Site Sponsors
Results 1 to 18 of 18

Thread: Testing for decentered lens

  1. #1
    Senior Member
    Join Date
    Sep 2009
    Posts
    1,587
    Post Thanks / Like

    Testing for decentered lens

    Hi all,

    I have an inkling that my Rodenstock 90mm HR-W is decentered and am wondering how I might go about methodically testing for this? I loathe doing test shots, but this might be one instance I need to suck it up and do some just to put my mind at ease.

    Basically I've been shooting some 6x13 ratio, stitched panoramic images with my Credo 60 and the left side of the frame seems to fall off, with the plane of focus having the appearance that the front standard has some slight swing engaged. Of course there is always a small chance that the front standard of my Techno is not exactly parallel, but the problem doesn't seem to exist with my Rodenstock 55mm APO–SD. I'm shooting almost exclusively at f11.

    Thanks,

    Tim

  2. #2
    Senior Member
    Join Date
    Feb 2011
    Location
    Eads, Tennessee
    Posts
    1,032
    Post Thanks / Like

    Re: Testing for decentered lens

    I have tested all of my Digitar lenses for skweing (decentered) and found three out of the four to need servicing. That's the bad news...... the good news is that Schneider is extremely easy to work with and 'fixes' the problem the first time under warranty - at least for me.

    I usually start with a row of buildings - usually a strip mall. With a 90mm lens I would get back probably 200 feet perpendicular to the buildings. Shoot an image centered wide open and then shift left and right 10mm for two more images. Flip the lens 180 degrees and repeat the shooting sequence. Check the edges at 100% pixels and if they don't match for sharpness the lens is probably decentered. Next I look for a scene which will allow for depth analysis...... usually a fence where the far beyond the fence and near in front of the fence can be seen and analyzed. Focus on the fence at - probably closer than 200 ft and shoot as above. Check the edges and see if you can find the focus point - should be either far or near for the side that is out of focus. If you are convinced the lens is skewed then contact Rodenstock for service. I have always dealt with Schneider Germany direct. I should add that I use my distances because I have always felt that going further out increases dof and can mask decentering issues.

    Good luck......

    Victor
    Last edited by vjbelle; 5th August 2015 at 07:11.
    Thanks 1 Member(s) thanked for this post

  3. #3
    Senior Member ondebanks's Avatar
    Join Date
    Feb 2009
    Posts
    518
    Post Thanks / Like

    Re: Testing for decentered lens

    What I do is shoot city lights from a suitable vantage point, tilting the camera to run the line of lights diagonally across the sensor, like this:



    Compare the point-source light images to each other on opposite sides of the frame. You are looking for consistency, at a given radial distance off-centre.

    Most problems show up worst when the lens is wide open, but you should run the test at all apertures.

    For real peace of mind, run it a second time with the horizon running across the other diagonal.

    Ray
    Thanks 1 Member(s) thanked for this post
    Likes 2 Member(s) liked this post

  4. #4
    Senior Member
    Join Date
    Dec 2007
    Location
    Little Rock AR
    Posts
    1,925
    Post Thanks / Like
    Images
    3

    Re: Testing for decentered lens

    The test with city lights is a good one. But any subject that has a nice even line to it works. I live near a river and can easily setup on one side and pan along the bank.

    I have come to believe that the QA of the Rodenstock glass is not what it used to be. I have had several lenses "new" from Rodenstock that all had some slight off centering issues. Mainly only showing when shifted.

    The issue you run into, is can Rodenstock make it better? or will they make it worse. I have come to realize that Rodenstock may not have any high end digital backs to test with (60 or 80MP). And Arca only seems to have a 33Mp back in France, that may be wrong, but it was what I was told when I had a problematic Rodenstock recently.

    The 33Mp back IMO is not the best test for focus and or decentering testing, when you are shooting an 80 or 60MP back, which don't allow for anywhere as much errors.

