9 lenses @ 2.5X *edit: now 10 lenses*

[morfa]

I decided to run a simple comparison between these lenses at a magnification I often like to use in the field: 2.5:1.

It's not a scientific test and I have only included one aperture setting per lens. I've tried to choose the setting so as to get the most out of every lens or lens combination. In some cases I chose it from experience (I know where many of them have their sweet spots) and in other cases I had to try a couple of different settings to decide.

The test subject is the head of a dead hover fly. I know this is not an ideal subject for a standardized lens test. However, I'm not so much interested in corner performance, flatness of field etc since I mainly work in the field on living subjects.

Each test shot is stacked from 25 exposures in Zerene Stacker. Whitebalance was fixed at 4800K. The plastic cup diffuser I used here adds some warmth and that is why I didn't go with 5300K-5600K which is the standard flash light temperature. I have cropped the images from the sides to reduce the dead space around the subject but otherwise the images are as shot.


3100px version
6600px version (12Mb)

Individual test-images (Click links for high-res versions!)

Kiron 105mm f2.8 & Raynox +6 diopter (effective focal length: 65mm)
Nikon Micro-nikkor 105mm f4 & Raynox +6 diopter (effective focal length: 65mm)
Ernst Leitz Milar 65mm f4.5
Carl Zeiss Luminar 63mm f4.5
Nikon Nikkor 50mm f1.8 (used reversed)
Nikon EL-Nikkor 50mm f2.8 N (used reversed)
Wollensak Micro-Raptar 50mm f4.5
Carl Zeiss Jena Mikrotar 45mm f4.5
Ernst Leitz Milar 40mm f4.5

I won't comment much on the performance since I haven't had time to carefully examine and compare these images.

Overall I think they are all quite good and the differences aren't huge. The top quarter/third of the Wollensak Micro-raptar image indicates that there might be a problem with how it is mounted on the bellows - perhaps it wasn't centered properly. I'll look into it.

Setup

If I wanted to test the optical performance of these lenses under ideal conditions there would have been better ways to do it. This is not what I set out to do here though! Instead, I wanted this test to reflect field conditions to at least some degree. Doing the test in the field was pretty much out of the question - it's freezing cold here and everything would have been covered in snow and ice.



Camera body:
Fujifilm S5pro

Focusing, and positioning:
Nikon PB-6 Bellows unit
Newport Linear stage + Starrett micrometer
Manfrotto 410 geared head

Light
Nikon SB-400 flash unit
Aluminum reflector
Short coiled TTL-cord
Manfrotto 819-1 Hydrostatic arm
Custom Brackets CB Mini-RC side bracket
To aid focusing (not relevant to the exposures): 1W LED mounted on small Cullman ballhead attached at the end of the bellows rail

Additional support
Novoflex panorama base
Stabil Macro, wooden tripod
Living room couch


This shows how the lighting was arranged: SB-400 from top right, aluminum reflector from bottom left, plastic cup diffuser around the subject.


The hover fly head was attached to a needle at the center of the plastic cup.

This kind of lighting/diffusion is usually not possible with living subjects and my initial thoughts were to use one of the DIY macro beautydish diffusers I usally work with. However, I soon realized that when changing lenses and having to accomodate for the different working distances it would have been difficult to get the lighting consistent enough with that type of diffusion. I could have mounted the flash completely off camera but decided to use this kind of diffuser instead and it seemed to work reasonably well in terms of consistency.

/John

[morfa]

I've continued my testing and found another combination I like a lot:
Leitz Summar 12cm f4.5 + Raynox lenses

Here is a test shot with the Summar + Raynox +12diopter:


High res. http://farm3.static.flickr.com/2687/427 ... c04e_o.jpg

I've tried to duplicate the conditions from the test above but it's not a perfect match.

A compelling aspect is that with the Summar 12cm, a bellows and a set of Raynox diopters (+6, +12 and +25) I cover the range between infinity focus up to 9:1. There is a drop in sharpness towards the edges at close to infinity focus but otherwise the performance is nice throughout the range (wonderful portrait rendition).

