Stacking - Rail or Focus Ring

[Stagecoach]

Andre's thread 'Experiment in Stacking' Experiment in Stacking
had some discussion regarding the merits of focus rail versus focus ring with a couple of interesting articles linked.

Having read one of the articles from Zerene https://zerenesystems.com/cms/stacke...ringversusrail
it mentions the "change in shape" (as opposed to the obvious change in scale) of the subject between the nearest and farthest shot when using a rail which is the first time I had heard of this. The article shows a gif where it can clearly be seen what he refers to.

So being my normal inquisitive self I decided to set up a test to see if I could replicate this and if I would come to the same conclusion.

Firstly, in this thread it is not my intention to prove or disprove the merit of either method, or do I have enough experience from using both methods to have led me to any conclusions.

1. Test intention

Using both focus rail and focus ring (two separate tests) to take a shot of the near edge and then the far edge of a subject and see if these can be superimposed on each other after correcting for the 'scaling' differences and to see how much "change in shape" was apparent with the rail use.

2. Test Method

Camera set on tripod with focus rail. A small aperture was used to get max DOF to get best clarity when the two shots were merged. The subject was a 50mm diameter shot glass. Camera D800 with 105mm VR micro Nikor

For each separate test the start point (nearest) was with the camera/lens at the stop point on the rail with the lens focused on the near rim. For the rail test the rail was wound forward for the second shot to the point where the rear rim was in sharp focus. For the focus ring test the lens barrel was rotated to the point where the rear rim was in sharp focus.

The four shots were opened in PS and titled. (Note, the titling was placed in exactly the same position within each frame with respect to the bottom left corner).

For both the Rail and Ring test shots their 'rear focused one' was merged onto the 'front focused one' at 50% opacity and 'scaled' to align.

3. Results

Using Lens Focus Ring


Using Focus Rail



I can detect 'no change in shape' with the focus rail example as seen so obvious in the Zerene article, why? What have I done or used differently?

Grahame

[george013]

Andre's thread 'Experiment in Stacking' Experiment in Stacking
had some discussion regarding the merits of focus rail versus focus ring with a couple of interesting articles linked.

Having read one of the articles from Zerene https://zerenesystems.com/cms/stacke...ringversusrail
it mentions the "change in shape" (as opposed to the obvious change in scale) of the subject between the nearest and farthest shot when using a rail which is the first time I had heard of this. The article shows a gif where it can clearly be seen what he refers to.

So being my normal inquisitive self I decided to set up a test to see if I could replicate this and if I would come to the same conclusion.

Firstly, in this thread it is not my intention to prove or disprove the merit of either method, or do I have enough experience from using both methods to have led me to any conclusions.

1. Test intention

Using both focus rail and focus ring (two separate tests) to take a shot of the near edge and then the far edge of a subject and see if these can be superimposed on each other after correcting for the 'scaling' differences and to see how much "change in shape" was apparent with the rail use.

2. Test Method

Camera set on tripod with focus rail. A small aperture was used to get max DOF to get best clarity when the two shots were merged. The subject was a 50mm diameter shot glass. Camera D800 with 105mm VR micro Nikor

For each separate test the start point (nearest) was with the camera/lens at the stop point on the rail with the lens focused on the near rim. For the rail test the rail was wound forward for the second shot to the point where the rear rim was in sharp focus. For the focus ring test the lens barrel was rotated to the point where the rear rim was in sharp focus.

The four shots were opened in PS and titled. (Note, the titling was placed in exactly the same position within each frame with respect to the bottom left corner).

For both the Rail and Ring test shots their 'rear focused one' was merged onto the 'front focused one' at 50% opacity and 'scaled' to align.

3. Results

Using Lens Focus Ring


Using Focus Rail


I can detect 'no change in shape' with the focus rail example as seen so obvious in the Zerene article, why? What have I done or used differently?

