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!

Alignment of IQ back

cunim

Well-known member
I recently picked up an RM3di (used) and have noticed that the IQ180 and the camera disagree about pitch level.

1. The bubble level on the camera frame shows a bit more than a half bubble of tilt when the electronic level in the IQ shows no tilt.
2. The IQ level shows a bit more than 1 notch (not sure what those correspond to) of backwards tilt when the camera bubble shows level.
3. A carpenter's level agrees with the camera bubble. When the camera shows level, so does the carpenter's level.
4. I think the camera frame is fine. The rear extension "box" (held by the rear clamp) lies in the same plane as the lens axis (held in the front clamp). The misalignment is most likely in the Mamiya adapter or the back itself.

To verify, I made some shots of a flat target with the apo digitar 150 at f5.6, at a distance of about 12'. Electronic level on the IQ set to zero. At 0 lens tilt, both top and bottom of the target go a bit soft. At 0.5 down tilt, much worse. At 0.5 up tilt everything looks good across the frame. At 1' up tilt it is terrible.

From all this I conclude the sensor plane is tilted about 0.5' back from the lens axis. On my old Alpa I would have shimmed the adapter but that is not possible here. Tell the truth, I never much cared about shimming the Alpa for infinity, but pitch and yaw parallelism are important. Any suggestions? Anyone know what Arca considers a reasonable tolerance for this type of thing?
 

miska

Member
Hi,
Could there bit of play between the digital back and the camera adapter plate (from Arca Swiss) ?
At least mine is not 100% snug, and a similar effect could cause what you are seeing. Although 0.5 degrees seems like a lot...
 

GrahamWelland

Subscriber & Workshop Member
You can reset the back's level setting if you need to. Press and hold the level display and you should be able to reset it.

However, if what you are seeing is softness due to lack of physical alignment then I don't think that you have any option but to get everything recalibrated together.

Generally speaking, I trust the levels on my Alpa and not those on the IQ back. Even reseting them doesn't seem 100% effective as there seems to be drift (more so with the IQ260 vs my 160 which seemed more consistent). I found that you have to pick one set of levels and stick with it otherwise you'll enter into a whole world of frustration. See earlier threads here about trying to calibrate bodies, backs and Cubes together. Here's what you'll experience: :banghead::banghead:
 

cunim

Well-known member
Here I am, foaming (slightly) at the mouth - pixel peeper on a technical blood trail. Last time I was this excited, the A7r shutter shake issue was big.

There is no slack I can see between the adapter and the camera frame. The Alpa had more. Miska, I misspoke. Hard to tell absolute tilt on the 3di. It's a coarse adjustment. I think the error is more like 0.25' than 0.5'. Possibly, you'd never notice it with a normal or wide lens, or with a longer lens at infinity.

Graham, I am not overly concerned with absolute level. I am concerned with having the sensor plane precisely orthogonal to the lens axis.

So I whipped out the Monolith and mounted the back on that. Centered all the bubble levels and reset the IQ level indicator as Graham suggested. All the little bubbles show DNL (dead nuts level). Shot a series at about 0.1' intervals using the Orbix to tilt. Problem not present. Zero tilt on the view camera yields even focus across the plane.

That means it is something in the 3di, or some mismatch between the 3di and the Mamiya adapter that I use with the Monolith. I'll have another adapter in a little while and will see how it does.

Anyone else with a 3di feel like shooting a wall with a long lens, wide open? Interesting to know if I am in or out of tolerance limits.

Peter
 

jlm

Workshop Member
my comments, maybe you already considered them.

bubble levels are made to varying degrees of accuracy; carpenter's levels are not so good, machinist's levels are much better and much, much better.

when i shoot flat art, i use a laser enlarging alignment tool to make sure, first, that my lens is perpendicular to the flat plane of the art. (basically place the laser source flat against the plane, so now the laser line is orthogonal both ways to the plane, then place a flat glass strip across the lens barrel in such a way the the laser dot reflects off the strip right back on itself and confirm the reflection goes right back into the beam where it originates.) now you have the lens axis set right and don't need to play with the bubble levels or even be concerned with level.

http://www.bhphotovideo.com/c/product/157407-REG/Versalab_PARALLEL_Parallel_Alignment_Gauge.html

now take a picture and check corner/center focus. ideally, you would then tilt/swing the back to get the back orthogonal, using corner/center sharpness as a guide.
you would have to figure a way to make the back do the moving, not the lens.

with the Cambo, it was possible to turn small adjusting screws in the body, effectively tilting/swing the back a tiny bit
 
Last edited:

jagsiva

Active member
Based on what you did with the Monolith, sounds like you can isolate the problem to the RM3Di. Also eliminates any issue of the IQ180 being flush with the Arca M645 mounting plate (again assuming you used the same plate).