    For the price point of the glass, it's a bit surprising to me that the lenses can't come from the factory with perfect alignment. But it also might be a problem with the camera mount. Anyone's guess. It might just happen when they are shipped as harsh handling might cause the copal shutter to get slightly out of alignment. No doubt the weak link in the chain is the copal shutter as with the heavier front element retrofocus Rodenstocks, the added weight can cause mis-alignment of the shutter. That has to go all the way back to Rodenstock for re-colimation.

    I have found that on many of many tech lenses, slight adjustments, mainly swing on shifts may need to be done to get a tack sharp focus across the shifted frame @ F11 or so.

    Paul
    Paul Caldwell
    [email protected]
    www.photosofarkansas.com
    Thanks 1 Member(s) thanked for this post

  5. #5
    Senior Member
    Join Date
    Sep 2009
    Posts
    1,587
    Post Thanks / Like

    Re: Testing for decentered lens

    Thanks all for the very helpful replies, I appreciate people taking the time to answer. I'll try get out and test end of this week and report back.

    With regards to Rodenstock having a back to test with, when I send the same lens in to query if the central flare I was seeing was normal – the now well documented flare that pops up as a big, hazy blob centre of frame. Turns out it is normal :-( – they sent back a set of test images shot on an IQ180.

    Other than the flare, which I don't seem to experience when using a Lee hood, I've found it to be a stunning lens, all be it on 6x7cm film. On film I've shifted 20mm and never noticed an decentering issues, but film is far more forgiving. Stitching with 8mm back fall and -/+15mm left and right on a 60mpx back pushes the lens much harder.

    Anyway, thanks again for all the replies.

    Tim

  6. #6
    Senior Member
    Join Date
    Aug 2010
    Location
    Maryland
    Posts
    475
    Post Thanks / Like

    Re: Testing for decentered lens

    Quote Originally Posted by Paul2660 View Post
    I have come to realize that Rodenstock may not have any high end digital backs to test with (60 or 80MP).
    My calendar must be way off.

    I didn't realize today was April 1st.

    Do you seriously think an optical company is going to test a lens with something as inaccurate as a camera sensor?

    Optical test equipment is far more precise than anything you could buy in the photographic product arena.

    If they want to do some "real-world" tests to confirm compatibility, I'm sure they can afford a back.

    - Leigh

  7. #7
    Senior Member
    Join Date
    Dec 2007
    Location
    Little Rock AR
    Posts
    1,925
    Post Thanks / Like
    Images
    3

    Re: Testing for decentered lens

    Quote Originally Posted by Leigh View Post
    My calendar must be way off.

    I didn't realize today was April 1st.

    Do you seriously think an optical company is going to test a lens with something as inaccurate as a camera sensor?

    Optical test equipment is far more precise than anything you could buy in the photographic product arena.

    If they want to do some "real-world" tests to confirm compatibility, I'm sure they can afford a back.

    - Leigh
    Leigh,
    Not sure the reason for the poke, but if you have a reason to poke at me, whatever. I personally find your tone a bit offensive. I will never understand the reason some people feel it's ok to make a personal attack on a web forum.

    To be honest, I guess your knowledge is far greater than mine. I just use the gear.

    However as a layperson, I would feel that no matter how good the rocket scientist optical design equipment is, a good final test would be to use the lens on a 80MP back as a test, as that is what the HR-W series is designed for. Net, Rodenstock designed the HR-W for the 80MP backs, period. Go ask as you obviously don't feel my opinion is worth anything more than a joke, i.e. April 1st.

    The actual joke, is paying 7.5K for a lens that is obviously not centered on a 60 or 80MP back and then getting into a long lasting discussion with Rodenstock on the process to fix the lens. No that's not a joke that's reality at least mine.

    I personally find your tone a bit offensive.

    As the owner, user of 4 of the HR or HR-W lenses, 3 of which where purchased new and all arrived off center to my 60MP back, I can clearly state, that Rodenstock does not test the equipment as well as it could be as for 5k to 7.5K per lens having a lens arrive off center, or off center after a shift of 10mm or less tells me that said lens did not get tested very well after leaving Rodenstock.

    You obviously don't hold my opinion very high, which is your choice, but if you take just a bit if time and search this forum and others related to medium format tech lens issues you will find many similar posts of uses with issues.