Grahame

I think there's another item involved to get a sharp stack. Distance and particular the change in distance. In the article it was mentioned there was no problem with a raisin but as soon the size in depth is increasing the problems are increasing.
I think the main difference between ring and rail is the change in distance. With the ring the distance is changed too but a little, with a rail the change in distance is bigger.
The formula I base this on is M=(di-f)/f or M=f/(do-f). In your example the difference in magnification between the front and rear of the cup is bigger with the rail as with the ring. Just a theory I thought of for a while. Didn't check it.

George

[Stagecoach]

I think there's another item involved to get a sharp stack. Distance and particular the change in distance. In the article it was mentioned there was no problem with a raisin but as soon the size in depth is increasing the problems are increasing.
I think the main difference between ring and rail is the change in distance. With the ring the distance is changed too but a little, with a rail the change in distance is bigger.
The formula I base this on is M=(di-f)/f or M=f/(do-f). In your example the difference in magnification between the front and rear of the cup is bigger with the rail as with the ring. Just a theory I thought of for a while. Didn't check it.

George

George, you are right regarding the differences in magnifications with use of the rail and ring methods, and the wording was put there to show that. It can also be seen with the 'lighter' 50% opacity area outer of the merged rail shots.

I'm not investigating 'sharpness' in any stacking method but the phenomena suggested in the Zerene article of a "change in shape" (not to be confused with a change in scaling).

[Stagecoach]

Update................

I'm now suspecting the "difference in shape" (as opposed to scale) is there but as this was only a 50mm distance front to back it is not enough to show by merging the images in this way. I have made a gif which suggests it's there so will attempt to downsize it for posting tomorrow and do a comparison one with the focus ring method also.

[george013]

Update................

I'm now suspecting the "difference in shape" (as opposed to scale) is there but as this was only a 50mm distance front to back it is not enough to show by merging the images in this way. I have made a gif which suggests it's there so will attempt to downsize it for posting tomorrow and do a comparison one with the focus ring method also.

Within 1 image the magnification is depending on the distance. The front and back of your cup have a different magnification.
M=f/(do-f)Let the width of your cup be 5cm or 50mm, focal length = 105.
Mfront = 105/(do-105), Mback = 105/(do+50-105)
Now you move your camera on the rail so the backside of the cup is sharp. You've to move the camera 50mm
Mfront=105(do-50-105), Mback=105/(do-105)
If your focus distance was 50cm or 500mm you get
first image: Mfront =0.2658, Mback=0.2359
second image Mfront=0.3043, Mback=0.2658

A flower with leaves between camera and background is more complicated off course.

Hoping I didn't make any faults

George

[DanK]

Thanks for posting this. As someone who stacks a lot, I know that this took substantial time.

Your images do show the phenomenon Rik wrote about, but the shape and placement of the object you photographed makes it hard to see it.

Start instead with the two captions. The stretching that is entailed in a change of shape is quite evident. The second title is much more displaced relative to the first in the image taken with a rail. Moreover, the difference is much greater on the horizontal axis, which is exactly what I would expect because of the aspect ratio: the start of the title is much farther from the center on the horizontal axis.

Rik's flower image extends almost to the edge of the frame. Moreover, it is asymmetrical on both axes, which makes the change in perspective more obvious.

[JohnRostron]

Surely, for the rail, the magnification remains constant for whatever the lens is focussed on. The subject distance, image distance and focal length are all constant. It is only in the focus ring that the magnification changes.

John

[rpcrowe]

I use my Canon 6D Mark-2 for focus stacking and it it the best setup I have ever used.

I use the 5D2 on a tripod and use a soft tipped stylus to touch the area of the image that I want to focus on. I have the camera set so that the touch will both focus and trip the shutter. The touch is very light and doesn't move the camera at all...

I can select exactly where I want the focus to be.



I shot the above at the San Diego Model Railroad Museum using my 6D3 and a 100-400mm f/4.5-5.6L IS ii lens.

