Tube lens tests on D800E full frame

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

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Greenfields
Posts: 116
Joined: Thu Nov 25, 2010 10:54 am
Location: Nottinghamshire, England

Post by Greenfields »

Hi Enrico,

The author means: "A telecentric lens stop is created by placement of a lens stop at the front focal point."

That appears in a caption to two of several illustrations of tube lens design.

I assume that "lens stop" means the exit pupil of the objective or, if there is one, a field stop behind the objective.

The book includes notes and ray fan plots of four tube lens designs: including two achromatic doublets and the example used by Nikon in their patents. Of course, we will never know whether Nikon's example represents the MXA20696. It appears to fit - but that's not conclusive.

Henry
Feel free to edit my images.

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Greenfields wrote:The author means: "A telecentric lens stop is created by placement of a lens stop at the front focal point."
There are still a couple of problems with that.

First, the separation between tube lens and objective is "100-200" mm in the recommended configuration, but the front focal point is out around 240 mm.

Second, the clear diameter at the rear of an MXA20696 is only 25 mm. At infinity focus, that location is slightly over 150 mm away from the image plane, so at f/20 (10X NA 0.25) the exit cone will be over 7.5 mm wide at that point. In a perfectly telecentric configuration, that would leave a non-darkening field less than 18 mm wide and a non-black field less than 33 mm wide. This seems a bit narrow even for Nikon's intended use with microscope eyepieces.

I'm thinking the word "essential" is a bit too strong in Seward's description.

--Rik

Greenfields
Posts: 116
Joined: Thu Nov 25, 2010 10:54 am
Location: Nottinghamshire, England

Post by Greenfields »

Maybe Seward's comment refers specifically to "The doublet pair based on the Abbe number" which is the design he is discussing.

He goes on to talk about another design of doublet pair based on "real glass" which looks similar but has better performance. This is how the description continues.

I obviously don't want to go beyond "fair use" by including too much so I have not included the illustrations.

"The doublet pair based on the Abbe number (shown in Fig. 13.4) may be considered an achromat for the d and g lines. Figure 13.5 displays the axial color of the doublet pair. The “U shape” indicates complete correction for two wavelengths. However, the plot does not display the “S shape” of an apochromat, where a vertical line indicates three corrected wavelengths. The lens is certainly an achromat; however, it is not apochromatic. The doublet pair might be called a semiapochromat because it is substantially corrected for three wavelengths.

Implementation of real glass from the 1990s yields improved performance of the doublet pair as shown in Fig. 13.6. The chromatic correction is greatly improved by application of known glass types to the prescription. The g line displays less lateral color. However, there is a cost to correct lateral color. The lanthanum flint LaFN7 and the short-flint special KzSFN4 are highly reactive with the atmosphere and cleaning agents. The schwer krone SK10, the barium schwer flint BaSF4, and the short-flint special KzSFN4 are obsolete as of 2010.

Implementation of this design requires new glass types along with new values for radius and thickness. See Chapters 17 through 19 for additional information on glass types and correction of lateral color.
The doublet pair with real glass in Fig. 13.6 should be considered an achromat. Figure 13.7 displays the axial color of the doublet pair with real glass. The lens with real glass is barely an achromat. However, the doublet pair may be called a semiapochromat because it is substantially corrected for three wavelengths."


Henry
Feel free to edit my images.

papilio
Posts: 243
Joined: Tue Jul 16, 2013 11:53 am
Location: St. Paul, MN

Post by papilio »

Thank you for this post Rik! So glad I found it yesterday, especially since it's showing the performance of the Raynox/Mitty combo as well as the camera I'll soon be using. Amazing how the Raynox seems always to out-perform the dedicated tube lenses.

At this time my image sharpness is still limited by my own technique, as well as (except for the Mitutoyo) my lenses, hopefully in time I'll be able to realize more of the D800E's potential. What really grabbed me in the short time I was able to play with my friend's D800E was the overall image quality, wonderfully deep and rich.

