Home
SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in

RF wide angle lens performance on Sony A7-series
View previous topic :: View next topic  


PostPosted: Wed Mar 06, 2019 8:48 am    Post subject: RF wide angle lens performance on Sony A7-series Reply with quote

I recently had a fierce discussion in another forum where I stated that I'm assuming that the new type BSI-CMOS Sensor performs better with some RF wide angle lenses (which are already well known for their restrictions) also in term of corner smearing. I've already seen really bad examples from A7-models with the old style CMOS sensors and compared it with the results from my RF lenses on my A7R II camera with the newer 42MP BSI-CMOS sensor and found them to be not that bad from my camera.

Some folks stated that all A7 models (irrespective of sensor model) perform all exactly equally bad with such lenses. Their theory behind is that the sensor stack is the only influencing factor and all models must have the same sensor stack as otherwise the E-mount lenses wouldn't perform equally good across the whole E-mount platform.

I've found a "sensor stack database" from Roger Cicala on Lensrentals. Obviously the sensor stack theory is wrong as there are E-mount cameras with sensor stacks from 0.6 to 2 mm listed.

Unfortunately nobody was able to present a direct comparison done with the same lens on different sensor models hence this issue is still open.

As I don't own a "regular" A7 I can't do that myself but I would really be curios to see such a comparison.

Not even after extensive search I've found anything on internet. Is anybody able to help?

Any feedback is welcome.


PostPosted: Wed Mar 06, 2019 9:54 pm    Post subject: Reply with quote

Upgrading to A7R (1st gen FF BSI-CMOS) eliminated most of the veiling flare from variety of my manual lenses, and significantly changed ghosting patterns.
Just as you are, I can't conclusively say, if this is due to shiny wiring being now hidden behind the photo cells, or just the different filter stack.

One has to have both A7 and A7R to make a definitive claim.


PostPosted: Wed Mar 06, 2019 11:03 pm    Post subject: Reply with quote

I have done some simple tests long time ago. I have my hands on A7 II and A7R III, and I also use to have the A7R (first gen). Colour cast is as good as history on most lenses with the newest sensors and corner smearing and vignetting is much better indeed. Colour accuracy and sharpness is also much better Wink

These samples are from the classic 20mm Russar which is the most challenging symmetrical wide angle RF lens I've had.

A7 II


A7R III


A7R III with some simple PP


An almost useless lens bacame usefull!



PostPosted: Thu Mar 07, 2019 12:05 am    Post subject: Reply with quote

Nordentro wrote:
I have done some simple tests long time ago. I have my hands on A7 II and A7R III, and I also use to have the A7R (first gen). Colour cast is as good as history on most lenses with the newest sensors and corner smearing and vignetting is much better indeed. Colour accuracy and sharpness is also much better Wink

These samples are from the classic 20mm Russar which is the most challenging symmetrical wide angle RF lens I've had.


Thank you, Lars.

Well, the color cast/vignetting issue was out of question. However, corner smearing isn't really visible on your examples, not even on the old 24MP sensor. Obviously the Russar isn't that bad in that respect.
Have you got the Jupiter-12 or any older Voigtländer RF wide angles in M39/LTM for testing as well? These lenses show definitely much more smearing than your Russar, particularly at wider apertures at infinity.

Would really be great to see any difference in corner smearing as well.

BTW, I haven't tried my J-12 on my A7R II yet, will do that tomorrow.

Cheers,


PostPosted: Thu Mar 07, 2019 9:25 am    Post subject: Reply with quote

Well, I doubt the J-12 is sharp wide open in the corners on any FF format wide open at infinity Wink Should still be possible to draw some conclusion because smearing tends to look slightly different than just unsharp corners.

I have the J-12 and can do a comparison between A7II and A7R III next week 😀 I also have classic lenses like the Nikkor 2.5cm, Canon 25mm and Angulon 21mm in Leica thread and many 28-35mm RF lenses too.

I believe that corner smearing can be reduced with both better sensors and thin filter stack...


PostPosted: Thu Mar 07, 2019 1:02 pm    Post subject: Reply with quote

Nordentro wrote:
Well, I doubt the J-12 is sharp wide open in the corners on any FF format wide open at infinity Wink Should still be possible to draw some conclusion because smearing tends to look slightly different than just unsharp corners.

I have the J-12 and can do a comparison between A7II and A7R III next week 😀 I also have classic lenses like the Nikkor 2.5cm, Canon 25mm and Angulon 21mm in Leica thread and many 28-35mm RF lenses too.

I believe that corner smearing can be reduced with both better sensors and thin filter stack...


You're right, the Jupiter-12 isn't able to produce sharp corners at infinity, not even at F16. Nevertheless I've made some test pictures today and compared them to the pictures form this test report: https://phillipreeve.net/blog/review-jupiter-12-35mm-2-8/
As this review was done with the 12MP A7S I had to downsize my pictures to 12MP as well. Due to this downsizing the pictures from the 42MP A7R II benefit a little bit in quality; i.e. the perceived sharpness.

This may allow the conclusion that the impression that the 42MP sensor delivers lesser corner smearing in comparison to the old 24MP sensor is based on this downsizing prozess for direct comparison.
I've also made a comparison beween my old A850 (same 24MP sensor like the A7) and my A7R II using the same lens. This also showed that the downsized picture from the A7R II wins slightly in critical pixel peeping mode when comparing both results in 24MP size on screen (32" EIZO UHD/4K).

Nonetheless, a direct comparison of the same smearing lens on both A7 versions may be helpful to be absolutely sure; i.e. to end all the theory or hypothesis.

Therefore if your time would allow I would really appreciate if you could provide such pictures taken with the same lens on both cameras with the focus on corner smearing; preferably in the same size (24MP) and low ISO.
1 picture each would be good enough. Typically with any infinity landscape scenery with some details in the lower corners.

Here is one of my test shots from the Jupiter-12 / A7R II at F16 where even when downsized to 2000px on the long side the corner smearing is still visible, at least on my monitor:



Cheers,


PostPosted: Tue May 14, 2019 10:23 pm    Post subject: Reply with quote

Well, I did a comparison recently AND it seems like the corner smearing is almost identical on both A7R III and A7 II. The R is sharper in center and has less vignetting and almost no colour cast in the corners on any lens, but the smearing is still quite obvious. I wonder if some of the smear comes from field curvature, but there is no doubht about that filter thinkness is the biggest reason for bad corner performance on these wide angle lenses.

The Sony sensors have avarage filter thickness compared to other cameras, it's just that Leica have very thin filters and do a better job in this sense. I'm starting to dream about a kolari mod sensor again Rolling Eyes