The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Finally a "proper" stackshot/viewcamera combo

zachary_goulko

New member
The main reason that I added an automated solution for focus stacking to the rear standard of my Actus is because I was getting parallax errors when the entire view camera was on the rail. When I focus stack, it is rare that I combine the stacking with tilts/swings, and if I do it's a small amount in order to reduce the total number of shots required.
By moving the rear standard the lens remains stationary, and there are no parallax errors on overlapping objects. This also happens when focus stacking with an autofocus lens.
I stack primarily watches and jewelry, and I've tried all the methods. This by far gives me the best quality stacks, with the least amount of errors/halos.
Here's a good explanation with a chart:
https://zerenesystems.com/cms/stacker/docs/troubleshooting/ringversusrail
 

zachary_goulko

New member
The main reason that I added an automated solution for focus stacking to the rear standard of my Actus is because I was getting parallax errors when the entire view camera was on the rail. When I focus stack, it is rare that I combine the stacking with tilts/swings, and if I do it's a small amount in order to reduce the total number of shots required.
By moving the rear standard the lens remains stationary, and there are no parallax errors on overlapping objects. This also happens when focus stacking with an autofocus lens.
I stack primarily watches and jewelry, and I've tried all the methods. This by far gives me the best quality stacks, with the least amount of errors/halos.
Here's a good explanation with a chart:
https://zerenesystems.com/cms/stacker/docs/troubleshooting/ringversusrail
 

zachary_goulko

New member
Apologies for the double post, but there seem to be some issues with posting.
It took 5 attempts to submit the post, but now I see 2 of them.
 

Steve Hendrix

Well-known member
I think this is a thread that could be very helpful to a lot of photographers.

There are a number of fantastic photographers involved in this thread (including numerous CI clients on both sides of the debate) who produce amazing photography and employ focus stacking. I trust them and accept that they know what they are doing and the proof is in their results.

I am not mathematically advanced - let’s say I’m challenged - yet I have a somewhat scientific mind, from the standpoint that I like to prove the theory. So when I encounter theory, no matter how accepted it may be, I often still want to see if applied in a real world situation. It comforts my limitations in being able to review a written theory and wrap my head around it completely.

I read the article from Zeneer Stacker about a year or so ago and thought it made a lot of sense. But in particular I thought the little table they put together of different stacking approaches matrixed with different subject matter was the most interesting part of it, in light of the accepted theory. This theory would not be the first in history that was proven to be correct, only to later determine that under different circumstances, the theory did not hold 100%.

I would be very interested in seeing some of the participants put the theory to the test and produce some examples that show results from rear standard focus stacking vs whole camera focus stacking (all with tilt), maybe with some variable subject matter. As Ed Harris said in Apollo13 - "Let's work the problem, people!".

I think everyone would benefit.



Steve Hendrix/CI
 

gerald.d

Well-known member
I hope You can solve the maths. I'm not so good in math.
I made a simple test at close up distance, not even macro.
What I see is that the angle of focus plane is changing, but not badly, I wonder, I could make a good 6cm thick focus stack of eg. 20 shots.
The equipment was Cambo Actus mini, Hassy Distagon cf fle 4/50 mm ,used wide open, and as sensor worked Canon 6d.
I set the maximal tilt, 10 degrees downward and shot two rulers. The distance (from lens front end) to the nearer one was ~35 cm. The second ruler was around 45-46 cm.
The rulers and the rear standard was vertically levelled. Rear shift up 12 mm (max).
I took two shots, first one focussed moving the rear standard (with live view) at 10 cm in the near ruler, the second at 16 cm (with rear standard).
In the resulting pictures I tried find sharp part at the far ruler.
The distance between rulers was 10,4 cm. (I often make mistakes when measuring things, looking at the picture, I think I focussed the first shot at 9,9 cm)
Then I draw picture of focal planes. The grey line is parallel with 16 cm line to show focal plane angle change.
Aha - I think this was the missing post you were referring to previously?

Thank you for doing this - it's a really brilliant practical demonstration of exactly the point I have been making, which I will return to in a bit. But first some context for the discussion so far.

