Schneider 72mm Apo Digitar Focus Shift when stitching??

[Viewmaster]

Hello,

Longtime reader, first time poster. I have a newbie question with regards to LF/MF digital lenses on tech cameras or in my case, a Cambo Actus/GFX Combo.

I have a 72mm schneider apo digitar that I'm not sure is a good copy or not. I was doing a shift test earlier today outside trying to gauge the sharpness of the lens and I noticed after I got the files into the computer that the lens appears to change focus as I shifted to the right (image pan left). I didn't change focus on the camera, all I did was pan left and right taking pics. It appears that as I shifted right, the plane of focus left the distance object I originally focused on (a fence) and jumped forward maybe 30 ft? The fence in the shifted image is now out of focus and a closer object that was slightly soft in the center image is now in focus in the shifted image. No changes to Fstop were made.

Is this normal? I would think that Large format film and digital lenses wouldn't exhibit this behavior since there aren't any moving lens groups in the lens (and they are usually cemented). I used to shoot 4x5 a little, but I never stitched images on film, so it's not something I might have come across.

If this lens were to have a decentered element somehow, would it show this kind of behavior?

Trying to figure out if I have a bum lens. Anybody have any idea what might be happening?

 

[PhiloFarmer]

Well....if you're "shfting"...that's one thing. If you're "panning" that's another thing.

The "panning" (rotating around a tripod-center-point) focus will change as subject-to-lens distance has changed.

If you're "panning" using a nodal-point-head...where the center-point of rotate-pan remains centered at the nodal-point of the lens...there should be no focus change.

Now...if you're "shifting" (moving either front-standard or rear-standard) then it depends on which standard you use...front-standard usually changes only perspective...while rear-standard has the potential to change geometry. Some "movements" are not precisely "equal"...and thence there can be differences camera-to-camera.

Now...if you're "shifting" in order to "stitch"...I'd say using the standards is a poor substitute for using a "sliding-back." The sliding-back allows you to maintain geometry and perspective (as the lens remains in the same relationship to the subject), and the size of the Image Circle (IC) is then the limiting factor for effectiveness of the "sliding-back-shift."

So...it depends on your "flavor" of "shifting"...so to speak.

I'm intrigued if others try to use standard-shift for stitching? I can't imagine why....but enlighten me...

If I'm "stitching" I want the lens to remain fixed...and my sensor-film-plane to expand...to the right and to the left for overlap.

Let me know....

Cheers,

Hank

 

[Audii-Dudii]

Anybody have any idea what might be happening?

I don't know about the latest production units or your particular model, but I know that my early production Actus is notorious for drifting slightly off its zero positions and applying small amounts of tilt and/or swing movements without my noticing it until I'm home and looking at my photos on my 27" monitor, where I frequently find the opposite sides of the composition (i.e., left and right, top and bottom) slightly soft as a result of the face of the sensor and plane of focus being slightly skewed in one direction or another relative to each other.

Is there any chance that your Actus GFX is doing something similar? Or that, for some reason, the front and rear standards aren't perfectly parallel to each other?

To do a quick-and-dirty check, remove the bellows and move the two standards toward each other until they almost touch, then look closely to see whether the width of gap between them varies at any point.

Then slide the rear standard left and right and check the width of the gaps again.

If so, ignore the degree markings that are silk-screened on the body and adjust the front standard until the gap between the front and rear standards is equal all around.

That said, off the top of my head, I suspect the depth of focus of a 72mm lens should be sufficient to cover for a small amount of misalignment between the standards, but I'm not certain. Personally, I use wide and ultra-wide angle SLR lenses with my Actus, which exacerbates any alignment issues due to these lenses having a much shallower depth of focus, so perhaps my experience is unique to me?

That's my two-cents worth, anyway ... good luck!

 

[Gerd]

.....Trying to figure out if I have a bum lens. Anybody have any idea what might be happening?

When shifting and stitching please never use the front standard.

Moving the Front Standard can change the item's size. This can be visible depending on the subject's nearness and depth of field in the picture. But the bigger problem is, you get parallax errors.

Therefore, always use the Back Standard to shift and stitch. The same goes for focus stacking - if you want to do that later.

I do not know what Cambo Actus you have. The GFX fits on the Actus Mini and the Actus XL. I have the Actus XL. Since it can happen that the material can easily twist under pressure and great weight. This looks like the lens is slightly decentered.

Greeting Gerd

 

[Viewmaster]

Hi All,

Thanks for your input it has been helpful. I ended up returning that lens and was able to find another, used copy and am currently testing lens #2. I do notice the second copy is better, but there still appears to be a slight change in what areas are in focus as I move around the complete image circle as I stitch. I'm not sure if this is a de-centering issue or perhaps just the way the lens is, but I will try and upload a series of pics so that maybe everyone can see what I'm seeing and let me know if this is normal behavior for this lens.