And as soon as the new camera arrives first thing I'll do is turn that Raynox around!


p.s. In looking back I see that you were comparing the camera's .jpgs. I believe that I've read in at least two articles that the potential sharpness of the D800E can be realized in the RAW/converted images but not so much in the .jpgs -- have you found this to be the case?
-- Michael


My flickr

Nikon D800E, Sigma 150mmOS Apo, Canon MP-E65, Mitutoyo Plan Apo 10X/NA0.28

ChrisR
Site Admin
Posts: 8662
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Amazing how the Raynox seems always to out-perform the dedicated tube lenses.
Errm, well, to refer once again to one of Rik's previous writings:
http://www.photomacrography.net/forum/v ... 364#126364
(And there was another I haven't found)
though the recent full frame tests look good, for APS sensors the dedicated TLs had the edge.

papilio
Posts: 243
Joined: Tue Jul 16, 2013 11:53 am
Location: St. Paul, MN

Post by papilio »

Oops, apologies ... I saw the results on this post, and remember one or two others which Rik linked me to as well which gave me the same impression about the Raynox. Sorry for speaking too soon. In Riks initial post here, what I seemed to take from the test shots was that the Raynox had better edge and corner performance but fell short in the center. Would this be accurate or am I mixed up?

I am pleased though that it works as well as it does, and surprised.
-- Michael


My flickr

Nikon D800E, Sigma 150mmOS Apo, Canon MP-E65, Mitutoyo Plan Apo 10X/NA0.28

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

ChrisR wrote:Errm, well, to refer once again to one of Rik's previous writings:
http://www.photomacrography.net/forum/v ... 364#126364
(And there was another I haven't found)
though the recent full frame tests look good, for APS sensors the dedicated TLs had the edge.
Oh, it's even murkier than that. In the tests reported there (April 2013), I actually did take a shot at evaluating full frame performance, by using an APS-C sensor shifted off-axis so as to cover one quadrant of a full-frame sensor. Even so, the conclusion came out -- based mostly on lateral CA -- that the official tube lenses had a slight edge over the Raynox.

But those tests were of the tube lens in isolation, imaging resolution targets at 20+ feet. That might or might not reflect how well the lenses would work when used behind an objective. That aspect, combined with the difficulty of working with a shifted sensor, left multiple "soft spots" regarding what the tests might mean with regard to actual practice.

Now 14 months later, I've taken here a quite different approach. No theoretical justifications, no extrapolations, not even any generalizations. I just put the tube lenses between the best objective and the best camera I could buy, shot the tests, and showed the images. There are still lots of soft spots here, most notably regarding lateral CA if you happen to be shooting anything other than a Nikon D800E in JPEG mode. But at least they're different soft spots.

With luck, the additional experience results in better recommendations, which in the end is what I care about.

--Rik

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

papilio wrote:I believe that I've read in at least two articles that the potential sharpness of the D800E can be realized in the RAW/converted images but not so much in the .jpgs -- have you found this to be the case?
The question is simple, the answer is not.

As part of familiarizing myself with the camera, I took some test shots of a shingled roof across the street, so I could compare in-camera JPEG against PS CS5 & Lightroom 4 development of the corresponding NEFs. The CS5 and Lightroom results were visually identical, but obviously different from the in-camera JPEGs, with the converted NEFs giving to me an impression of finer detail. But the case was not ironclad; when I ran the images past a friend who has looked at a lot of Nikon images, he pointed out that every place he looked, the same details were visible in both images although they definitely looked different. It's like they had been sharpened with different filters. So I took a harder look at that question, shooting instead some USAF resolution targets. That nailed down that yes indeed there was significant difference in the sharpening (for example some pretty serious halos around the bars), and in addition the converted-from-raw image shows about 1 finer set of bars.

Based on those results, I would definitely choose to work from the NEF if I wanted the very highest quality results and I was willing to pay the additional cost in storage and conversion.