I have to be honest here - reading between the lines, I don't actually believe a great number of people have been doing much (if any) focus stacking with tilt at macro-level. I must stress that I am in full agreement with Egor and others that the ability to rack the rear standard for a stack to prevent parallax issues will be of huge benefit to many where little to no lens tilt/swing is applied.

This solution clearly introduces the possibility of setting front standard tilt/swing and the racking the rear standard, but I don't actually believe that the consequences of being able to do that were fully thought through, either by Cambo or those arguing against the practicalities that I have raised. There is actually no mention of stacking with a tilted focal plane in the first 8 posts of the thread - Egor didn't even bring it up in the initial discussion.

Cambo then commented in post 9 "The idea behind the way we use the Cognisys motor is, that the front standard is used to get the plane of focus where you need it. After that, it can stay where it is. The actual stacking is done by moving the rear standard in small increments."

As stated before, I do not believe the person making this comment understood at that point the error they were making. In my view, and in the context of the discussion, I think they actually believed that the angle of the focal plane would not change as the camera was focused through the stack.

rodenstock then shares his great self-built solution that he's been successfully using for a year or so. Again - this solution provides the ability to tilt the lens and then rack the rear standard for the stack, but crucially, he later in the thread states "The main reason that I added an automated solution for focus stacking to the rear standard of my Actus is because I was getting parallax errors when the entire view camera was on the rail. When I focus stack, it is rare that I combine the stacking with tilts/swings, and if I do it's a small amount in order to reduce the total number of shots required."

In the interchange between myself and Cambo, it is totally evident that Cambo simply did not understand the problem. Yes, so this solution came about from the requirements of a jewelry photographer, but there is absolutely no suggestion that the photographer concerned actually intended to use this solution with tilt applied - only that "he managed to speed up his workflow whilst obtaining perfect results by doing it this way."

It would be very helpful if the photographer concerned would comment on what their workflow is.

Even after multiple challenges, Cambo still insisted that I was "Not correct. Maybe the animation is not clear enough here, but the front standard is only used to set the plane of focus before you start stacking. It doesn't change during stacking."

Well, you've just demonstrated with a brilliantly conceived practical example that in fact, it is highly relevant - even when we are not remotely close to a macro situation. And the great thing (for me), is that I can just do the maths on your practical demonstration, rather than a theoretical example.

What your test shows is the problems people are going to run into if they rely on Cambo's advice, set tilt on the front standard in order to get the plane of focus where they want, and then rack the rear standard.

The maths

If I may interpret your test this way, you are simulating an attempt to shoot a "box" that is 113mm in length (the hypotenuse of the 104mm base by 45mm high (144mm height on the rear ruler less 99mm height on the front) right angled triangle).

This box is 61mm deep (160mm less 99mm) - to keep things simple here, I will assume the box has a parallelogram profile. The large flat surface of the box presents an angle of around 23 degrees to the horizontal (the angle to the horizontal where you mark 99mm in your diagram), and the complexities of the Scheimpflug calculations means that a 10 degree tilt will get the whole of the surface of the box in focus when the near edge of the box is 99mm from the camera.

Now - imagine this "box" is not solid. Imagine it is simply created from thin bits of wood that form the 12 edges of the box.

You now rack the rear standard to get the nearest edge of the rear surface of the box into focus at 160mm from the camera. The tilt angle remains the same of course.

But what about the far edge of the rear surface of the box? Well, on the last image of your stack, it will be nowhere near in focus. Because the focal plane now falls fully 17mm beyond the position of the far edge of the rear surface. (17mm being the difference between (160-99) and (222-144)).

The focal plane angle with respect to the horizontal has changed from 23 degrees to 31 degrees.

Conclusion

I'd say that with the help of the test you performed, we can pretty much categorically conclude that in many situations - even those not remotely approaching macro level maginifcation, those who rely on fixing movements on the front standard and then racking the rear standard to perform a focus stack - as recommended by Cambo - are going to run into some pretty significant challenges.

They will end up with the focal plane angle changing substantially, and they will have to over-shoot the stack in order to ensure they get the entire volume in focus that they require. No prizes given for who works it out, but if anyone wants to have some fun - tell me what the stack parameters would have to be in order to get the cube I discussed to explain the maths behind your example fully in focus.