This is very much like shooting with the camera tethered to a computer except that I don't need the computer and I can do it in the field. Of course, a tripod is a necessity.

I have not yet used it this way but, my Sony A6500 will focus using the touch screen which should allow me to use it in basically the same way a I use the 6D2. However, the Sony will only focus using the LCD viewer and I would need to trip the shutter independently...

[DanK]

Surely, for the rail, the magnification remains constant for whatever the lens is focussed on. The subject distance, image distance and focal length are all constant. It is only in the focus ring that the magnification changes.

John

There are three ways to use a rail:
1. use a bellows, fix the lens, and use the rail to move the body
2. Use a bellows, fix the body, and use the rail to move the lens
3. Leave the lens fixed on the body and use a rail to move the combination.

I think people here are discussing #3, and for #3, what you wrote is incorrect. The image distance changes as you move the combination toward or away from the subject, and with it, the magnification changes. If you have a rail and macro lens, you can try this and see that magnification changes as you change image distance with the rail.

[george013]

Surely, for the rail, the magnification remains constant for whatever the lens is focussed on. The subject distance, image distance and focal length are all constant. It is only in the focus ring that the magnification changes.

John

Look at my calculations. It's not only the point where you focus on, also the points that are out of focus, other distances. That's where the stacking is for. Yes, the magnification remains equal with the same distance. In my example that was 0.2658 both for the front and the back of that cup in two shots. But for both shots the magnification of the other side of the cup was 0.2359 and 0.3043.

George

[vaez]

nice

[Stagecoach]

A flower with leaves between camera and background is more complicated off course.

Hoping I didn't make any faults

George

Yes it would certainly be more difficult to measure George but the principles would be the same I think.

[Stagecoach]

Thanks for posting this. As someone who stacks a lot, I know that this took substantial time.

Thank you Dan. Much time was spent with the scaling and alignment of the front and rear frames but once aligned simply flattened and saved. What I really struggled with was accurate cropping of the image separately to produce the gif. I learnt a lot regarding what we can do with 'guide lines'.

Your images do show the phenomenon Rik wrote about, but the shape and placement of the object you photographed makes it hard to see it.

I agree, luckily it shows well in the gif. If I had used a longer distance between near/far it would have shown better.

Start instead with the two captions. The stretching that is entailed in a change of shape is quite evident. The second title is much more displaced relative to the first in the image taken with a rail. Moreover, the difference is much greater on the horizontal axis, which is exactly what I would expect because of the aspect ratio: the start of the title is much farther from the center on the horizontal axis.

You have given me an idea. We can measure both vertical and horizontal distances accurately of the near and far shot and from that see a change in ratio.

[george013]

Yes it would certainly be more difficult to measure George but the principles would be the same I think.

The point of view is changing too. Leaves/subjects are hiding the background, coming closer so that the background is becoming more visible. Or even the inside of your cup.
Back to the formula, you can see that the smaller the change in distance is in relation to the distance, the less influence it will have. In the example the change was 50mm and a distance of 500mm. If you could make it 50mm at a distance of 1000mm the difference would be less.

George

[Stagecoach]

Surely, for the rail, the magnification remains constant for whatever the lens is focussed on. The subject distance, image distance and focal length are all constant. It is only in the focus ring that the magnification changes.

The distance moved was 50mm. (would have been slightly more due to angle)

John

Here's the shots taken with the Focus Rail John that show the increase in magnification of the rear focused image compared to to front focused image.





The magnification/scaling being further explained by George and Dan.

[Stagecoach]

As mentioned in my earlier post my objective was to replicate what the Zerene article had mentioned and demonstrated very well regarding the apparent "change in shape" the stacking software has to deal with when a focus rail is used.

I think most of us are aware of the differences in magnifications between the use of rails and focus ring, but straight forward magnifications/scaling differences between slices should not be a hard thing for stacking software to deal with.