    I have been told that during a test of 4 32mm HR-W lenses at one of the tech camera companies, it was very apparent that from the 4 lenses, only 1 was stellar across the shift range of 12mm. So with all that great optical equipment that is so much better than a digital back, I guess they are just not paying close attention or possibly they don't care, go figure.

    Obviously, your mileage may vary, and I wish you the best in your pursuits.

    Paul
    Paul Caldwell
    [email protected]
    www.photosofarkansas.com
    Likes 1 Member(s) liked this post

  8. #8
    Subscriber and Workshop Member MGrayson's Avatar
    Join Date
    Nov 2010
    Location
    NYC
    Posts
    1,575
    Post Thanks / Like
    Images
    4

    Re: Testing for decentered lens

    Paul,

    I recently came back from a workshop with a decentered SK35XL in Cambo mount. The lens went off to Netherlands ... and came back worse. After the second trip it was better, but I no longer have faith that paying for technical camera equipment even raises the probability of getting well calibrated gear, never mind guaranteeing it.

    There's a lot to be said for using gear that's available in quantity from rental houses so that when, not if, something goes out of whack there's a quick fix.

    Best,

    Matt

  9. #9
    Senior Member
    Join Date
    Aug 2010
    Location
    Maryland
    Posts
    475
    Post Thanks / Like

    Re: Testing for decentered lens

    Quote Originally Posted by Paul2660 View Post
    I personally find your tone a bit offensive.
    Hi Paul,

    No offense intended. I was poking fun at the opinion, not at the author thereof.
    I would have made the same comment regardless of who posted that.

    I'm like a certain popular Presidential hopeful... I say what I think. Honest opinions, 100% of the time.

    Quote Originally Posted by Paul2660 View Post
    As the owner, user of 4 of the HR or HR-W lenses, 3 of which where purchased new and all arrived off center to my 60MP back, I can clearly state, that Rodenstock does not test the equipment as well as it could be as for 5k to 7.5K per lens having a lens arrive off center, or off center after a shift of 10mm or less tells me that said lens did not get tested very well after leaving Rodenstock.
    "... after leaving Rodenstock." ???

    So who would test the lens after it leaves the manufacturer?

    Sinar did that, buying lenses in quantify, running their own tests, and keeping only those that passed.

    I don't think that was your intent.

    Admittedly, a de-centered lens should never leave the factory in the first place.
    But that's a fantasy in the 3rd Millennium.

    Back in the "good ole days", top-quality lenses were individually tested before shipment.
    Any problems were caught and corrected.

    With modern computerized manufacturing and statistical process control, they may only test 1 in 10 or 1 in 100.
    We lucky folks in Consumer Land get to test the other 99 (or 9).

    =====

    Regarding centering...

    A finished lens is not centered as a unit.

    Every lens element must be evaluated individually for centering.
    With modern computerized lens grinding this may be less of an issue than in earlier decades.

    Then the individual elements are assembled into cells, with the optical axes of individual elements all aligned.
    The optical axis of the resulting cell may be off-centered.

    That is supposed to be corrected in final assembly.

    To change the centering of a finished lens, it must be disassembled and that final correction step repeated.
    That is a complex procedure, requiring much time and effort by trained techs with expensive equipment.

    Given that modern products often use adhesives where fasteners were previously used, disassembly can be very difficult, possibly destroying some cosmetic or functional components of the lens.

    We're not talking about torquing a setscrew by 1/4 turn and sending the lens back to the owner.

    - Leigh

  10. #10
    Senior Member ErikKaffehr's Avatar
    Join Date
    Oct 2012
    Location
    Nyköping Sweden
    Posts
    1,191
    Post Thanks / Like

    Re: Testing for decentered lens

    Hi,

    My impression used to be is that image projectors are mostly used to test assembled lenses.

    Using a digital sensor is quite demanding, the P65+ sensor is about 6 micron pitch corresponding to 83 lp/mm. Most published MTF-data go to 40 lp/mm (10/20/40), even if Schneider and Rodenstock publish data at 60 lp/mm (15/30/60).

