Tube lens comparisons for MFT with Mitu 10x
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Tube lens comparisons for MFT with Mitu 10x
I have been meaning to do a systematic test of all my possible tube lenses, and finally I did it. These are the contestants:
100mm Tokina macro lens, newest model f/2.8
125mm Raynox (DCR-250) reversed
135mm Vivitar/Komine f/2.8
135mm Carl Zeiss Jena f/3.5
135mm Pentax f/3.5
150mm Schneider-Kreuznach Apo-Componon HM enlarger lens
170mm DCR-5320Pro Raynox 72mm diameter close-up lens set
200mm Nikkor-Q f/4
208mm Raynox (DCR-150) not reversed
210mm Sironar f/5.6 view camera lens Edit: accidentally used in reverse!
This set represents a wide variety of lens types and quality, so I did not know what to expect. In particular I had high hopes for the very expensive Apo-Componon HM and the Sironar view camera lens. I was sure that some of these would do better than my first, and cheapest, purchases, the 135mm Vivitar/Komine (just $60) and the old 200mm Nikkor-Q (also cheap though I forget how much).
I was wrong. The exotic solutions hardly differ from the cheapest, and are sometimes worse. $1000+ down the drain.
To be fair, there are some caveats here. The Pentax and Zeiss 135mm lenses were in M42 mounts and the adapter to MFT didn't come yet, so I used an M42-Nikon F adapter on my Nikon-MFT adapter. This does not permit these lenses to focus at infinity, so they were focusing slightly closer. I don't know if this mattered; I should do another test to check that. Or redo the test when my adapter comes.
Another caveat: neither the Pentax nor Zeiss lenses were the "best" versions ever made; they were the versions that were available to me on eBay last time I was in the US.
Yet another caveat: when I got the Apo-Componon, I noticed that a lens wrench had removed some of the black paint in the slots on mounting ring of the front element. This means somebody once took it apart, and that can't be good.
Method: The objective was always placed as close as possible to the lens. This sometimes involved complex arrangements of step-up and step-down rings to get below the lens flange. The enlarger lens had no front threads so the whole front element was surrounded by step-up and step-down rings that were attached to the shutter [Edit: I just mean the mounting threads, there being no shutter], which did have threads.
My subject was the outer surface of the hindwing of a Vanessa altissima or related butterfly, a "universal" butterfly that you all can go out and catch for yourselves, so you can compare your tube lenses using the same subject. The wing is glued to a slide. Photos were taken with an Oly PEN F and a Stackshot, step size 4um.
I used FastRawViewer to make sure every exposure was using the entire dynamic range available. I used high speed flash to control for camera and environmental vibrations. I used heavy physical diffusion to reduce heat-induced butterfly scale movement.
I stacked from the RAW files. I used exactly the same Photoshop batch action for all the conversions: Sharpening at ACR default, 25%; exposure slider adjusted down, whites adjusted down. Nothing else touched. Final stacked results were adjusted again slightly with the levels slider to equalize them so they all had the same exposure.
Here is the full image for the SK 150:
The cheapest solutions were quite good, hardly distinguishable from the expensive ones. Here is the SK 150mm versus Vivitar/Komine 135mm, compared in the center of their frames at 100%. In spite of the greater m of the SK and its 7x higher cost and its fantastic reputation, its resolution is only very slightly greater on the subject.
SK 150mm (left) vs Vivitar/Komine 135mm (right).
Among the four 125mm -135mm options, the Raynox DCR-250 and the Vivitar/Komine 135 were best. The Raynox was slightly contrastier than the camera lenses, but had some very slight red fringing on highlights in the center, while the Viv 135 had no color fringing in the center. Stronger (and redder) fringing on CZJ esp in corners, PMax. Edges of the Raynox may be slightly sharper than those of the Vivitar, but with more purple fringing than the Vivitar.
Comparing the Raynox DCR5320 (170mm) to the SK 150mm, the Raynox had more magenta fringing and perhaps very slightly less resolution at the same FOV.
The Raynox DCR-5320 had less contrast and slightly lower resolution in center than the 200mm Nikkor-Q, and shows stronger purple fringing on the corners compared to the Nikkor-Q, which had only slight fringing on the brightest highlights. Nevertheless the corners of the DCR-5320 might have slightly more resolution than those of the Nikkor-Q.
Between the 200mm Nikkor-Q and the 210mm Sironar view camera lens, the Nikkor is contrastier and slightly sharper. In the corners, the Sironar showed more purple fringing. PMax shows less fringing than DMap. Edit: I used the Sironar in reverse accidentally, so this result may improve when used correctly.
