03-12-2012, 07:59 AM
I read the other thread and I still do not understand the issue, sorry.
You are reporting massive problems on the NEX 7 corners with some lenses, which do not appear on the A700 or NEX 5n. I understand that "massive" problems mean that the lenses are reasonably sharp on the A700 and NEX 5n and noticeably unsharp on the NEX 7. I have not seen any sample pictures, however.
So I have tried a comparison with what I have: The NEX 5 (not n) and some A mount lenses. I do not see much difference. What I see is:
-the 24mm f/2.0 prime has poor corners on both cameras. This appear to be in accordance with the published MTF.
-the 35mm f/1.4 and CZ zoom at 50mm have reasonably sharp corners on both cameras, so the NEX 7 can have reasonably sharp corners on some lenses.
From the other thread, which I am not sure I understand correctly, it seems that the problem only appears on wide angles when wide open or on the 85mm CZ. I checked the published MTFs for all wide angle lenses available in A and E mount and for the 85mm CZ. None of them is going to show perfect corners on a 24 mpix APS-C sensor. According to dpreview, the sensor has a cutoff frequency of about 3400 lines per picture height, that is about 109 line pairs per mm: a very high figure for any lens. Published MTFs do not even go that high, they stop at 30 (Sony) or 40 (Zeiss) and even at these values the corners of all wide angle lenses or of the 85mm CZ show poor values. The A700 has a much lower resolution, about 70 line pairs per mm. The corner unsharpness may be measurable but will not be visible on normal subjects for many of these lenses.
Is it possible that what you are reporting is just due to the fact that the NEX 7 will allow you to measure the lenses more accurately? I understand that you are using imatest. That software has been criticized for extrapolating resolution values beyond the sensor cutoff frequency, could that be an explanation? Imatest would give too optimistic values on the A700 and show the lenses for what they really are on the higher resolution sensor?
You are reporting massive problems on the NEX 7 corners with some lenses, which do not appear on the A700 or NEX 5n. I understand that "massive" problems mean that the lenses are reasonably sharp on the A700 and NEX 5n and noticeably unsharp on the NEX 7. I have not seen any sample pictures, however.
So I have tried a comparison with what I have: The NEX 5 (not n) and some A mount lenses. I do not see much difference. What I see is:
-the 24mm f/2.0 prime has poor corners on both cameras. This appear to be in accordance with the published MTF.
-the 35mm f/1.4 and CZ zoom at 50mm have reasonably sharp corners on both cameras, so the NEX 7 can have reasonably sharp corners on some lenses.
From the other thread, which I am not sure I understand correctly, it seems that the problem only appears on wide angles when wide open or on the 85mm CZ. I checked the published MTFs for all wide angle lenses available in A and E mount and for the 85mm CZ. None of them is going to show perfect corners on a 24 mpix APS-C sensor. According to dpreview, the sensor has a cutoff frequency of about 3400 lines per picture height, that is about 109 line pairs per mm: a very high figure for any lens. Published MTFs do not even go that high, they stop at 30 (Sony) or 40 (Zeiss) and even at these values the corners of all wide angle lenses or of the 85mm CZ show poor values. The A700 has a much lower resolution, about 70 line pairs per mm. The corner unsharpness may be measurable but will not be visible on normal subjects for many of these lenses.
Is it possible that what you are reporting is just due to the fact that the NEX 7 will allow you to measure the lenses more accurately? I understand that you are using imatest. That software has been criticized for extrapolating resolution values beyond the sensor cutoff frequency, could that be an explanation? Imatest would give too optimistic values on the A700 and show the lenses for what they really are on the higher resolution sensor?