    So, clearly a 60 MP sensor clearly takes lenses into the low MTF region. So I would say for certain that a 60 or 80 MP sensor is a challenge for any lens.

    Quite a few years ago Joseph Holmes published an article about frequent issues with MFD equipment and your observations may lay in line with that article.

    I would assume that MFD is not alone having issues with tolerances. Photozone.de often reports on decentered lenses. On the other hand, Roger Ciala is looking into sample variation on different lenses like here: http://www.lensrentals.com/blog/2015...e-angle-lenses

    I got the impression that new lenses usually are quite OK, but they may wear out. Simple designs may be better than more complex designs. Please not that Roger's tests are normally at maximum aperture. All lenses perform better when stopped down.

    Best regards
    Erik

    Quote Originally Posted by Paul2660 View Post
    Leigh,
    Not sure the reason for the poke, but if you have a reason to poke at me, whatever. I personally find your tone a bit offensive. I will never understand the reason some people feel it's ok to make a personal attack on a web forum.

    To be honest, I guess your knowledge is far greater than mine. I just use the gear.

    However as a layperson, I would feel that no matter how good the rocket scientist optical design equipment is, a good final test would be to use the lens on a 80MP back as a test, as that is what the HR-W series is designed for. Net, Rodenstock designed the HR-W for the 80MP backs, period. Go ask as you obviously don't feel my opinion is worth anything more than a joke, i.e. April 1st.

    The actual joke, is paying 7.5K for a lens that is obviously not centered on a 60 or 80MP back and then getting into a long lasting discussion with Rodenstock on the process to fix the lens. No that's not a joke that's reality at least mine.

    I personally find your tone a bit offensive.

    As the owner, user of 4 of the HR or HR-W lenses, 3 of which where purchased new and all arrived off center to my 60MP back, I can clearly state, that Rodenstock does not test the equipment as well as it could be as for 5k to 7.5K per lens having a lens arrive off center, or off center after a shift of 10mm or less tells me that said lens did not get tested very well after leaving Rodenstock.

    You obviously don't hold my opinion very high, which is your choice, but if you take just a bit if time and search this forum and others related to medium format tech lens issues you will find many similar posts of uses with issues.

    I have been told that during a test of 4 32mm HR-W lenses at one of the tech camera companies, it was very apparent that from the 4 lenses, only 1 was stellar across the shift range of 12mm. So with all that great optical equipment that is so much better than a digital back, I guess they are just not paying close attention or possibly they don't care, go figure.

    Obviously, your mileage may vary, and I wish you the best in your pursuits.

    Paul

  11. #11
    Senior Member ondebanks's Avatar
    Join Date
    Feb 2009
    Posts
    518
    Post Thanks / Like

    Re: Testing for decentered lens

    Quote Originally Posted by Leigh View Post
    Hi Paul,
    I'm like a certain popular Presidential hopeful... I say what I think. Honest opinions, 100% of the time.
    Presumably, apart from the dress and the bobbed hairstyle. Go Hillary!

    Ray

  12. #12
    Senior Member
    Join Date
    Aug 2010
    Location
    Maryland
    Posts
    475
    Post Thanks / Like

    Re: Testing for decentered lens

    Quote Originally Posted by ondebanks View Post
    Presumably, apart from the dress and the bobbed hairstyle. Go Hillary!
    Ray
    Sorry, Ray. You missed on both counts.

    I have no hair, and I don't wear dresses.

    - Leigh

  13. #13
    Senior Member ondebanks's Avatar
    Join Date
    Feb 2009
    Posts
    518
    Post Thanks / Like

    Re: Testing for decentered lens

    Quote Originally Posted by Leigh View Post
    I have no hair, and I don't wear dresses.
    Leigh, that's what I said: "apart from" those things...

