[UnknownVT]

Just got the paper magazine Popular Photography - April, 2011 issue -
full review of the K-r - scan of their very useful test results panel:

[loco]

Not possible. I read somewhere else the image quality is just "decent".

[macTak]

Interesting to see a measure of the SR efficiency--they did not have that with the K-5 review (I also note it gives a better result than DPR's test). What surprises me is that at every ISO from 100 (save at 200 where it's a tie) the K-r does better in noise than the K-5 (a very different result from dxo and I'd say from every studio comparison I've seen). This is especially the case at the higher ISO's; the K-r ISO 25600 is rated 9.5; whereas the K-5 ISO 12800 is rated 9.2 and 25600 is rated 15.3. Something doesn't seem quite right with these results....

[Laurentiu Cristofor]

Something doesn't seem quite right with these results....

+1

But at least they're not shooting the camera down for the wrong reason.

[UnknownVT]

Interesting to see a measure of the SR efficiency--they did not have that with the K-5 review (I also note it gives a better result than DPR's test). What surprises me is that at every ISO from 100 (save at 200 where it's a tie) the K-r does better in noise than the K-5 (a very different result from dxo and I'd say from every studio comparison I've seen). This is especially the case at the higher ISO's; the K-r ISO 25600 is rated 9.5; whereas the K-5 ISO 12800 is rated 9.2 and 25600 is rated 15.3. Something doesn't seem quite right with these results....

I would tend to agree with you on the HighISO results -
eg: almost all reviews I've seen say that the K-r and K-x are very similar -
because I own a K-x and wanted to see if there was any improvement of the K-r over the K-x
I've scrutinized the standardized sample test images over at Imaging-Resource.com -
there was very little difference -
if anything, I thought the K-x images may have been better at HighISO -
But there were others that did not agree -
which only really goes to show it was very, very close -
probably well within margin of experimental error,
and of course any personal bias

Just for comparison/interest from PopPhoto.com K-x test panel:


But despite these discrepancies -
I still think the Popular Photography review was very worthwhile -
as they often have insights and test certain aspects that others tend to leave out -
like the all important AF speeds -
not just at some arbitrary light level -
but over the full range of light levels as given by the specs.

[e6zion]

pretty impressive

[er1kksen]

I've always gotten the impression that PopPhoto's high-ISO measurements are heavily skewed by the jpeg engine's NR. Perhaps comparing the ISO it goes up to while still rating "very high" resolution is more meaningful, since that provides a better insight into whether a camera is getting a lower noise rating by just smearing everything away.

[UnknownVT]

I've always gotten the impression that PopPhoto's high-ISO measurements are heavily skewed by the jpeg engine's NR. Perhaps comparing the ISO it goes up to while still rating "very high" resolution is more meaningful, since that provides a better insight into whether a camera is getting a lower noise rating by just smearing everything away.

I think you're right that reviews ought to compare noise levels at given resolution levels
- kind of like MTF readings.

But I have the feeling that PopPhoto are probably using "industry standard" tests - ie: likely to be what everyone else is also using - the Imatest analysis.