Kind regards,


Gerald.
 

MGrayson

Subscriber and Workshop Member
I don't know if this image is helpful, but it is true. Forgetting the film plane for a second, and assuming an ideal thin lens without field curvature, a lens takes any plane in front of it to another plane behind it. These two planes are in focus with respect to each other. The two planes are either parallel, or intersect on a line that also lies in the lens plane. If you try to imagine families of these planes, you see what happens in stacking experiments. There are purely geometrical ways to determine paired planes of focus (Scheimpflug and hinges) or you can use the fact that, along the lens axis, pairs of points in focus satisfy 1/a + 1/b = 1/f, where a and b are the distances to the lens and f is its focal length.

Here's an example showing what happens as we rack the film plane forward while keeping a tilted lens fixed. Note the second hinge point, although I did these by the thin lens equation.



Best,

Matt
 

MGrayson

Subscriber and Workshop Member
More to the point, if you have an object at some angle to the lens, its in-focus points on the other side of the lens (labeled, confusingly, "Image") will not be arranged in any easy way to capture:



The film plane's relation to the lens will not ideally be fixed wrt the lens. You would want the film plane to lie along either set of red lines, depending on which way you wanted to stack.

--Matt
 

dchew

Well-known member
This post is not about macro, so in that regard it is off-topic. I do think there are a few things to point out that are relevant and might help, as per Steve Hendrix’s request. At least for those of us who don’t have access to Gerald’s CapCam! Also, as I mentioned above, I look at this stuff a bit differently than most so this post may help dispel how crazy I am (I’m still a little crazy, but hopefully not that crazy).

I noticed over several years that most people think of lens tilt as tilting the plane of focus. In other words, there is a causation we have in our brains between the degree of tilt and the degree to which the plane of focus is “laid down” flat. In my opinion, this perceived connection is one of the major sources of confusion associated with lens tilt and the Scheimpflug principle.

There is a correlation between lens tilt and plane of focus angle, but not when focused at infinity. In fact, even when focused in from infinity the rear DoF plane angle does not change at all as you dial in more or less tilt, so I find it more useful to think of it this way: As you dial in more tilt, you are simply pulling the DoF wedge up through the ground closer to the camera because the hinge point is moving up.

First, a model description. To keep this simple, everything described below is with the camera level to the horizon both in pitch and roll:
Imagine a threaded rod attached to the bottom of the camera that extends down through the ground infinitely far. The threaded rod is fixed and in line with the film plane, but the threads spin as you dial lens tilt in and out.

As soon as you dial in any small amount of tilt, say 0.01 degrees, something magical happens: the normal parallel planes in the standard DoF (near DoF, plane of focus, far DoF) flip from being parallel planes out in front of the camera to being a wedge, or cone of DoF, with a point of origin (where the lines converge) mounted on that threaded rod. I will refer to that as the “head” of our cone of DoF. As we dial in more tilt, the threads rotate and pull that point of origin (head) up from infinitely far underground toward the base of our camera.

If the lens is focused at infinity, the plane of focus is 90 degrees from the film plane and, since our camera is oriented with the horizon, the plane of focus is also parallel with the ground that extends out to infinity in front of the camera. It remains 90 degrees from the film plane regardless of how much tilt we dial in. Focused at infinity, the only thing more tilt does is pull that hinge line up through the ground closer to the base of the camera; the plane of focus remains flat, 90 degrees from the film plane.



Now we add the focusing variable: Adding to the threaded rod model, let’s add a rope that extends from the lens down, and is tied some point out from the head, at the “waist” of the plane of focus. Close to the point of origin where the near DoF, plane of focus and far DoF converge, but just a wee bit out from that point of origin. Like this:



I know that picture isn’t an accurate description of the angles and the math. But for this explanation it should be good enough…

Now a little more magic: Whenever the lens is focused at infinity, the rope is just long enough to keep the plane of focus 90 degrees from the film plane. As we pull in focus from infinity, the rope will get shorter and pull the plane of focus up from horizontal.

To prove all this, I need to give a shameless plug to Torger’s Lumariver DoF app:
www.lumariver.com. Below are several screenshots from Lumariver.

