p.6 #2 · Sony A7 - first impressions and some tests
Many thanks for posting your test photos, Phillip. Your real-world cat photos come closest to how I'd use the camera; how responsive do you find the viewfinder and shutter, and do you feel the focussing accuracy is sufficient for the kitty?
p.6 #4 · Sony A7 - first impressions and some tests
Sorry for being late, but without knowing the chart testing methodology, I have an hypothesis about what can be happening.
It can be that simple as the corners are not in focus.
If the focusing distance is short the distance difference between the center and the corners of the chart can be big enough for the corners to be OOF, even more with that high magnification.
The focusing plane isn't actually a plane but a sphere around us. The chart is perfectly flat.
If this would be correct, the lenses shown as better would be those that have field curvature towards infinity.
To prove if it was the case it would be that simple as repeating the test with exactly the same methodology with just one lens (e.g. the FD 35 f2), but instead of focusing on the center focusing on the corner.
p.6 #5 · Sony A7 - first impressions and some tests
There was some talk about high resolution sensors and old lenses some pages before.
Remember that this is the A7, 24MPx, almost the same as the Canon 5d mk2 which has been successfully adapting manual lenses for a long time. The equivalent pixel density is 16Mpx on APS-C and 12Mpx on 4/3, nothing fancy.
p.6 #6 · Sony A7 - first impressions and some tests
rafa1981 wrote:
To prove if it was the case it would be that simple as repeating the test with exactly the same methodology with just one lens (e.g. the FD 35 f2), but instead of focusing on the center focusing on the corner.
Or print out a concave test pattern :-P
Now I'm wondering if the 35-70 has field curvature on not. Phillip, any real-world shots with that yet (and thanks so much for the examples already!)
p.6 #7 · Sony A7 - first impressions and some tests
AmI wrong in wondering whether out-of-focus is the same as smearing? IN my experience, among others with NEX 5, which did smear corners with quite a few lenses, never did the result look like those of a lens with field curvature and misfocused corners. So why am I reading so many comments about FC and mis-focused corners? Am I missing something?
p.6 #8 · Sony A7 - first impressions and some tests
isn't APS-C 16mpx equivalent to ~37mpx on FF? I think the A7 is equivalent to a ~10.5mpx APS-C cam. Even the A7R is slightly lower pixel density than the NEX-6
p.6 #9 · Sony A7 - first impressions and some tests
Spot on Derek. If I need to end up using larger and larger lenses, it kind of defeats the purpose of such a camera (for my needs anyway). I need a smallish kit with incredible lens performance - just what my Leica gives me....
There is much to that logic, and while I want the Sony A7r to work with all my m lenses, the M240 may remain the most elegant solution for someone who uses the 20-135 range, and mostly 35-50 within that.
p.6 #10 · Sony A7 - first impressions and some tests
Yes. You divide pixel count by 2.25 (1.5x1.5) to go from a FF to 1.5x crop
I hadnt thought about this before but does this imply, absent the removal of the AA filter, that the A7R will have lower resolving power over the frame than the 24Mp Nex7 ?
morpheus2891 wrote:
isn't APS-C 16mpx equivalent to ~37mpx on FF? I think the A7 is equivalent to a ~10.5mpx APS-C cam. Even the A7R is slightly lower pixel density than the NEX-6
p.6 #11 · Sony A7 - first impressions and some tests
philber wrote:
AmI wrong in wondering whether out-of-focus is the same as smearing? IN my experience, among others with NEX 5, which did smear corners with quite a few lenses, never did the result look like those of a lens with field curvature and misfocused corners. So why am I reading so many comments about FC and mis-focused corners? Am I missing something?
To be honest, it doesn't really matter, whether the corner softness is caused by smearing or glass cover induced exaggerated field curvature, as the end result still remains soft corners. My thinking is that smearing and exaggerated FC are both caused by the glass cover but are two separate phenomenons. Smearing is caused by coma, exaggerated FC is caused by the shifting of the focusing plane as the light rays enter the glass at a steep angle. The reason why you didn't see much FC on the Nex 5 is because you are using the sweet spot of the lens that is normally quite flat field, which leaves the other aberration to be seen. Just my educated guess as I have not used the Nex 5 with alt glass.
p.6 #12 · Sony A7 - first impressions and some tests
wolfloid wrote:
There is much to that logic, and while I want the Sony A7r to work with all my m lenses, the M240 may remain the most elegant solution for someone who uses the 20-135 range, and mostly 35-50 within that.
I always said that Sony would make no effort to make its camera work with third-party lenses. If they did, it would be coincidental, and a bonus. The results from the FE 35mm indicate that the most likely outcome is that the best lenses for A7R will be specially designed, principally by Zeiss, but may be others as well (Sony, Sigma).
p.6 #13 · Sony A7 - first impressions and some tests
philber wrote:
AmI wrong in wondering whether out-of-focus is the same as smearing? IN my experience, among others with NEX 5, which did smear corners with quite a few lenses, never did the result look like those of a lens with field curvature and misfocused corners. So why am I reading so many comments about FC and mis-focused corners? Am I missing something?