That said, I nonetheless decided to use JPEG for the tube lens tests because a) the images weren't going to challenge the D800E's resolution even in JPEG mode, b) I did care about the additional cost in storage and conversion, and c) I figured that what really mattered was just to get uniform treatment for all the lenses. That worked fine -- I think.

But here's where things get a little confusing.

After I realized that lateral CA had been removed from the JPEGs, I went back and shot one more stack as NEF + JPEG to get a comparison for the tube lens tests. I ran the JPEG's straight through ZS and ran the NEFs through ZS using the Lightroom plugin, making no attempt to remove CA during the raw conversion. As expected, there was quite a lot more CA in the second case. But not as expected, the result from JPEG also showed a lot better detail, especially in the corners. This result persisted after I used Photoshop to remove CA in the result, and it further persisted even when I had Lightroom remove the CA during its raw conversion, and re-stacked from there.

Bottom line is that as I write this, I don't know exactly what's going on. Clearly I am suffering from NTCID (New Technology Cognitive Insufficiency Disorder). I think the next step is to install Capture NX2 and see if I can use that to reproduce the same nice result that I got from the D800E's internal JPEG processing, then try tuning it to do even better.

--Rik

Peter M. Macdonald
Posts: 187
Joined: Tue Jan 20, 2009 2:59 pm
Location: Berwickshire, Scotland

Post by Peter M. Macdonald »

Rik,

Is the additional detail in the JPEG tests over the raw stack coming from the different sharpening which the camera is applying when it forms the JPEGs?

Peter

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Peter M. Macdonald wrote:Is the additional detail in the JPEG tests over the raw stack coming from the different sharpening which the camera is applying when it forms the JPEGs?
That's probably one contributor. I've spent some time this morning poring over the various inputs and outputs. It's clear that the DMap outputs posted above have failed to track some of the fine scale structure, resulting in loss of detail on the front surface of small blobs. The loss is different with different inputs, and that's something I might expect to be altered by sharpening.

Again, I don't think this is an aspect that would alter the lens-versus-lens comparisons, but it's definitely an interesting topic for investigation in its own right.

--Rik

Inseewincesee
Posts: 70
Joined: Wed Jun 04, 2014 1:08 am
Location: Sydney, Australia

Post by Inseewincesee »

As a relatively new D800e owner myself, I've been finding this topic very interesting.

Regarding the differences found with 'In Camera' processing, it rang a bell, and I remembered reading on a way to alter it.

You can subtly change the way the D800/e in camera filters work .

Look from P163 of the owners manual.
Image Enhancement - Picture Controls.
You have control over Sharpening/ Contrast/ Brightness/ Saturation / Hue.

You can then create your own custom preset, save it and name it whatever you like.
And if you find one that you think other people might fine useful, you can then share your presets with other D800/e owners using ViewNX2 software.

If this effects the differences found with the processing of JPEG's, I don't know, but it might be worth at least testing if changes made in these sub menus alter the results you have been getting.

Why do I think this might be something to do with it?

Well the D800/e seems has a underlying JPEG processing happening regardless of menu choice.
If you were to have a single memory card installed, and selected RAW as your preferred capture mode, it will do so, but, viewing the same RAW captured image via the screen on the back of the camera, it is a JPEG reproduction of the RAW data.
Somewhere along the line, there is a overriding filter coming into effect.

Have you checked your cameras system software version?
Nikon released a new one for the D800/e around a month ago.
https://support.nikonusa.com/app/answer ... F-b%3A1.10

As for computer based Chromatic Aberration Filters, you might find this of interest.
C'T Digital Photography magazine [issue 9, P76-82] Reviewed the performance of these software programs and their effectiveness, with both JEPG's and RAW.
ACDSee Pro4 / Capture ONE / Apple Aperture / DxO Optics Pro / Lightroom / Nikon Capture NX2 and RAW Therapeee.

