Sensor differences affect choice of tube lens

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Lou Jost
Posts: 5945
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Sensor differences affect choice of tube lens

Post by Lou Jost »

As I mentioned here ( http://www.photomacrography.net/forum/v ... highlight= ) I have been testing 200mm tube lenses (and nearby focal lengths). I am using a 20Mp Micro-Four-Thirds (MFT) sensor on a shift adapter, so that I am actually capturing the whole lower right quarter of a full-frame image. I thought this would allow all of us to use the results of my tests, regardless of the format of a reader's chosen camera.

This assumes that that lens defects are independent of the sensor. However, when I tested my favorite tube lenses (200mm f/4 Nikkor-Q, 180mm f/2.8 ED Nikkor, 200mm f/4 AI Nikkor) on this shifted MFT sensor, they all showed strong cyan/red fringing (cyan on the outside edges, red on the inside edges) near the corners, which I had not noticed when using these on my Nikon cameras.

I was using a new 7.5x Mitutoyo objective, so I thought maybe the objective was bad. I repeated the tests using my known-good 10x Mitutoyo, which I had used extensively, and got the same fringing. So it wasn't the objective.

I then went back and took the same picture, with exactly the same subject and lighting (everything was still set up from the previous test) with my Nikon D5300 APS camera, whose pixels are roughly the same size as those of my MFT sensor. I then compared a subject feature that was the same distance from the aerial image center in both the MFT and APS pictures, near the corner of the APS frame. At 200% magnification, the MFT picture (left below) had strong cyan fringes while the APS picture (right below) did not! Some defect is indeed visible on the APS sensor, but it is not conspicuous.

Image

Perhaps the extra thickness of the MFT sensor pack (4.2mm for MFT rather than about 2mm for Nikon) might play a role in this? The light was reaching part of the sensor at a much bigger angle than in normal photography, because the sensor was shifted so far off center (so that its far corner would be in the same place as the corner of a full-frame sensor). These oblique rays would "see" a thicker filter pack:

Image

Has anyone else noticed this?

The curious thing was that a few of my tube lenses showed less of this effect. The Raynox DCM150 showed mild cyan fringing. The DCM 5320 showed yellow fringing instead of cyan, and purple instead of red, and virtually no fringing on the APS sensor. The Sinaron S 210mm large-format lens showed a similar but more diffuse yellow/purple fringing on the MFT sensor. The Nikkor 80-200 2.8 ED showed only slight cyan fringing, which was less conspicuous and greener on the APS sensor.

[Edit: The Sinaron on the APS sensor shows almost no fringing at all.]

I am puzzled.

rjlittlefield
Site Admin
Posts: 23562
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

At first blush, this sounds like the sensor cover acting like a prism to split oblique bundles into rainbows.

The effect should vary depending on the extent to which the objective & tube lens form a combo that is more or less telecentric on the sensor side.

--Rik

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Does anyone know if all 4/3 sensors have similarly thick glass?
Chris R

Lou Jost
Posts: 5945
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Yes Chris, it is part of the Four-thirds standard. All lenses intended for Four-thirds cameras take this sensor thickness into account, and the Micro Four Thirds lenses (whose flange distances are much shorter than the original Four Thirds standard) are often advertised as telecentric on the image side. This suggests Rik is on the right track with his explanation of the problem.

Of course there are no Micro Four Thirds lenses designed to cover an APS or FF sensor, so they can't be used as tube lenses except on a Four Thirds sensor (where they can do rather well, though my Oly 40-200 zoom showed some minor cyan fringing on the brightest black and white edges).

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I think I remember reading that "Digitar" lenses were made with larger exit pupils, more telecentric on the sensor side, to work better with glass-covered Bayer filtered digital sensors. More of an issue with short lenses I daresay. But if you have a few Apo Digitars lying about...
Chris R

Chris S.
Site Admin
Posts: 4042
Joined: Sun Apr 05, 2009 9:55 pm
Location: Ohio, USA

Post by Chris S. »

Lou, my bet is that you've already eliminated the following as a possibility, but just in case:

In my experience, Nikon is particularly good at canceling out some forms of chromatic aberration in camera (for jpegs) and in raw conversion software (for NEF files). Even some Canon jpegs I sent through a Nikon raw converter had their CA very convincingly removed, although the camera and lens had nothing to do with Nikon. Last time I looked, Nikon removed CA by default for both jpegs (in camera) and raw files (in conversion), so that one could be removing CA in processing without realizing it.

Thus the difference might not be sensor size, but in the way the camera or conversion software handles the files. But again, you've probably taken steps to avoid this.

Cheers,

--Chris S.

Lou Jost
Posts: 5945
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Thanks for that suggestion, Chris. That's a very good point. I had always thought that Nikon needed lens information in order to be able to do that, so I did not consider this, since most of these lenses have no electronic communication with the camera.

The lens most likely to "cheat" would be my Nikon 80-200 ED, which does have electronic contacts. But there is no lens metadata in the files, as if the lens was not recognized (it's old). Nevertheless I just checked a jpg versus NEF (opened in Adobe Camera Raw with lens aberration correction turned off) from one of the images in the stack, and the two images are nearly identical. So I don't think that is what is happening.

However, I do see that checking Lens Corrections/Profile/Remove Chromatic Aberration in ACR does clean up the aberrations a bit with this lens. Maybe that would improve some of the others...

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I had always thought that Nikon needed lens information in order to be able to do that,
Not sure, but I think it just does it "by eye"!
Chris R

Lou Jost
Posts: 5945
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Spurred by your comment, Chris, I tried making a stack from RAW files with the Chromatic Abberations Correction checkbox checked in ACR's Lens Corrections/Profiles. This almost completely eliminated the CA even though the lens (in this case an old all-manual 200mm AI lens) does not communicate with the camera and there is no profile involved. I did NOT use the manual CA corrections. I am pleasantly surprised. Thanks Chris for the suggestion!

Stack near the extreme corner of FF, made from uncorrected jpgs on left, stack made from corrected RAW files on right:

Image

Now which test results to present? On the one hand, it is nice to know what the lens is really doing, withpout software hiding its defects, but on the other hand some aberrations will be more easily correctable than others, and that is important to determine, so that the corrected output is also useful to see. I guess present them both?

rjlittlefield
Site Admin
Posts: 23562
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Lou Jost wrote:I guess present them both?
Sounds that way to me.

--Rik

Post Reply Previous topicNext topic