encountered it in the tether + Zerene environment (or perhaps I should add "not so far, at any rate).

[DanK]

I'm going to try to sum up:

1. The problem Andre noted can arise from too great a distance between images or by a large distance between two surfaces at an edge.
2. Andrew found that the former was his main problem. Software can't compensate for that.
3. Re the latter (halos present even when there are enough images): various stacking algorithms suffer less from this than others. In the case of Zerene, DMax will typically show larger halos than PMax. (DMax is superior in other respects for flowers, so when I have a problem, I touch up edges from a PMax composite onto a DMax composite.) I haven't compared Zerene to other software packages in this respect, but Bill found that Zerene was better than the other two he tried.
4. Again re halos present when there are enough images: using a rail will not lessen this problem and can generate other artifacts unless the object photographed is very small.

[xpatUSA]

Is not the root cause 'perspective' distortion, as described here?

Two forms of distortion are described: 'extension' and 'compression'. Distance also gets a mention as a factor.

Since the above seems like "Photography 101", I'm probably missing something ...

[pnodrog]

Considering the DOF required to produced the shot it looks pretty successful to me. I would prefer the secondary flower to be toned down a bit more but that is an aesthetic observation rather than technical.

[Round Tuit]

Considering the DOF required to produced the shot it looks pretty successful to me. I would prefer the secondary flower to be toned down a bit more but that is an aesthetic observation rather than technical.

Thanks Paul. I agree with you that the photo is successful as is but it isn't as good as it could be. Ideally, I should have removed the top flower before taking the shot. I agree that toning it down would improve the composition.