Between the 200mm Nikkor Q and the 208mm Raynox DCR-150, the Nikkor outresolves the Raynox in the center but the Raynox seemed slightly more contrasty. On the edges, the Nikkor shows slight blue-purple fringes on the outer edges of bright highlights, but the Raynox shows stronger blue fringing and it is on the inner (more central) edges of the highlights.
So that's my surprising report.
100mm Tokina macro lens, newest model f/2.8
125mm Raynox (DCR-250) reversed
135mm Vivitar/Komine f/2.8
135mm Carl Zeiss Jena f/3.5
135mm Pentax f/3.5
150mm Schneider-Kreuznach Apo-Componon HM enlarger lens
170mm DCR-5320Pro Raynox 72mm diameter close-up lens set
200mm Nikkor-Q f/4
208mm Raynox (DCR-150) not reversed
210mm Sironar f/5.6 view camera lens Edit: accidentally used in reverse!
This set represents a wide variety of lens types and quality, so I did not know what to expect. In particular I had high hopes for the very expensive Apo-Componon HM and the Sironar view camera lens. I was sure that some of these would do better than my first, and cheapest, purchases, the 135mm Vivitar/Komine (just $60) and the old 200mm Nikkor-Q (also cheap though I forget how much).
I was wrong. The exotic solutions hardly differ from the cheapest, and are sometimes worse. $1000+ down the drain.
To be fair, there are some caveats here. The Pentax and Zeiss 135mm lenses were in M42 mounts and the adapter to MFT didn't come yet, so I used an M42-Nikon F adapter on my Nikon-MFT adapter. This does not permit these lenses to focus at infinity, so they were focusing slightly closer. I don't know if this mattered; I should do another test to check that. Or redo the test when my adapter comes.
Another caveat: neither the Pentax nor Zeiss lenses were the "best" versions ever made; they were the versions that were available to me on eBay last time I was in the US.
Yet another caveat: when I got the Apo-Componon, I noticed that a lens wrench had removed some of the black paint in the slots on mounting ring of the front element. This means somebody once took it apart, and that can't be good.
Method: The objective was always placed as close as possible to the lens. This sometimes involved complex arrangements of step-up and step-down rings to get below the lens flange. The enlarger lens had no front threads so the whole front element was surrounded by step-up and step-down rings that were attached to the shutter [Edit: I just mean the mounting threads, there being no shutter], which did have threads.
My subject was the outer surface of the hindwing of a Vanessa altissima or related butterfly, a "universal" butterfly that you all can go out and catch for yourselves, so you can compare your tube lenses using the same subject. The wing is glued to a slide. Photos were taken with an Oly PEN F and a Stackshot, step size 4um.
I used FastRawViewer to make sure every exposure was using the entire dynamic range available. I used high speed flash to control for camera and environmental vibrations. I used heavy physical diffusion to reduce heat-induced butterfly scale movement.
I stacked from the RAW files. I used exactly the same Photoshop batch action for all the conversions: Sharpening at ACR default, 25%; exposure slider adjusted down, whites adjusted down. Nothing else touched. Final stacked results were adjusted again slightly with the levels slider to equalize them so they all had the same exposure.
Here is the full image for the SK 150:
The cheapest solutions were quite good, hardly distinguishable from the expensive ones. Here is the SK 150mm versus Vivitar/Komine 135mm, compared in the center of their frames at 100%. In spite of the greater m of the SK and its 7x higher cost and its fantastic reputation, its resolution is only very slightly greater on the subject.
SK 150mm (left) vs Vivitar/Komine 135mm (right).
Among the four 125mm -135mm options, the Raynox DCR-250 and the Vivitar/Komine 135 were best. The Raynox was slightly contrastier than the camera lenses, but had some very slight red fringing on highlights in the center, while the Viv 135 had no color fringing in the center. Stronger (and redder) fringing on CZJ esp in corners, PMax. Edges of the Raynox may be slightly sharper than those of the Vivitar, but with more purple fringing than the Vivitar.
Comparing the Raynox DCR5320 (170mm) to the SK 150mm, the Raynox had more magenta fringing and perhaps very slightly less resolution at the same FOV.
The Raynox DCR-5320 had less contrast and slightly lower resolution in center than the 200mm Nikkor-Q, and shows stronger purple fringing on the corners compared to the Nikkor-Q, which had only slight fringing on the brightest highlights. Nevertheless the corners of the DCR-5320 might have slightly more resolution than those of the Nikkor-Q.
Between the 200mm Nikkor-Q and the 210mm Sironar view camera lens, the Nikkor is contrastier and slightly sharper. In the corners, the Sironar showed more purple fringing. PMax shows less fringing than DMap. Edit: I used the Sironar in reverse accidentally, so this result may improve when used correctly.