You are reading about smeared corners because:
1) The new e mount results in the sensor being very close to the lens which results in light coming in at a exteme angle (not direct) when using wide rangefinder lens. This has worked with film in range finders because the film is less sensitive to this issue. Leica has solved [hidden] this problem with a software adjustment. But now sony has the problem. It results in smeared edges when using rangefinder (small) lens at <50mm lens.
2) However it has got people pixel peeping at the edges and what they are finding is those lens that they thought were great (because they were not looking in the edges) are not so geat.
So a real problem for low mm rangefinder lens has got people noticing that even longer lens that are thought to be good may not be so good in the edges.
The discussion should be good for lens in the future (attention will cause it to be fixed in newer lens) but not so good (people are not going to buy them as much) for historic lens which have weak edges.
I suspect there may be other issues as well. All the adapters have tolerances and so do lens and the sum of two tollerances can either be good (cancelling) or bad (multiplying) and this may be different side to side. So some of it could be tolerance effects related to adapters, which really gets noticed with a 36mpx camera.
p.6 #14 · Sony A7 - first impressions and some tests
Scott Stoness wrote:
You are reading about smeared corners because:
1) The new e mount results in the sensor being very close to the lens which results in light coming in at a exteme angle (not direct) when using wide rangefinder lens. This has worked with film in range finders because the film is less sensitive to this issue. Leica has solved [hidden] this problem with a software adjustment. But now sony has the problem. It results in smeared edges when using rangefinder (small) lens at <50mm lens.
2) However it has got people pixel peeping at the edges and what they are finding is those lens that they thought were great (because they were not looking in the edges) are not so geat.
So a real problem for low mm rangefinder lens has got people noticing that even longer lens that are thought to be good may not be so good in the edges.
The discussion should be good for lens in the future (attention will cause it to be fixed in newer lens) but not so good (people are not going to buy them as much) for historic lens which have weak edges....Show more →
Can you please explain how did Leica solve/hide the smearing problem with software? Thanks.
p.6 #15 · Sony A7 - first impressions and some tests
philber wrote:
I always said that Sony would make no effort to make its camera work with third-party lenses. If they did, it would be coincidental, and a bonus. The results from the FE 35mm indicate that the most likely outcome is that the best lenses for A7R will be specially designed, principally by Zeiss, but may be others as well (Sony, Sigma).
I certainly agree with you, I guess where i am surprised, is that the requirement for rf wides and sony lenses do not align. It seems as if the ability to create lenses that are highly compact and sit close to the sensor would be desirable for Sony. If they are forced to create more telecentric lenses, then the size advantage starts going away. The whole a7 push seems to be about compactness, no?
Maybe if not Sony, then someone else (other than Leica) will take up that flag and wave it. One can hope.
p.6 #16 · Sony A7 - first impressions and some tests
edwardkaraa wrote:
Can you please explain how did Leica solve/hide the smearing problem with software?
it is my understanding that they adjusted , in camera, the raw file to remove the vignetting and refocus the different light colors so they aligned digitally. [If you use the high end Leica, with a newer lens the camera recognizes which lens and adjusts it in camera. Or you can just set it.] However any such adjustment degrades the image on the edges, as compared to perfect optics. This is why I say [hidden] because they did not fix the optical problem, they just did the equivalent of vignetting and ca corrrection in photoshop.
In addition both Leica and Sony did some adjustments to hardware to anticipate bent light. But the problem is that if they do too much of this it causes distortion to lens that don't need the correction, so they have to have a compromise. The real question for me is whether the hardware fixes degrade the edges of good lens that have direct light. Because I don't want the A7r for lightness but for the dynamic range and resolution on my canon mount zeiss and L lens that don't have the ray angle problem because they are longer.
p.6 #18 · Sony A7 - first impressions and some tests
Jochenb wrote:
I'm not sure if this irony or serious? Nobody's forcing you to buy one you know...
What do you think? Would I have traded my D800 and X100s for an 7R if I really believed it was crap. I was making a sarcastic point about the content of the thread.
p.6 #19 · Sony A7 - first impressions and some tests
edwardkaraa wrote:
The reason why you didn't see much FC on the Nex 5 is because you are using the sweet spot of the lens that is normally quite flat field, which leaves the other aberration to be seen. Just my educated guess as I have not used the Nex 5 with alt glass.
For the same focal length, field curvature is not significant compared to the fact that for having the whole chart in the cropped frame one has to focus farther:
-The relation between "lens to the center of the chart" and "lens to corner of the chart" is smaller the farther the focusing distance is.
-The absolute DOF (in length units) is bigger the farther the focusing distance is.
That effect alone is in my opinion more relevant than field curvature. "APS has more DOF" (note the quotes).
Which situation can require sharp corners in a near focused pic at f2 when the whole background and foreground is blurred?
In my opinion in real world pics sharp corners are needed the wider the angle, the farther the distance and the smaller the aperture.
p.6 #20 · Sony A7 - first impressions and some tests
morpheus2891 wrote:
isn't APS-C 16mpx equivalent to ~37mpx on FF? I think the A7 is equivalent to a ~10.5mpx APS-C cam. Even the A7R is slightly lower pixel density than the NEX-6
My fault, but even more relevant to the fact that a 24Mpx FF sensor is not that demanding, maybe the opposite.
Thanks for all the samples, Phillip. 
