Lens testing continued, Zeiss S-Sonnar 62mm f/2.5

dickb · Post by **dickb** » Mon Dec 17, 2018 12:16 pm

I am still in the process of setting up the perfect lens testing system, well perfect for me that is. Currently I use a full frame Canon 5D mk II tethered to my pc, running qDslrdashboard. The camera is mounted on my Nikon Multiphot stand, the subject is on an Olympus CHC focussing block. In an attempt to get perfect alignment between subject and camera, I use an inverted ball head (at the moment an enormous Burzynski head, which may be overkill). I lower the lens mount onto the ball head's surface and lock the movement of the head when alignment appears to be optimal.

I found out the hard way some care must be taken when doing this - after releasing the wrong knob at the Multiphot I now have several small pieces of silicon wafer rather than an unwieldy large one..

At the moment I use two Jansjo LEDs for lighting. Diffused flash is probably going to be better for serious lens comparison, but I'm having some trouble getting my Yongnuo YN622C-TX to work properly with my Canon flashes.

Anyway, I'm having fun testing some of my more unusual lenses such as the Zeiss S-Sonnar 62mm f/2.5 M1:2,12, a copy lens designed for 1:2,12 magnification, as the name implies:

At magnification 2:1:

And a 100% crop from the centre:

dickb · Post by **dickb** » Mon Dec 17, 2018 12:57 pm

The previous photos were with the lens in normal, non-reversed direction, the following are with the lens reversed - much better, especially in the corners of this full frame image. One way this image would benefit from diffused light is that the raised edges of the metal squares reflect differently from the two LEDs in a way that can be confused with CA..

At magnification 2:1:

And a 100% crop from the centre:

And a 100% crop from the edge:

ray_parkhurst · Post by **ray_parkhurst** » Mon Dec 17, 2018 2:42 pm

At some point I will find time to compare my 62SS versus the 95PN. I've had the 62SS for some time (thanks typestar!) but have never fully tested it. It has larger max aperture than the 95PN, but even the 95PN prefers f3.3.

dickb, what did you find was the best nominal aperture at m=2:1, and also what was the working distance?

dickb · Post by **dickb** » Mon Dec 17, 2018 3:29 pm

Ray, the working distance at 2:1 is about 55mm with the lens reversed, and 2 or 3 cm more when not reversed. A refreshingly long distance as the lens I tried before this one was an S-Biogon 40/5.6 with a less than ideal 5mm at 2:1 when used in non-reversed orientation.

I can't say for certain what the best aperture is, I have to test it further. It is about the only lens I can think of in my collection with the apertures marked as effective values rather than theoretical values when focussed at infinity.

ray_parkhurst · Post by **ray_parkhurst** » Mon Dec 17, 2018 3:37 pm

The 74SP (74mm S-Planar) has aperture marked for 1:1. But the 62SS is marked for m=0.47 (1:2.12), so a bit confusing when trying to use at 2:1...Ray

edited to add: the 62SS is also one of a handful of lenses I own that actually open wider than their rated max aperture. This seems to be a positive thing for a few lenses, where they actually perform very well at the max rated aperture, and yet you can open them even further for critical focusing or light gathering. The Lomo 100mm Mikroplanar, and the 75mm Magnagon also have this characteristic.

dickb · Post by **dickb** » Mon Dec 17, 2018 4:15 pm

Indeed, that was the one I wasn't sure about, but you're right, that one is marked as f/8 at 1:1. So I take it Zeiss marked specialised copy lenses for their specific magnification. Other S-(Ortho)-Planar lenses have normal aperture values inscribed.

I just did a quick run at various apertures, if anyone is interested in pixelpeeping have a look here:

Single images at the apertures marked 3.6,5.6,8.0 and 11 respectively.

Bear in mind please:

- this is a 21MP full frame sensor
- I just focussed once and then changed the aperture, these are not the best single files from a large stack
- lighting is done with 2 undiffused LEDs

https://drive.google.com/open?id=1ZRbwJ ... uMtw4A9Xh5

Anyway, the way I see these files is the corners improve going from 3.6 to 8.0. The centre may be best at 5.6, so 1 stop down. YMMV if you use a smaller sensor though..

ray_parkhurst · Post by **ray_parkhurst** » Mon Dec 17, 2018 5:44 pm

dickb wrote: Anyway, the way I see these files is the corners improve going from 3.6 to 8.0. The centre may be best at 5.6, so 1 stop down. YMMV if you use a smaller sensor though..

Thanks, I was just starting to set up to test the lens. Your results are about what I expected based on some early tests. I see what looks like a fair bit of LoCA. Also, needing to go to nominal f3.8 to get the centers sharp, and f5.4 to get the corners, eliminates this from contention with the 95PN. Too much to hope...Ray

dickb · Post by **dickb** » Mon Dec 17, 2018 5:53 pm

The diffusor I intended to use for flash works reasonably well for Leds as well, here is a 100% crop from the centre, again 2:1 reversed 62SS at 1 stop closed (marked 5.6), no CA as far as I can tell

I've uploaded another set of photos with my previous set, now with diffused LEDs, the difference should be obvious..

dickb · Post by **dickb** » Mon Dec 17, 2018 5:58 pm

ray_parkhurst wrote:
dickb wrote: Anyway, the way I see these files is the corners improve going from 3.6 to 8.0. The centre may be best at 5.6, so 1 stop down. YMMV if you use a smaller sensor though..
Thanks, I was just starting to set up to test the lens. Your results are about what I expected based on some early tests. I see what looks like a fair bit of LoCA. Also, needing to go to nominal f3.8 to get the centers sharp, and f5.4 to get the corners, eliminates this from contention with the 95PN. Too much to hope...Ray

Ray, on what camera will you be using this lens? Corners on full frame are a challenge for most lenses. Unfortunately I sold my 95PN long ago, so I can't test them side by side.

ray_parkhurst · Post by **ray_parkhurst** » Mon Dec 17, 2018 7:11 pm

dickb wrote:Ray, on what camera will you be using this lens? Corners on full frame are a challenge for most lenses. Unfortunately I sold my 95PN long ago, so I can't test them side by side.

I've been working on APS-C. I will probably get a FF camera eventually, but the center being sharpest at f3.8 will be limiting even on APS-C.