Uncropped 8X sample with the Summar + Raynox 25diopter (30mm effective focal length) and the Summar stopped down to between the "24" and "48" marks:

High res: http://farm5.static.flickr.com/4067/427 ... f891_o.jpg

[rjlittlefield]

John, every time I get a closer look at Raynox lenses I get more impressed.

Can you list the model numbers that you're using, please?

The only +25 that I see in their lineup is the MSN-202. They had a set numbered CM-3500 with +6, +12, and +24, but the Raynox web site says that this was discontinued in Jan 2008.

By the way, I congratulate you on the high quality of this comparison. By no means would I consider this "unscientific"!

--Rik

[morfa]

Yes, I have the CM-3500 set and they are marked 6x,12x and 24x. I've always presumed these number corresponded to diopters and they sort of do. According to their website: 6x = 5.9 diopters, 12x = 11.8 diopters and 24x = 25 diopters :-)

http://www.raynox.co.jp/english/35mmacc/egcm2000.htm

It could be a typo but I think it's more likely they use the same lens as in the MSN-202. I have the MSN-202 as well and as far as I can tell it seems identical to the 24x except for the print around the front element. I guess 6x, 12x, 24x worked better with the marketing department.

/John

John, every time I get a closer look at Raynox lenses I get more impressed.

Can you list the model numbers that you're using, please?

The only +25 that I see in their lineup is the MSN-202. They had a set numbered CM-3500 with +6, +12, and +24, but the Raynox web site says that this was discontinued in Jan 2008.

By the way, I congratulate you on the high quality of this comparison. By no means would I consider this "unscientific"!

--Rik

[Charles Krebs]

John,

That's a lot of work. Thanks.

I downloaded all the higher res files and took a quick look. For me the most remarkable thing is just how little difference there is, especially comparing the ones with and without the Raynox lenses. (But we've avoided the corners of the frame, and that's the place supplementary close-up lenses sometimes have problems).

The 50 El Nikkor looks lower in contrast, as if there was a bit of flare. And the Mikrotar and 50/1.8 show a very slight amount of "color" around small white dots. As you mentioned there's something going on with the Wollensak in the upper part. But all and all, very similar.

The final stacked image is a "real world " test, but I'm curious if you feel the stacking process tended to "equalize" the results. That is, if you look at a similar frame in the source files is there a more noticeable difference?

[ChrisR]

'Tis a lot of work. Fascinating results.
Did you use any sort of lens hood, especially on the reversed lenses?
Sorry I'm to lazy tolook your camera up - how big is the sensor in one of those? This affects "edge" of course for other cameras.
I had to smile at how well the 50/1.8 standard lens came out against the more exotic competition. Is the 120 Summar specifically a macro lens?

I suppose with hindsight it would have been good to include one of the first 9 with the tenth lens, so the conditions would be the same, for comparison.

Perhaps most of these lenses are beating the sensor's resolution, so they tend to look similar.

[morfa]

The 50 El Nikkor looks lower in contrast, as if there was a bit of flare.

Yes, given my previous experiences it surprised me to see the low contrast from the EL-Nikkor. It was probably a flare issue. I had decided to use all lenses without shades. Susceptibility to flare is a pretty important issue for me when I use them with natural light in the field. Of course I need to do a lot of tests with different ligthing setups before any conclusions can be drawn. But I'd say the EL-nikkor failed this particular one.

Did you use any sort of lens hood, especially on the reversed lenses?

No shades. Unfair perhaps but it was a conscious descision. All things equal, it's better not to need a shade since working distance can be critical in the field. I intend to make more tests though, with and without shades.

The sensor of my camera is 23mm wide – I should've mentinoned that!

Is the 120 Summar specifically a macro lens?

Yes, optimized for 0.2X - 4X according to macrolenses.de: http://www.macrolenses.de/ml_detail.php?ObjektiveNr=60

I suppose with hindsight it would have been good to include one of the first 9 with the tenth lens

I'll keep it in mind!

And I think you're right - most of these lenses probably do outresolve the sensor at this magnification – it's only 6MP effective pixels.