First, I show what happens as more tilt is dialed in with the lens focused at infinity. Note how the hinge line gets pulled up through the ground closer to the base of the camera, from 11 ft @ 1 degree to 5.6 ft at 2 degrees:



Second, showing the “standard” landscape application, where the far DoF is set along the horizon, 90 degrees from the film plane. We do this by dialing in focus from infinity:



Third, now that the rear DoF plane is horizontal we dial in more tilt.


A few things to highlight:
  1. More tilt pulls the point of origin up the threaded rod
  2. More tilt scrubs (diminishes) DoF
  3. The far DoF remains horizontal if we don’t change the focus distance dialed on the lens.
  4. The plane of focus does, indeed, “lay down.” But I like to think of this as being due to the DoF being scrubbed or squashed. You, of course, can look at it however you want!

Fourth, a sequence of maintaining the same tilt (1.5 degrees), but dialing in focus from infinity. Hinge point doesn’t move, but our rope gets reeled in, which pulls up the cone of DoF.



Fifth, the difference between dialing in more tilt to get something close in focus, vs reeling in the focus point. In many cases, focusing closer helps because the DoF remains the same width. More tilt brings the plane of focus closer to the ground, but at the expense of height. The ground is all sharp, but the top of the tree or mountain might get blurry as you dial in more tilt. The upper right corner of each image depicts the DoF wedge projected at infinity on the film plane. In the third image, it is the highest, 31mm above the center line of the film/sensor (assuming no rise/fall, but that's another story...).



In summary, I think it is more useful to think of focus as the tool that tilts the plane of focus and cone of DoF. The degree of lens tilt defines how far the head of that cone sits below the camera. It is true that when focused in closer than infinity, changing the degree of tilt does “tip” the plane of focus. But it does not change the angle of the rear DoF line at all.

Again, thanks to Torger!

Dave

GO TRIBE!
 
Last edited:

MGrayson

Subscriber and Workshop Member
Dave,

Who are we kidding? People want gizmos that work. Projective geometry is only useful if you want to understand WHY they work.

:chug:,

Matt
 

Ori

New member
I’m the guy to blame (-:

WoW what an amazing thread to read!
Hi guys, I’m Ori Livney and I’m the guy behind this invention
I’m a jewelry photographer from Israel
I came up with this idea about 2 years ago when I realized that moving my whole cambo ultima was possible with cognisys stackshot it was far from being the best solution, moving the camera created challenges that helicon focus had a hard time dealing with, after realizing this I set the goal to solve this in a way that will a speed up my workflow and b creat better results, after building a working prototype and working with it for several months, I contacted Cambo and they were great and together we made this a viable working solution to a: stacking with the ultima and b: remote focusing
Although I use the ultima’s movements in various situations, I don’t use the tilt and stacking together, for me the 2 approaches are means to an end and that can vary depending on my needs,
I use a Nikon D850 and Schneider 80mm macro digitar and 120mm macro digitar for 99% of my work
I use front standard movements to shoot usually flat necklaces on a surface or use swing for shooting odd angles, but in general I just focus stack to get my needed results. The sharpness and quality of this setup is above and beyond anything else I have tested.
The stacking bracket that makes this possible is an awesome ultra important part of my workflow, it saves me so much time and enables me to get the highest level of performance both from an accuracy and efficientcy stand point, being an industrial engineer these are a part of my “DNA”
So to summarize thing up, the discussion on the tilt and the effects of stacking and tilting is fascinating and I learned a lot!
But the truth is that there’s no real reason to do both at the same time, getting the schinflug principal right and getting a good accurate results can take more time than stacking the shot, especially with this power combo.
Stacking 5 vs 10 images plus having to deal with the correct tilt angle doesn’t really make sense to me,
Bottom line, for me it’s a life changer when it comes to my studio workflow and I’m sure it will be for you if you use the cambo ultima for table top photography
I will be more than happy to answer any questions you might have!
Nice to meet you guys!
Ori
 

gerald.d

Well-known member
No - you're the guy to thank!

Hi Ori -

Thanks for taking the time to join the forum and share the background to this camera.