Well worth searching out and reading

I think you are now able to purchase a digital version of this issue directly from their recently updated site.
http://www.ct-digiphoto.com/back_issues ... ue-9-2012/

If having problems, you can also get back issues via Zino

RobertOToole
Posts: 2627
Joined: Thu Jan 17, 2013 9:34 pm
Location: United States
Contact:

Post by RobertOToole »

Inseewincesee wrote:........

Well the D800/e seems has a underlying JPEG processing happening regardless of menu choice.
If you were to have a single memory card installed, and selected RAW as your preferred capture mode, it will do so, but, viewing the same RAW captured image via the screen on the back of the camera, it is a JPEG reproduction of the RAW data.
Somewhere along the line, there is a overriding filter coming into effect.
This is important, most photographers even many professionals believe that when they shoot an image in RAW format the review image is a RAW file. This is not the case, each RAW NEF (and JPEG) file created in camera has jpeg files embedded in it.

In the RAW only (not RAW+JPEG) setting a Nikon camera writes a 160x120 uncompressed thumbnail, a 570x375 JPEG image and a full-size jpeg in addition to the actual RAW file. The small jpeg is used to display the entire image on the LCD, the full size jpeg is used for the magnified view.

AFAIK Canon cameras are similar but they largest embedded jpeg is only medium sized.

BTW there are lots of useful programs to extract the embedded jpegs. I like to use them as proofs sometimes. There is a good free utility out there called
IJFR. Google: Instant Jpeg From Raw
Inseewincesee wrote:........

Have you checked your cameras system software version?
Nikon released a new one for the D800/e around a month ago.
https://support.nikonusa.com/app/answer ... F-b%3A1.10
Yes this is a good update. It supposedly fixed a CLS flash misfire bug after almost 2.5 years after the fact!


Robert

RobertOToole
Posts: 2627
Joined: Thu Jan 17, 2013 9:34 pm
Location: United States
Contact:

Re: Tube lens tests on D800E full frame

Post by RobertOToole »

rjlittlefield wrote:Recently I've been testing several tube lenses for use with full-frame cameras. This post is an initial report of what I've found.
Great info and really appreciate, thank so much for sharing!

Robert

Blame
Posts: 342
Joined: Fri May 14, 2010 11:56 am

Post by Blame »

Congratulations on buying that D800E. I think you will find it delivers more detail than your previous APS-c. Partly because of the additional pixels but I think its superb dynamic range is going to be the important bit. A full frame sensor has the area to store more and that one is the best going.

The more dynamic range the more aggressively sharpening can be set which is what I think you are seeing with JPG.

One shouldn't be surprised at seeing a 3 element raynox lens win. The longer the focal length and the smaller the aperture the fewer elemenst that are needed. I suspect that the dedicated tube lenses are more optimized for field flatness which isn't important for stacking.

rguerra
Posts: 2
Joined: Tue Oct 20, 2015 8:24 pm
Location: New York City

separation between objective and tube lens

Post by rguerra »

Hi, just stumbled across your forum, so apologies for reviving an old thread. I've used microscopes for years and it never occurred to me that it was possible to stick a microscope objective on a full frame camera: very cool!

Image sensors on the kinds of microscopes that the ITL200 and Nikon tube lenses are meant for are rarely larger than 1-2cm across, so I'm surprised they work as well as they do. However, I find the massive amount of longitudinal CA in the images where Nikon objectives are paired with the Raynox lens intriguing.

Objective manufactures are quite secretive about their lenses, but it would seem like Nikon is offloading some of its CA correction onto the tube lens (I believe Zeiss does that as well). Do images taken with a Nikon objective and Nikon tube lens look better? Also, despite being "infinity corrected", the objective/tube lens combination works a bit better when they are spaced like a Keplerian telescope, which would mean that the objective should sit ~200mm away from the tube lens (the back focal plane of these objectives is usually inside the lens itself). Has anyone tried doing that?

Cheers,
Rodrigo

Post Reply Previous topicNext topic