Also, I accidentally started two threads about this at the same time. one didn't show up until a few days later I think, so the admin may want to merge the two together for clarity

 

[Viewmaster]

Sorry, What is the image maximum size (dimensions) I can upload? I've been trying to upload some images, but they keep erroring out saying they are too large? Shrinking the file size for jpg compression doesn't seem to work, so I think I'm exceeding maximum dimensions?

 

[darr]

I have employed rear horizontal shift (shifting magazine/back) with my SK 72, ALPA Max and CFV50c & P45 with no problems.
Technique may have something to do with it, but I did have an Arca Swiss ML2 that had a bit of slop in it that drove me nuts at times.

Kind regards,
Darr

 

[Viewmaster]

Hi all,

I've been able to attach a 9 shot pan of the 72mm hitting all the corners as far as my camera will also (cambo actus/GFX). These SOC only changed in size so that I can get them uploaded here. I wish they were a bit bigger so as to better see what might be happening.

The focus point was on the closer red fence (with the black car in front) I centered that shot and then panned to all other other areas. what I notice as I pan around to stitch is that certain points closer to the camera start to move in and out of focus a bit. The Swingset is a good object to hone in on, in some shots it's somewhat out of focus, then as I pan towards the edge of the image circle, the swingset moves into sharper focus.

I thought I was pretty careful setting up the shot to make sure the buildings, fence, etc.. were parallel to the camera, as well as the camera standards being parallel, but I could be wrong.

I know this is a wide open shot, which is not at the sweet spot, but I thought it would help to see if something is happening with the focus point, or if there's a lens decentering issue.

Any thoughts appreciated.

 

[darr]

"I know this is a wide open shot, which is not at the sweet spot, but I thought it would help to see if something is happening with the focus point, or if there's a lens decentering issue."

This may be troublesome. When you say wide open, are you saying you shot at f/5.6? If so, your DoF is limiting; why not shoot at f/11? The SK 72mm is a normal lens, not a wide lens and thus will not have a large DoF wide open. More info is needed:

1. What is your exposure settings?
2. Describe your panning equipment.
3. Describe your panning technique.
4. Are you refocusing before each exposure?
5. Anything else that may have changed between shots?

We can offer help, but more info is needed.

 

[Shashin]

These lenses are not designed to be used wide open--maximum aperture is for focusing. You should stop down to f/11, as Darr suggests.

 

[Viewmaster]

Hi all,
Here is some more info:

Equipment is a Cambo actus mini setup with Fuji GFX.
I don’t recall the exposure off the top of my head but it was set once and then used for the whole set for consistency.

I centered the camera back first in the 0,0 position for lack of a better term,used that to focus on the fence next to the black car. Focus was then locked down and I panned to the top right of the camera’s max shift, panned across left, then left side center, across to the right, then panned down to the lowest position bottom right and back across left. Basically the order that the pics are in.

I did not refocus after the first focus set. That is what I was testing. To see us there is drift or decentering as you move around the inside of the image circle.

I did a series of tests. F 5.6-16. I only had time to upload the f5.6 one. But f5.6 is where it’s easiest to see where the focus is and where it might be moving to. I understand that f5.6 is not where you want to be shooting in actuality, but I’m trying to see if this lens is working first.

When I look at the swing set in the center photo of the image set it’s slightly out of focus. When I pan to the right the swing set is now in sharp focus. My thinking (which might be wrong) is that the swing set should remain out of focus through any other images in the pan it shows up in. It’s not on the edge of the image circle. It’s still in what I think is the central portion of the Lens. That part shouldn’t change unless there’s an issue, correct?

 

[Gerd]

The question is which standard have you moved to Stitsch?

When you move the Front Standard, you change the object's width (object distance) and the focus can move. Furthermore, you get parallax error.

Greeting Gerd

 

[PhiloFarmer]

Well......if you are focusing with the standards...you will ALWAYS have to deal with either PERSPECTIVE-SHIFT (front-standard)....or GEOMETRY-SHIFT (rear-standard)....

You get MUCH BETTER results SHIFTING with a SLIDING-BACK.....NO CHANGES in perspective OR geometry....

Cheers....

Hank

 

[MrSmith]

presuming you used the rear standard to shift across on the actus. did you lock once shifted? that might be enough movement to alter focus, i know when shooting close up locking the focus will shift the image slightly, might be the same for shifting

 

[Viewmaster]

The actus only allows shifting on the rear standard only. The front has tilt and swing, both of those were left alone.

Mr. Smith, I noticed that about the actus focus changing slightly when locking the knob down. But I checked that and compensated for it before I started shifting.

 

[vjbelle]

What you are describing does sound like the front and rear standards not being parallel. How did you check for parallelism? I suggest you find something that has enough detail at a distance of maybe 50 feet and check for parallelism visually at 100% pixels. You should be able to shift maximum left and right and have equal sharpness on the edges of the image. If not then make a slight adjustment of the front standard swing until both visually match. This can be accomplished in a very short period of time. At that point take your images and check them on your computer. They should very closely match what you saw on the LCD.