I would check the RM3Di at the back where the adapter plate goes in. The bottom has a slot that you set the plate into before locking with the wheel at the top. Sometimes, dirt can get into the bottom slot and not allow the mounting plate to sit properly before it is locked with the wheel.

I would suggest using some compressed air first to see it helps. You can then move to more aggressive methods to clean it out.

When mounted flush, all four corners should be flat with the camera body, run a finger over each corner and see it is sitting the same all around.

In almost every case, I would go with the camera bubbles rather than the electronic one in the IQ, but in this case, sounds like the issue is not with the levels, but with the camera and back being parallel when mounted.

Good luck...
 

Bill Caulfeild-Browne

Well-known member
A suggestion for calibrating the level in the IQ backs - find a truly flat horizon. An ocean or a large lake will do! No need to use carpenters levels - just take pix with the horizon near the top (or bottom) of the frame. When it's truly level, adjust the back accordingly.

Still, I think there is some drift in the IQ back and I recalibrate every few months.
 

cunim

Well-known member
I believe I have my answer. An Arca dealer has told me that he has seen this before, and that some "Tum" units (the tiltable frame/bellows in the 3di) are not precisely aligned. Precision Camera (Arca repair in the US) is not equipped to correct this. The camera must be sent to France where Martin will do the necessary.

Might be a good idea for Arca owners to check this on their own R-series cameras. Mind you, a tilt of <0.5' would only matter under conditions that require precise focal plane positioning. Subject should be flat (e.g a wall), imaged with an open aperture and a longer lens. More an architecture or macro type of thing than landscape. Under these conditions, the misalignment is obvious.

God knows, the lenses go out easily enough (I have had to have two Rodies realigned) that I really don't want to be worrying about the camera as well. Guess I will send mine off soon.
 
I believe I have my answer. An Arca dealer has told me that he has seen this before, and that some "Tum" units (the tiltable frame/bellows in the 3di) are not precisely aligned. Precision Camera (Arca repair in the US) is not equipped to correct this. The camera must be sent to France where Martin will do the necessary.

Might be a good idea for Arca owners to check this on their own R-series cameras. Mind you, a tilt of <0.5' would only matter under conditions that require precise focal plane positioning. Subject should be flat (e.g a wall), imaged with an open aperture and a longer lens. More an architecture or macro type of thing than landscape. Under these conditions, the misalignment is obvious.

God knows, the lenses go out easily enough (I have had to have two Rodies realigned) that I really don't want to be worrying about the camera as well. Guess I will send mine off soon.
A misalignment between the lens, the camera and the back is a fact of life just as dust inside lens. Just like you cannot expect 4 corners of a lens to have equal resolution (see various lens data at slrgear), the accumulative systematic error can cause a drift of 10-40 pixels along any direction of the picture when you rotate the lens or the digital back on an Alpa camera body. Alpa suggested this to be within specification and tolerance. I have tested this matter on just many Alpa camera bodies (SWA, STC, FPS) as well as Cambo and Sinar, and they all behaved the same way. This in real world scenario does not significantly affect sharpness after correction of keystones in post-processing, but is unfortunately a nuisance fact for perfectionists to accept.
 

cunim

Well-known member
A misalignment between the lens, the camera and the back is a fact of life just as dust inside lens. Just like you cannot expect 4 corners of a lens to have equal resolution (see various lens data at slrgear), the accumulative systematic error can cause a drift of 10-40 pixels along any direction of the picture when you rotate the lens or the digital back on an Alpa camera body. Alpa suggested this to be within specification and tolerance. I have tested this matter on just many Alpa camera bodies (SWA, STC, FPS) as well as Cambo and Sinar, and they all behaved the same way. This in real world scenario does not significantly affect sharpness after correction of keystones in post-processing, but is unfortunately a nuisance fact for perfectionists to accept.
True in general, but I respectfully disagree in this case. If AS can build the Monolith - with all of its movement points - so that it carries the sensor reasonably parallel to the lens, they should be able to do the same with the little 3di with one tilt axis to worry about. 0.25' is a fairly gross optical misalignment and has real world consequences. The dealer mentioned that a small proportion of 3di bodies have shown this type of error and that it is correctable - in France.

