02-20-2012, 10:32 AM
[quote name='mst' timestamp='1329732137' post='15944']
Allow me to disagree. The D7000 results have been verified by two additional copies of the lens, just to be sure it's not a bad sample.
The D7000 generally shows a tendency to give weaker results towards the borders (which may be a common problem with high density sensors, see the Nex-7 issues Klaus is facing). Compared to the D200 tests, some things have changed, for example we (have to) use a different RAW converter. The higher pixel density means that we measure at a different frequency and different lenses may behave slightly differently here (compared to the lower res D200).
We believe it's a results of the different RAW converters used for different systems. The different demosaic algorithms may lead to slightly different amounts of CAs by emphasizing the issue or by including a small amount of default CA correction.
In any case: from the very first beginning Klaus has always emphasized that the results are not comparable across different test systems. There are simply too many parameters that have an influence on the measured values to allow direct comparisons.
-- Markus
[/quote]
You can't just disagree just to disagree though... I understand that you have verified additional copies of the lens. Apparently you too felt something more was going on.
You skip my point though.... Just look at the Tamron 17-50 f2.8 example I gave. Same D7000, same RAW converter, I suspect?
Yet it gives the results compared to the D200 we would expect, the 14-24mm f2.8 D200/D7000 reviews do not.
So, my point remains, there is something else going on.
The CA issue... It is always an issue. I remember an MFT test, where the amount of CA "measured" in pixels did not at all correspond to the much higher CA clearly evident in the images, which spread over a lot more pixels. I remember a Zeiss prime test from the same lens sample, where the D3X with a bit higher pixel pitch "measured" quite a bit lower CA than the 5D mk II test "measured".
That is not your guys fault, though, I have to presume it is the software test suite you use which gives such "unreliable" results regarding CA.
Anyway.. the CA oddity I pointed out is not very important. What is, in my opinion, is the weird difference between the D200/D7000 14-24mm f2.8 results, and for instance the D200/D7000 tamron 17-50mm f2.8 results.
Or take another lens tested on both, like the Nikon 35mm f1.8 for instance. On D200 and D7000, it shows what we expect. Now compare that to what the 14-24mm results show. Nikon 16-35mm f4 VR, same story.
I stay with my point that the 14-24mm f2.8 is NOT a good example, and that something more is going on there between those two reviews.
Allow me to disagree. The D7000 results have been verified by two additional copies of the lens, just to be sure it's not a bad sample.
The D7000 generally shows a tendency to give weaker results towards the borders (which may be a common problem with high density sensors, see the Nex-7 issues Klaus is facing). Compared to the D200 tests, some things have changed, for example we (have to) use a different RAW converter. The higher pixel density means that we measure at a different frequency and different lenses may behave slightly differently here (compared to the lower res D200).
We believe it's a results of the different RAW converters used for different systems. The different demosaic algorithms may lead to slightly different amounts of CAs by emphasizing the issue or by including a small amount of default CA correction.
In any case: from the very first beginning Klaus has always emphasized that the results are not comparable across different test systems. There are simply too many parameters that have an influence on the measured values to allow direct comparisons.
-- Markus
[/quote]
You can't just disagree just to disagree though... I understand that you have verified additional copies of the lens. Apparently you too felt something more was going on.
You skip my point though.... Just look at the Tamron 17-50 f2.8 example I gave. Same D7000, same RAW converter, I suspect?
Yet it gives the results compared to the D200 we would expect, the 14-24mm f2.8 D200/D7000 reviews do not.
So, my point remains, there is something else going on.
The CA issue... It is always an issue. I remember an MFT test, where the amount of CA "measured" in pixels did not at all correspond to the much higher CA clearly evident in the images, which spread over a lot more pixels. I remember a Zeiss prime test from the same lens sample, where the D3X with a bit higher pixel pitch "measured" quite a bit lower CA than the 5D mk II test "measured".
That is not your guys fault, though, I have to presume it is the software test suite you use which gives such "unreliable" results regarding CA.
Anyway.. the CA oddity I pointed out is not very important. What is, in my opinion, is the weird difference between the D200/D7000 14-24mm f2.8 results, and for instance the D200/D7000 tamron 17-50mm f2.8 results.
Or take another lens tested on both, like the Nikon 35mm f1.8 for instance. On D200 and D7000, it shows what we expect. Now compare that to what the 14-24mm results show. Nikon 16-35mm f4 VR, same story.
I stay with my point that the 14-24mm f2.8 is NOT a good example, and that something more is going on there between those two reviews.