Firstly, as per the post title edit, I actually think people should be very grateful that you came up with this solution and worked closely with Cambo to see it develop into a final product - a product that I am sure will be of great benefit to many photographers. Kudos to Cambo for working closely with their customers to create a new solution.

You raise a few interesting points in your post that I think are worth addressing - particularly in context of the discussion that has occurred in this thread.

Although I use the ultima’s movements in various situations, I don’t use the tilt and stacking together, for me the 2 approaches are means to an end and that can vary depending on my needs,
Thanks for clearing that up.

It is as I anticipated, and serves to underline the point I was making earlier in the thread that - contrary to claims made by others - this is NOT a viable solution for those who have a requirement to tilt/swing and stack at the same time.

So to summarize thing up, the discussion on the tilt and the effects of stacking and tilting is fascinating and I learned a lot!
But the truth is that there’s no real reason to do both at the same time, getting the schinflug principal right and getting a good accurate results can take more time than stacking the shot, especially with this power combo.

Stacking 5 vs 10 images plus having to deal with the correct tilt angle doesn’t really make sense to me,
Actually, there are very good reasons for wanting to do both at the same time, but what prevents most people from actually being able to do it is quite simply that their camera system cannot fulfill the requirement.

Where it makes sense to want to do this is when you are shooting something that is not flat on with the non-swung/tilted focal plane, presents a large surface area to get in focus, but is not deep - i.e. when viewed straight on, x and y axes are significantly larger than z.

Something like a watch head, a bangle/bracelet, or a ring.

In this example, you could be looking at the difference between shooting half a dozen images to get the shot in focus if you can align the focal plane with the surface of the object, or twenty to thirty shots if you cannot.

Let's put some numbers to it. Assume you have a bangle/bracelet 10cm in diameter and 1cm thick, and you want to shoot it at an angle of 40 degrees to the lens axis. Assume a depth of field of 2mm.

If you can align the focal plane with the plane of the bangle, you need to shoot 5 images for your stack. If you cannot, and choose instead to shoot with no lens movements, then because the near point of the bangle is 5.8cm closer to the camera than the far point, you are going to have to shoot 29 images to get it all in focus.

Now - all other things being equal - what would you rather? Do you want to have to work with 29 images, or 5?

Of course the problem is - with the kit that I guess almost everyone uses - all other things are not equal. It simply isn't feasible to nail that swung focal plane for each shot in the stack.

Basically, with the kit you have chosen to use, you don't really have a choice - you're either going to have to shoot 29 images, or simply choose not create the image.

Personally, I'd rather have a solution that allowed me to get the shot in only 5 images.

The other reason why one might want to be able to stack perpendicular to the face of an object is if for aesthetic reasons, one wanted the focus to drop off perpendicular to that face. Example -



Here the desire is to have the watch "head" in focus, but then have the focus fall off as we move away from the head. Naturally, this image has been shot with a swung and tilted focal plane, and then (IIRC) about 3 shots for the stack to get what I wanted in focus.

Had this been shot with a flat focal plane, then not only would it have required a lot more shots, but more importantly from an artistic perspective, a lot of the bracelet on the right hand side of the frame would have been in focus, whereas all of the bracelet on the left hand side would be out of focus. This would have given a very odd look to the image.

It is close to impossible to create this "look" (assuming you need to stack to get your desired magnification) with any other system other than the camera I use. Whether or not one sees any benefit in being able to deliver an image like this of course is a whole other matter entirely. I don't doubt that if a client asked a photographer to come up with this shot, and the photographer was unable to do so, the photographer could come up with a very convincing argument as to why the client shouldn't really be asking for such a shot.

But for the sake of this discussion, let's stick to the functional capabilities and practicalities of the solutions under discussion.

Bottom line, for me it’s a life changer when it comes to my studio workflow and I’m sure it will be for you if you use the cambo ultima for table top photography
I will be more than happy to answer any questions you might have!
Nice to meet you guys!
Ori
In wrapping up - to reiterate, this is a great solution, and I can envisage a lot of photographers finding it enormously beneficial in their work.

It's just a shame that the announcement of it got side-tracked due to spurious claims being made with regards to its capabilities, but at least hopefully as a result of that, there has at least been some educational benefit.

