I’m about to move on from my testing to see how well various rangefinder — in my case, Leica M-mount — lenses do on the Sony a7S. Here’s what I’ve found out so far.
- Leica 16-18-21mm f/4 Tri Elmar ASPH — works fine at 16mm; also at 21mm.
- Leica 18mm f/3.4 Super-Elmar ASPH — works fine
- Leica 24mm f/3.8 Elmar ASPH — works fine. A tiny bit of “Italian flag” casting
- Leica 28mm f/2.8 Elmarit ASPH — a bad fit. Really soft corners at wide openings.
- Zeiss 35mm f/2 Biogon ZM — works well
- Leica 50mm f/1.4 Summilux ASPH — not good. Blurry corners at wide apertures.
- Leica 90mm f/2 APO-Summicron — fine. I didn’t do an exhaustive test on this lens since it worked so well on the a7R.
- Leica 135mm f/3.4 APO-Telyt– fine. I didn’t do an exhaustive test on this lens since it worked so well on the a7R.
You must have a duff copy of the 50 Lux because I have zero blurry corners shot wide open.
I use a Voigtlander close focus adapter, shoot in shutter priority with a shutter speed of 1/125 or faster and the image quality is excellent.
Please go back and redo this or get another copy of the lens.
You are posting misleading results that nobody else is seeing…
The lens performs quite well on an M240. That wouldn’t happen if it were a bad copy.
I think the difference is the thickness of the sensor stack. The lens was designed for film, which has the equivalent of a zero-thickness stack.
http://www.lensrentals.com/blog/2014/07/sensor-stack-thickness-part-iii-the-summary
Jim
Jim, I agree with EvilTeddie about Summilux 50mm. But since you know you copy is OK on your M240 then the next thing could be your M-Nex adaptor, did you try to test the oposite border to see if there is a problem with adapter planarity?
Also, you should try two tests, once focusing on the center and one focusing on the edge, sensor stack could exaggerate focus plane curvature and for a smearing test, focusing on the border would be better….
nice site, first time here i’ll be reading your old posts 😉
I’ve looked at all four corners, and tried focusing on the corner. The latter helps a little, but not much.
Maybe I’ll do some tests with an ISO 12233 target to quantify the effects.
Jim
I’ve received my A7S and made a test with the A7, I can confirm your tests are right, there is no difference between the two about smearing 🙁
Yes, the smearing seems to come from the glass in the sensor stack, which is the same across the alpha 7x line. Corner color casts appear to be resolution-sensitive, however, with the a7S being the most resistant to the phenomenon.
Jim
Jim, but you did see better performance with the ZM 35, Leica 18 and 24 on the A7s compared to those on the A7r (how about compared to the M?).
While the coated film has a zero stack depth, ignoring the anti-scratch layer, the emulsion does have depth. So, in a colour film the red layer is usually towards the bottom. The colours are recorded at different depths compared to a sensor where all colours. in a conventional sensor, are recorded at the same depth using the Bayer, or equivalent, filter. Presumably the other limitations of film mask the effect that a high quality lens will be designed to focus all colours in one plane and the red layer will be degraded?
Now the question: Are digital systems set up to focus on the sensor surface, the top of the stack or the top of the actual well. Supplementary question: Will the refractive index of the sensor stack not have more effect than the depth, or at least an interaction?