Between the 200mm Nikkor Q and the 208mm Raynox DCR-150, the Nikkor outresolves the Raynox in the center but the Raynox seemed slightly more contrasty. On the edges, the Nikkor shows slight blue-purple fringes on the outer edges of bright highlights, but the Raynox shows stronger blue fringing and it is on the inner (more central) edges of the highlights.
So that's my surprising report.
Last edited by Lou Jost on Wed Dec 14, 2016 5:45 pm, edited 2 times in total.
Almost forgot, there was one other thing I did in these tests, with the Sironar view camera lens. I used a shifting bellows with that lens, and made a panorama of three overlapping shots, which I then stitched together manually so I could see how well the lens controlled distortions. It did very well. I made no effort to hide the stitched edges, except translations and one rotation of about 1/10 of a degree, and a slight exposure adjustment for one frame. This image is 10000 pixels wide.
It would be easy to put a Fotodiox shift adapter between the bellows and the camera, with its shift axis perpendicular to that of the bellows, for perfect stitching of three rows and three columns of overlapping images (effectively about 2.5 images in each row and column). Because all these images are actually part of a single aerial image, this kind of scanning does not involve any perspective shifts (the lens does not move) and does not benefit from telecentricity. [Edit--the lens does not move for the stitching, but it does move during the stacking, so telecentricity would help there.]
Edited note--corners are not too good, but with a little cropping it would be very nice. Will do more experiments along these lines later...
It would be easy to put a Fotodiox shift adapter between the bellows and the camera, with its shift axis perpendicular to that of the bellows, for perfect stitching of three rows and three columns of overlapping images (effectively about 2.5 images in each row and column). Because all these images are actually part of a single aerial image, this kind of scanning does not involve any perspective shifts (the lens does not move) and does not benefit from telecentricity. [Edit--the lens does not move for the stitching, but it does move during the stacking, so telecentricity would help there.]
Edited note--corners are not too good, but with a little cropping it would be very nice. Will do more experiments along these lines later...
Last edited by Lou Jost on Thu Dec 15, 2016 7:06 am, edited 2 times in total.
Great test Lou! Thanks for doing it.
I started similar tests which approached the same conclusion (cheaper M42 lenses are plenty good enough). Unfortunately, the Samyang 2/135 ended up being a very front-heavy arrangement and may have slightly bent my rail
So, my tests are incomplete and I'm off down a rabbit hole of testing what damage I've done - if any. This is in addition to entering another rabbit hole of flocking inner surfaces of adapters when I saw contrast differences during testing. Surprising how reflective that "black" anodising can be. Sigh!
Anyway - thanks again for publishing your results.
I started similar tests which approached the same conclusion (cheaper M42 lenses are plenty good enough). Unfortunately, the Samyang 2/135 ended up being a very front-heavy arrangement and may have slightly bent my rail
So, my tests are incomplete and I'm off down a rabbit hole of testing what damage I've done - if any. This is in addition to entering another rabbit hole of flocking inner surfaces of adapters when I saw contrast differences during testing. Surprising how reflective that "black" anodising can be. Sigh!
Anyway - thanks again for publishing your results.
Yes, very shiny indeed!! I've flocked most of my adapters with Protostar, it's the closest thing to an optical black hole I've come across. Pure black at just about any angle!!Beatsy wrote:Great test Lou! Thanks for doing it.
I started similar tests which approached the same conclusion (cheaper M42 lenses are plenty good enough). Unfortunately, the Samyang 2/135 ended up being a very front-heavy arrangement and may have slightly bent my rail
So, my tests are incomplete and I'm off down a rabbit hole of testing what damage I've done - if any. This is in addition to entering another rabbit hole of flocking inner surfaces of adapters when I saw contrast differences during testing. Surprising how reflective that "black" anodising can be. Sigh!
Anyway - thanks again for publishing your results.
Sorry about your rail, the Rokinon/Samyang 135mm f2 is quite heavy.
Mike, this does mount directly on my D90 and D5300, though these tests were on MFT using a Nikon F to MFT adapter (and it fit that as well). And yes, it has a 52mm front thread.
There is also a good report on this forum regarding a slightly newer Nikkor 200mm f4 all-black manual lens used as a tube lens. (My Nikkor Q has a silver basal segment.)
There is also a good report on this forum regarding a slightly newer Nikkor 200mm f4 all-black manual lens used as a tube lens. (My Nikkor Q has a silver basal segment.)
Lou,Lou Jost wrote:Mike, this does mount directly on my D90 and D5300, though these tests were on MFT using a Nikon F to MFT adapter (and it fit that as well). And yes, it has a 52mm front thread.
There is also a good report on this forum regarding a slightly newer Nikkor 200mm f4 all-black manual lens used as a tube lens. (My Nikkor Q has a silver basal segment.)
Would that be the Nikkor Non-Al 200mm f4? Do you recall the review, I looked but couldn't find it.
Thanks,
Mike