Kind regards,


Gerald.
 

Gerd

Active member
Hello everybody,

my name is Gerd and I have rather come across this discussion by chance. My English is not so good and I hope I have in the previous discussion nothing to read over what was already said.

I also use the Actus XL in combination with tilted focal plane and stacking, as Gerald described it. It is also important to note that the focal wedge opens the farther you fade away.
When focussing/stacking with the rear standard the sharpness level will swing, but my motive can still be in the wedge and thus in the focus. Of course not in all cases, that depends on the composition and the motive.

Attached a test picture with the Cambo Actus XL, FUJI GFX, Mamiya 140mm Macro F4.5 M/L-A.
Sorry for the bad photo model. The watch is a cheap replica of the flea market, scratched and could not be cleansed properly. The purpose of the image was to test only the Mamiya lens at the time










Greeting Gerd
 

Gerd

Active member
Hello everybody,

my name is Gerd and I have rather come across this discussion by chance. My English is not so good and I hope I have in the previous discussion nothing to read over what was already said.

I also use the Actus XL in combination with tilted focal plane and stacking, as Gerald described it. It is also important to note that the focal wedge opens the farther you fade away.
When focussing/stacking with the rear standard the sharpness level will swing, but my motive can still be in the wedge and thus in the focus. Of course not in all cases, that depends on the composition and the motive.

Attached times a test picture with the Cambo Actus XL FUJI GFX, Mamiya 140mm Macro F4.5 M/L-A.
Sorry for the bad photo model. The watch is a cheap replica of the flea market, scratched and could not be cleansed properly. The purpose of the image was to test only the Mamiya lens at the time







Greeting Gerd
 

Kuky

Member
To revive this thread... I am on the fence of buying a Cambo Ultima just for this feature. The Stacksot I already have it (10 cm rail version).
Went to the Cambo dealer to ask them about tips and tricks and they asked me what's a Stackshot. OMG

Bottom line, is any document outlining how you mount the Stackshot after you buy the adapter kit? Cambo didn't bother to put anything on their site.
Are there any limitations/incompabilites I should be aware of?
How do you dissasemble the focus wheel? Didn't seem to find any imbus slot for this purpose....

Regards,
Cristian
 

Boinger

Active member
To revive this thread... I am on the fence of buying a Cambo Ultima just for this feature. The Stacksot I already have it (10 cm rail version).
Went to the Cambo dealer to ask them about tips and tricks and they asked me what's a Stackshot. OMG

Bottom line, is any document outlining how you mount the Stackshot after you buy the adapter kit? Cambo didn't bother to put anything on their site.
Are there any limitations/incompabilites I should be aware of?
How do you dissasemble the focus wheel? Didn't seem to find any imbus slot for this purpose....

Regards,
Cristian
https://blog.cambo.com/2017/10/04/cambo-focus-stacking-solution/
 

Cambo

Member
The Cambo Stack Shot mount comes with instructions how to assemble it. And this small allen screw you're looking for is hidden behind the focus knob's rubber band.
 

Kuky

Member
The Cambo Stack Shot mount comes with instructions how to assemble it. And this small allen screw you're looking for is hidden behind the focus knob's rubber band.
Thanks,

Is there any electronic downloadable version of the instructions?
The Stackshot is the 10cm one which I have to disassemble or do I have to order separate parts from Cognisys?

Regards,
Cristian
 

Cambo

Member
Thanks,

Is there any electronic downloadable version of the instructions?
The Stackshot is the 10cm one which I have to disassemble or do I have to order separate parts from Cognisys?

Regards,
Cristian
Cambo only supplies the hardware to connect the Cognisys gear to an Ultima or Actus-XL. If you've got any questions regarding the Cognisys gear, just contact them. I'm sure they'll be happy to help. And it actually doesn't come as a surprise that your dealer is a bit puzzled when it comes to Cognisys related items, as they sell most of their stuff online.
 

Kuky

Member
Thanks, for some reason I thought you were representing Cambo.
Ofc I contacted Cambo on both emails info and sales but they did not respond. Maybe next week, let's hope.
 
Top