    Ray

  14. #14
    Senior Member
    Join Date
    Feb 2011
    Location
    Eads, Tennessee
    Posts
    1,032
    Post Thanks / Like

    Re: Testing for decentered lens

    Quote Originally Posted by Paul2660 View Post

    As the owner, user of 4 of the HR or HR-W lenses, 3 of which where purchased new and all arrived off center to my 60MP back, I can clearly state, that Rodenstock does not test the equipment as well as it could be as for 5k to 7.5K per lens having a lens arrive off center, or off center after a shift of 10mm or less tells me that said lens did not get tested very well after leaving Rodenstock.

    Paul
    I can relate to all of this. Schneider is no better and, for sure, Alpa doesn't help matters. I have, though, had great experiences with Schneider who actually DO fix the lens the first time. I'm amazed at the users who pay tons of money for this stuff and don't even know that a lens is out of wack. I do hope that Rodenstock treated you as well as Schneider treated me.

    Victor

  15. #15
    Senior Member
    Join Date
    Feb 2011
    Location
    Eads, Tennessee
    Posts
    1,032
    Post Thanks / Like

    Re: Testing for decentered lens

    Quote Originally Posted by Leigh View Post
    Do you seriously think an optical company is going to test a lens with something as inaccurate as a camera sensor?

    Optical test equipment is far more precise than anything you could buy in the photographic product arena.

    If they want to do some "real-world" tests to confirm compatibility, I'm sure they can afford a back.

    - Leigh
    I don't think that this is a compatibility issue but rather a manufacturing issue. Otherwise, how does a lens actually get repaired/aligned so as not to show decentering without the accompanying back? If these manufacturer's have all of that sophisticated test equipment then they should use it.

    Victor

  16. #16
    Senior Member
    Join Date
    Aug 2010
    Location
    Maryland
    Posts
    475
    Post Thanks / Like

    Re: Testing for decentered lens

    Quote Originally Posted by vjbelle View Post
    I don't think that this is a compatibility issue but rather a manufacturing issue. Otherwise, how does a lens actually get repaired/aligned so as not to show decentering without the accompanying back? If these manufacturer's have all of that sophisticated test equipment then they should use it.
    Victor,

    I'm sorry, but I fear you have a very distorted view of lens manufacture.

    A lens is not designed to work with a particular back, even if such lens is permanently attached to a camera body.

    The optical axis of a lens is supposed to be at the geometric centerline of the housing.
    There may be exceptions for special cases and special lenses, but that rule is the norm.

    The center of the sensor should correspond to its physical midpoint.

    For a lens to be "centered" in normal use, its optical axis must pass through the center of the sensor.

    De-centering is defined as the condition where that correlation fails.
    It could be a fault of the lens, the lens flange on the camera body, or the mounting of the sensor within the body.

    - Leigh

  17. #17
    Senior Member
    Join Date
    Feb 2011
    Location
    Eads, Tennessee
    Posts
    1,032
    Post Thanks / Like

    Re: Testing for decentered lens

    First of all, I appreciate your extensive knowledge regarding lens manufacturing/testing..... I really do! I don't have that kind of knowledge and so I have to make deductions with the limited visual information at hand. If a lens shows by visual inspection at 100% pixels that on a planar subject one side/top/bottom at the edge is out of focus compared to the same location at the opposite side of the file then something is surely amiss. I agree that it could be numerous issues including flange or sensor but with all of my circumstances the issues was resolved by sending my lenses back to Schneider (in my experiences). I can only deduce that the lens was at fault..... not the flange or sensor. I also always got an invoice (at no charge) showing that the lens was out of alignment and corrections made.

    Appreciate your input......

    Victor

  18. #18
    Member
    Join Date
    Oct 2009
    Location
    Arizona
    Posts
    128
    Post Thanks / Like
    Images
    5

    Re: Testing for decentered lens

    Quote Originally Posted by vjbelle View Post
    First of all, I appreciate your extensive knowledge regarding lens manufacturing/testing..... I really do! I don't have that kind of knowledge and so I have to make deductions with the limited visual information at hand. If a lens shows by visual inspection at 100% pixels that on a planar subject one side/top/bottom at the edge is out of focus compared to the same location at the opposite side of the file then something is surely amiss. I agree that it could be numerous issues including flange or sensor but with all of my circumstances the issues was resolved by sending my lenses back to Schneider (in my experiences). I can only deduce that the lens was at fault..... not the flange or sensor. I also always got an invoice (at no charge) showing that the lens was out of alignment and corrections made.