Over the years I had read many discussions regarding the merits of both methods but the vast majority came down to 'ease of use'.

So here's my results of attempting to replicate Zerene's explanation and example, and I have also included a 'comparison'

No 1 - Result achieved using Focus Ring


No 2 - Result achieved using Focus Rail


So as the article mentioned it does appear the glass/subject tilts and we can put this down to a change in shape/perspective

Also, using a rail gives you the possibility of the 'Rail Mechanical Travel' not being in 'Exactly' the same direction as the optical axis between sensor and subject between shots. So some pondering over this is also needed

[george013]

Some more calculations based on the former figures.

Using the ring you move the lens forwards or backwards.
focal length 105, distance 500mm, diameter of the cup 50mm
Using the lens formula the image distance is for the front of the lens 133mm When focusing at the back of the cup the image distance will be 130mm. So your lens moved only 3mm to change focus to the back of the cup. With the rail it would be 50mm.
I think the best way is what Dan wrote: fix the lens and move the body. But then you need a bellow setup.

George

[JohnRostron]

There are three ways to use a rail:
1. use a bellows, fix the lens, and use the rail to move the body
2. Use a bellows, fix the body, and use the rail to move the lens
3. Leave the lens fixed on the body and use a rail to move the combination.

I think people here are discussing #3, and for #3, what you wrote is incorrect. The image distance changes as you move the combination toward or away from the subject, and with it, the magnification changes. If you have a rail and macro lens, you can try this and see that magnification changes as you change image distance with the rail.

You say "The image distance changes as you move the combination toward or away from the subject". But what is the subject? My contention is that the subject is what you are focussed on, the plane in which your subject is in focus. Given this, then the subject distance (from the lens) remains constant; the image distance (from the lens) remains constant; magnification is the ratio of these and thus remains constant. I would concede that for any shot, any part of the image not in focus would have a different magnification.

John

[DanK]

I think the point of this thread could get lost in this disagreement about terminology.

I used subject to mean the thing you are trying to photograph. You would use it to refer to the point on that object that is in focus. One could do either, but the whole point of this discussion is the pros and cons of different methods for focusing on different parts of one thing, whether one calls that larger thing the 'subject' or something else. That's what focus stacking is.

Consider this image:



I was trying to photograph an Asiatic lily, not the anther at the top. To create this image, I had to focus on several different points in this flower to create a stack. To do this, I rotated the lens ring. This resulted in a set of images that had different magnification as well as different focal points. I could also have used a rail--the camera was in fact mounted on a rail--and moved the camera either forward or backward, relative to the flower, to achieve varying focus. Doing so would have created a set of images that differed in magnification.

That's my only point, and it is a key to understanding what this thread (and the previous thread that spawned it) were about. To create a stacked image of any thing--the thing I called the subject--you can either use a rail or the focusing ring to create a set of images with different focal points. Regardless of the method, the images in that set will also have different magnification.

The starting point was whether there is an advantage to using a rail to create this stack. The answer is: with objects larger than a raisin, you will do better using the lens barrel.

[JohnRostron]

I think the point of this thread could get lost in this disagreement about terminology.

I used subject to mean the thing you are trying to photograph. You would use it to refer to the point on that object that is in focus. One could do either, but the whole point of this discussion is the pros and cons of different methods for focusing on different parts of one thing, whether one calls that larger thing the 'subject' or something else. That's what focus stacking is.

Perhaps, rather than use 'subject', I should have used 'subject plane'. That would have made my explanation more explicit. When focus stacking, it is this subject-plane-in-focus which is the important component, which is why I concentrated on that aspect.

This sort of argument reminds me of Humpty Dumpty : "When I use a word,” Humpty Dumpty said, in rather a scornful tone, “it means just what I choose it to mean—neither more nor less.” “The question is,” said Alice, “whether you can make words mean so many different things.” “The question is,” said Humpty Dumpty, “which is to be master—that's all."

John (and Lewis Caroll)