Victor

 

[Geoff]

The issue of parallelism exists on other cameras, like the Linhof Techno. When testing lenses wide open, there was similar softness on one side vs. the other. Fiddling with swing didn't get it under control easily - as the correction is small and hard to regulate. Simpler answer is to shoot at f11 and not worry. With these lenses, shooting wide open may be a recipe for heartache.

 

[darr]

I am not convinced you are keeping the same plane of focus in your shots. I have never used an Actus camera, but I have used a few different bellows/rail cameras with digital backs. If you are going to use a rail type camera for stitching, you need to keep the sensor as flat as possible to the plane of focus. This is harder to do than it sounds with standards and rail focusing as the slightest movement of any component involved with the setup (including tripod), will cause a reaction that may show up in individual shots. From my experience, standards and rail focusing mechanisms all have a bit of slop in them, thus one reason pancake cameras and sliding backs were invented. You keep using the word panning which means you are moving the camera on the tripod from side to side, this introduces more possibility for precision problems.

You spoke of the chair in your pictures in and out of focus, and you are shooting at an aperture of 5.6 and panning from side to side. I would expect things that are not on the plane of focus to move in and out of focus when doing that. Here are a few snapshots to help you get the idea of what works for me to get a sharp panorama from the SK 72 at f/11:

1. The camera must be level.

2. The camera must be locked from movement.

3. Are you focusing on a particular subject, or hyperfocal distance focusing?
Your DoF with a 72mm lens is limited by design (pretty much a telephoto lens). Pick what you want in focus and shoot at f/11-f/16, or learn where your particular lens' hyperfocal sweet spot is.

4. Does the camera, lens, tripod, and/or standards move at all during sensor movement?

5. Example shot with the SK 72 @ f/11 with focus on fence and shot on an angle to show some lens DoF quality.
The image is squeezed down from 63.5". The fence and its components are in focus.

Hope this helps!

 

[Geoff]

I am not convinced you are keeping the same plane of focus in your shots. I have never used an Actus camera, but I have used a few different bellows/rail cameras with digital backs. If you are going to use a rail type camera for stitching, you need to keep the sensor as flat as possible to the plane of focus. This is harder to do than it sounds with standards and rail focusing as the slightest movement of any component involved with the setup (including tripod), will cause a reaction that may show up in individual shots. From my experience, standards and rail focusing mechanisms all have a bit of slop in them, thus one reason pancake cameras and sliding backs were invented. You keep using the word panning which means you are moving the camera on the tripod from side to side, this introduces more possibility for precision problems.

You spoke of the chair in your pictures in and out of focus, and you are shooting at an aperture of 5.6 and panning from side to side. I would expect things that are not on the plane of focus to move in and out of focus when doing that. Here are a few snapshots to help you get the idea of what works for me to get a sharp panorama from the SK 72 at f/11:

1. The camera must be level.

2. The camera must be locked from movement.

3. Are you focusing on a particular subject, or hyperfocal distance focusing?
Your DoF with a 72mm lens is limited by design (pretty much a telephoto lens). Pick what you want in focus and shoot at f/11-f/16, or learn where your particular lens' hyperfocal sweet spot is.

4. Does the camera, lens, tripod, and/or standards move at all during sensor movement?

5. Example shot with the SK 72 @ f/11 with focus on fence and shot on an angle to show some lens DoF quality.
The image is squeezed down from 63.5". The fence and its components are in focus.

Hope this helps!

Just one thing to add to this excellent post - that some of the DOF scales on the lenses are for film and are much more generous than for digital backs. DOF for digital is a much thinner zone than you might imagine.

 

[Audii-Dudii]

The issue of parallelism exists on other cameras, like the Linhof Techno. When testing lenses wide open, there was similar softness on one side vs. the other. Fiddling with swing didn't get it under control easily - as the correction is small and hard to regulate. Simpler answer is to shoot at f11 and not worry. With these lenses, shooting wide open may be a recipe for heartache.

When I was using my Cambo Actus in the field, I had enough difficulty keeping the front and rear standards sufficiently parallel that I eventually took to carrying a small, telescoping snap-gauge with me and used it several times during my outings to verify the gap between the front and rear standards was as to close the same distance top-to-bottom and side-to-side as possible:

At least on my Actus and with the wide- and ultra-wide angle lenses I used (which have very little depth of focus to cover any errors), the accuracy of the hash marks silkscreened onto the front standard for the tilt and swing movements wasn't sufficient for it to be possible to truly zero the alignment of the standard without actually measuring it.

Of course, the GFX version has a larger diameter rear standard, so I don't know if this method will work as well as it did on the original version of the Cambo Actus. But given the symptoms being described here, non-parallelism of the front and rear standards is the very first thing I would address if I was the original poster. <shrugs>