Regarding K-7's ISO accuracy/metering performance

[dnaxe]

From comparing the section "low light" in the imaging resource reviews below, it seems that the K-7 meters in low light, at any given ISO:

more than half a stop less sensitive than the D90 (this reduces as light levels decrease)
more than a third of a stop less sensitive than the 50D
somewhat less sensitive than the D300
has similar sensitivity to the 500D
to be more sensitive than the K20D

I think this could mean either that the K-7 is not as sensitive as it is rated or that it's tuned to give a brighter image or a combination of the above?

Having looked at some of the images, I suspect that the K-7 is not as sensitive as the rest and probably similar to the K20d (which, strangely, in those light conditions seems to be tuned to give a brighter image that the K7), although the difference is probably less than half a stop.

If anyone has one of those cameras (the D90 is probably the most interesting as it shows the greatest difference) would you care to compare the difference? (i.e. take shots at the same shutter/aperture/ISO and see which camera produces a brighter shot).

Something else which could be done is to see what the histograms look like for the images - I might do that later.

If anyone has any thoughts about the above, it would be interesting.

http://www.imaging-resource.com/PRODS/K7/K7IMAGING.HTM
http://www.imaging-resource.com/PRODS/E50D/E50DIMAGING.HTM
http://www.imaging-resource.com/PRODS/D300/D300A5.HTM
http://www.imaging-resource.com/PRODS/T1I/T1IIMAGING.HTM
http://www.imaging-resource.com/PRODS/K20D/K20DIMAGING.HTM
http://www.imaging-resource.com/PRODS/D90/D90IMAGING.HTM

 

[flowerpot]

i dun think there is this more or less sensitive... just how the alogrithm interpret the the scene and give the exposure it thinks is correct...

different cam interpret differently thats all...

as to which is more accurate. its a open question. Its the photographer who decides how the scene is to be exposed depending on the subjects, what the cam predicts might not be what the photographers wants but does not mean its not accurate...

the key is it has to be consistent...

 

[dnaxe]

Hm. I think it's less sensitive because in general even though the exposure settings look like more light was let in (shutter was open for longer) the output doesn't look as bright.

I suppose that it might also be an effect of the accuracy of the aperture setting (i.e. 2.8 might be different because of the chosen lens).

 

[flowerpot]

then its not the metering, rather the sensor and the true aperture of each lens... the f numbers are usually round off for most lens...

histrogram is a better gauge... should see a slight shift when exposeure settings are change....

 

[night86mare]

sorry, i am not sure what you are talking about here..

when i read iso accuracy, i think of whether iso100 is really iso100, i.e. dpreview seems to mention a while back that canon entry level dslr (that's the only one i remember, might be across the board).. iso100 seems to be more like iso125, etc.. i hope you get what i mean here.

from the links, i am not sure what is happening, it seems to be due to different metering, that's all.. of course every camera handles metering differently..

 

[flipfreak]

not surprising abt iso accuracy. for the iso 100 on 5d2, a900 and d3x, some people found a discrepancy among the 3 cameras with canon using iso 70ish while the other 2 brands using iso 160 when all the cameras indicate iso 100. its just a matter of how a brand wants the pic to turn out. lesser noise, flatter tones? or vice versa?

 

[creampuff]

Take these online reviews with a bag of salt. In attempting to sound informative and comprehensive, there is a lot of verbiage and so-called "conclusive results" thrown up but ask yourself this:

What lens was used to take the test shots?... can you be sure the lens is identical to a competitor's model?
Exposure sensitivity can vary with the lens being used.

Are you certain the base ISO is what the manufacturer claims?

Like flowerpot has stated, what is important is whether you can get consistent exposures with whichever camera you got...

 

[dnaxe]

sorry, i am not sure what you are talking about here..

when i read iso accuracy, i think of whether iso100 is really iso100, i.e. dpreview seems to mention a while back that canon entry level dslr (that's the only one i remember, might be across the board).. iso100 seems to be more like iso125, etc.. i hope you get what i mean here.

from the links, i am not sure what is happening, it seems to be due to different metering, that's all.. of course every camera handles metering differently..

We understand ISO accuracy identically.

If it was merely metering differences, fine.

However, don't you think that from the test images (not the text of the links) that the k-7 output isn't actually brighter? I was looking in particular at the 11 lux ISO100 and the 0.67 lux ISO800.

not surprising abt iso accuracy. for the iso 100 on 5d2, a900 and d3x, some people found a discrepancy among the 3 cameras with canon using iso 70ish while the other 2 brands using iso 160 when all the cameras indicate iso 100. its just a matter of how a brand wants the pic to turn out. lesser noise, flatter tones? or vice versa?

Yeah. Just something to know about when comparing noise. I suppose there's really no excuse for the higher noise of the K-7. Not that I really care, but...

Take these online reviews with a bag of salt. In attempting to sound informative and comprehensive, there is a lot of verbiage and so-called "conclusive results" thrown up but ask yourself this:

Note that they didn't actually talk about ISO accuracy.

What lens was used to take the test shots?... can you be sure the lens is identical to a competitor's model? Exposure sensitivity can vary with the lens being used.

Yes, I pointed this out above.

I suppose with that difference (ooh, whatever Pentax 2.8 lens they used is not quite 2.8!) plus the generally accepted "acceptable" deviation of +/- 20% for ISO accuracy, is the explanation.

Are you certain the base ISO is what the manufacturer claims?

What? All I assumed was that Imaging Resource was competent enough to set each relevant camera to ISO 100 for all shots stated to be taken at ISO 100.

Like flowerpot has stated, what is important is whether you can get consistent exposures with whichever camera you got...

I actually tend to dial in +0.3 or 0.7 for most shots with the K-7. Might just be my taste

 

[spheredome]

The question runs thru my mind when I browse at ephotozine's review stating iso 800 and iso1600 alike. What if the K-7 iso1600 is only iso1250 of the other brands/model?