I suspect the unit just needs adjustment so that the sensor axis is vertical when the tilt adjustment is in the detent. I'll check it out and report back.
 

MGrayson

Subscriber and Workshop Member
Forgive me, I am confused by the units. As an angular measure, 0.25' reads as 1/4 minute or 15 arc-seconds - smaller than the planet Jupiter as seen from earth unmagnified. Is that the error under discussion? Or is it 1/4 degree, which I agree would be a gross error.

Thanks,

Matt
 

cunim

Well-known member
Forgive me, I am confused by the units. As an angular measure, 0.25' reads as 1/4 minute or 15 arc-seconds - smaller than the planet Jupiter as seen from earth unmagnified. Is that the error under discussion? Or is it 1/4 degree, which I agree would be a gross error.

Thanks,

Matt
Sorry, Matt. I can't find the degree symbol on my keyboard. It is 1/4 degree - approximately because the tilt adjustment is not calibrated that finely.

Peter
 

jlm

Workshop Member
i'm sure he means 1/4 degree; the Tilt scales are all marked in degree increments.

over the about 3" span of the back, 1/4 degree represents a .013" discrepancy, about 1/64th inch, about 0.3mm. that is a huge amount; these things are shimmable to .001"
 

MGrayson

Subscriber and Workshop Member
Oh! Thank you for the clarification, Peter. That's huge, and really unacceptable in one of these systems.

Take an image on screen and rotate it by 1/4 degree. It's not subtle!

Matt
 

cunim

Well-known member
i'm sure he means 1/4 degree; the Tilt scales are all marked in degree increments.

over the about 3" span of the back, 1/4 degree represents a .013" discrepancy, about 1/64th inch, about 0.3mm. that is a huge amount; these things are shimmable to .001"
Thanks for clarifying. I hadn't realized it was quite this much.
 

jagsiva

Active member
Peter,

If you have been talking to Walter in Toronto, I suggest you take your camera into him and try your "TUM" module with his camera as well as his "TUM" with yours. If the DB was properly calibrated for level, then it should not be off relative to the bubble levels on the camera - indicating that the problem may not be with the "TUM".

I still think you may have some dirt where the adapter plate goes in the body. I would isolate the problem for sure before sending to Martin as that is going to take you a few weeks.

If you in SWO or close to London, you're welcome to try with mine.

Cheers...
 

stephengilbert

Active member
The degree symbol (on a Mac, at least) is Option 0: º

I've got a Matias Tactile Pro KB, which lists the symbols for each key.

So I can get Jürgen's name right. :)
 

cunim

Well-known member
Peter,

If you have been talking to Walter in Toronto, I suggest you take your camera into him and try your "TUM" module with his camera as well as his "TUM" with yours. If the DB was properly calibrated for level, then it should not be off relative to the bubble levels on the camera - indicating that the problem may not be with the "TUM".

I still think you may have some dirt where the adapter plate goes in the body. I would isolate the problem for sure before sending to Martin as that is going to take you a few weeks.

If you in SWO or close to London, you're welcome to try with mine.

Cheers...
No dirt or other visible issues.

Walter has just sent his demo 3di off to France to integrate with the new e-shutter. Guess I will do the same as I plan to install that system. As usual, Martin was quick to respond to my support request and suggested some tests that tend to confirm a misalignment in the camera as opposed to the IQ back. He says misalignment is rare and he confirms the unit must go back to France for adjustment. As part of these tests I found a fairly gross problem in the Rotaslide so that will go back too.

Dealing with Arca Swiss is refreshing. They may not have a slick web site or deliver product quickly, but you can still carry on a conversation with a knowledgeable principal. That is a rare thing these days.

Peter
 
Top