I shot an infinity cityscape scene, a ~1m bokeh scene with the 35s and 50s (including both new Zeiss lenses) and foliage as a sort of natural 'brick wall' test. All lenses below were used for the infinity test, though not all were used for the foliage test due to time and weather constraints.
I also wanted to shoot this on the M9 as a control reference, but it started to rain as I finished up the infinity test with the a7 and decided just to forge ahead with the bokeh and foliage scenes instead.
Lenses tested:
Voigtlander Ultra Wide-Heliar 12mm f5.6 Aspherical (LTM version which is identical to the current M mount version)
Voigtlander Super Wide-Heliar 15mm f4.5 Aspherical M
Leica Summilux-M 21mm f1.4 ASPH.
Zeiss Biogon T* 21mm f2.8 ZM
Leica Super-Elmar-M 21mm f3.4 ASPH.
Leica Summicron-M 28mm f2 ASPH.
Voigtlander Nokton 35mm f1.2 Aspherical VM II
Canon 35mm f2.0 LTM
Zeiss C Biogon T* 35mm f2.8 ZM
Sony Carl Zeiss Sonnar T* 35mm f2.8 (newly announced with the a7/a7R)
Voigtlander Nokton classic 40mm f1.4
Leica Summilux-M 50mm f1.4 ASPH.
Canon 50mm f1.4 LTM (a lens from the late 50s-early 60s)
Voigtlander Nokton 50mm f1.5 Aspherical LTM (optically identical to the current M mount model, though apparently there may be some coating differences)
Zeiss C Sonnar T* 50mm f1.5 ZM
Zeiss Planar T* 50mm f2 ZM
Sony Carl Zeiss Sonnar T* 55mm f1.8 (newly announced with the a7/a7R)
Sony 28-70mm f3.5–5.6 OSS (newly announced with the a7/a7R)
Leica Summarit-M 90mm f2.5
Only had time to skim through the results... heading out for the evening. Initial impression is there is detectable edge image smearing with all lenses. Some clean up by stopping down, but some, like the ZM21/2.8 and 28 Cron are pretty terrible. Coke bottle terrible in fact.
Hopefully the a7R fares better. As mentioned, it's really the camera I wanted to try, but only the a7 was available here at this time for this very brief period...
Incredible testing effort Ron and with a great write-up summarizing the issues and conclusion. You beat all the professional reviewers who pay lip-service to the idea of using 'Leica lenses' with the cameras but fail to generate any data points.
_julian_ wrote:
Incredible testing effort Ron and with a great write-up. You beat all the professional reviewers who pay lip-service to the idea of using 'Leica lenses' with the cameras but fail to generate any data points.
I only downloaded the 28 Cron and 21 SEL. You can see quite an amazing resolution, but the noise
seems quite bad for ISO400. I guess we really need to wait and see the RAWs.
28 Cron seems to be unusable and it doesn't really clean up closed down. I guess my 28R v2 will be used instead.
21 SEL - a huge improvement between f3.4 and f4. I think I can probably use this one with some kind of
plugin pass applied.
Ah, if Andrew can use his 21 SEL, then my Elmar 24 should be a piece of cake, and it gives me hope for my ZM 18 too...
Ron, many many thanks for your efforts!
Thanks Ron, great effort. I am luke warm about the results with RF lenses and I sure hope the A7r will perform a whole lot better - but I ain't holding my breath. The ZM18 sample I saw gave some hope but it seems like it was false hope.
There are a whole lot of things to like about the Sony A7/r and they have made a refreshing move in the right direction - it is just a pity that it looks like they didn't quite get there with this incarnation with regards to RF lens compatibility. With the right set of primes in the line up I still think it looks like a great camera. The only official sample I have seen from the 24-70 seemed to indicate a decentered lens. I really like that 55mm lens and a few more like that on the wide side and I think I would be quite interested.
philber wrote:
Ah, if Andrew can use his 21 SEL, then my Elmar 24 should be a piece of cake, and it gives me hope for my ZM 18 too...
Ron, many many thanks for your efforts!
21SEL - doesn't seem to be too destroyed/smeared in the corners, the pink vignette is there, but much less
than I remember this lens to look like on NEX-7. I think with modern Camera RAW, it should be fine.
It's really strange how huge the difference is between f3.4 and f4. Then it gets better, but the difference
between f4 and f8 is not that big.
I'm a little worried about my 28R v2, it was quite bad on my NEX-7 and it's similar to 28 Cron M...
I really hope it works, as it's such an amazing lens.
snowboarder wrote:
I'm a little worried about my 28R v2, it was quite bad on my NEX-7 and it's similar to 28 Cron M...
I really hope it works, as it's such an amazing lens.
Issues with an R lens? I would have thought that their much longer registry would put the exit pupil far enough from the sensor to avoid any issues whatsoever.
philber wrote:
Issues with an R lens? I would have thought that their much longer registry would put the exit pupil far enough from the sensor to avoid any issues whatsoever.
Yeah, that's why I sold it right away, maybe it was my NEX-7 copy, but it was clearly not great with that R lens.
Ron, thank you very much!!! Great work and it is wonderful to have your perspective Till now I don't feel there have been shots that have been presented are definitive enough to make a decision.
It would appear for now that the M9/M240 won't be replaced yet, with M lenses. I still have my A7r on order with the 55/1.8, but this may be put on hold shortly There is no point, if the M lenses do not perform on the A7's IMO, as I do not wish to invest in any more lenses for now. Thanks again!
You've created a gold mine Ron, thanks for the effort! Have to say after an initial glance @12-21mm that the CV15 and Biogon 21 look quite promising regarding (lack of) smearing on the F-stops I would use for a FF at infinity which would be F/8 or perhaps 11. Magenta/color cast is evident and quite heavy but that is secondary and fixable. Will alt-tab much more tomorrow :-)
and it doesn't really clean up closed down. I guess my 28R v2 will be used instead.