    Appreciate your input......

    Victor
    Victor you are correct in many of your findings etc. I will add more food for thought. When you have an AF lens it must be slightly loose relative to a manual focus lens, as very small motors must move that mass. Doesn't mean it can't be sharp, but it does mean it may not be quite as sharp as manual focus style lens, if both are right off the assembly line. But, as someone said, shipping it can result in a bump, fall, etc, in the box. And then an individual piece of the component glass inside the lens can slightly decenter. It only takes one.
    I always contended that if you gave someone at the top level, say Zeiss, a 'crapflex' lens, and let them assemble and build it, then right off the line, made images, they would probably rival any lens available, even much more expensive glass, for sharpness. But after walking a block, with it bouncing on its strap against your hip or chest, the cheaper lighter construction would let it down, with elements de-centering, etc.
    Now a 35mm SK for MF is not a cheap lens, but it is subject to the comment about AF lenses, as well as the shipping and un-gentle use parameters.

    In the case of a technical LF lens, the culprit is several fold. First the Copal shutter, and its competition, now gone, and with them out of the mechanical shutter business as well now, has to deal with newer and heavier glass. This can cause a torque on the central bass barrel the front and rear lens groups are screwed into.
    Setting a tripod down, with a camera, and with a lens mounted, firmly or with some alacrity, can cause a large front element group, to impact the brass mounting tube, and over time the front and rear groups, lose alignment with each other. The larger the element in front, especially, the more prone to this occurring, in the lens. This is one of the reasons for problems cropping up. But putting pressure on a lens in your bag or back pack while transporting or dropping the bag or pack, can also be a culprit.

    For a while now, both Rodenstock and Schneider, have often included notes in their lens boxes, when delivered new, warning of this possibility. The recommend not carrying the camera on a tripod over the shoulder, for sure. Each step you take with a shouldered camera tripod, increases the torque on the lens groups, the larger the front group, the worse it is.
    So be gentle...

    Another factor involved for digital is definitely focus shift. The lens designers, had by the 1880's and definitely in the 1890's figured out how to get the blue and green to focus on the same plane. The red wavelengths however would mostly, be slightly in front of the image plane. Those who used film and remember on manual focus cameras, that one had to move the focus slightly to get the sharpest images.
    They had a red index for this adjustment on the barrel of the lenses, to show this correction.

    But first with digital sensors and lenses, there is another issue yet. The red still moves further away from the focus plane, as you stop down the lens. And although the focus depth increases, like depth of field, as you stop down, the red may move further forward then the far plane of focus, and so red is out of focus, muddying up your image. All lenses have focus shift, more or less.
    And due to the onset of Diffraction showing up much sooner, as you stop the lens down, with digital imaging, than it did on film, you cannot close your lens down as far as you could for film. And the smaller the pixels the worse the situation gets. So, with wide angle lenses to standard focal lengths, on a 36mp or 43mp or 50mp DSLR, for instance we find that 5.6 is the limit before diffraction shows up, where with film, you could go down to F8-11. Some of it can be corrected, but it sometimes makes image look 'Digy'. With an 80mp back on MF, F8.5 is the limit with standard to wide lenses that you can stop down. With some lenses, the red moves further, than that far plane of focus, as I said, so they cannot resolve the highest MP backs. That is also why many older lenses do not do as well as newer designs.

    Hope this helps.
    Rod
    Last edited by RodK; 11th August 2015 at 13:43.
    US Representative, Arca-Swiss International
    R-Line Technical Cameras, Large Format View Cameras, Tripod Heads D4, D4m, P1, P0, Z1, Z2, C1 Cube.
    http://www.rodklukas.com/arca-swiss 480-755-3364
    Instagram @arcaswissusa Facebook @arcaswissusa
    Thanks 1 Member(s) thanked for this post

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •