K20D Noise Comparison against A700, 40D, D300, D3 - Page 2

robgo2 · 02-07-2008, 07:35 PM

Jeff,

Thanks for putting this together for the rest of us.

My first impression is that the K20D shots are better exposed than all of the others. The 40D shots look out of focus. I cannot believe that the Canon could be so soft. The K20D noise is certainly the closest to film grain and is superior to all except the D3, but some of this superiority may due to better exposure. All in all, it woud appear that the K20D may outperform its competition in high ISO IQ, and from what I have seen previously, in low ISO IQ as well. (Note: I do not consider the D3 to be its competitor.)

Rob

R@mbo · 02-08-2008, 06:17 AM

Originally Posted by robgo2

The 40D shots look out of focus.

I see so too!
The sharpness of the 40D is normally really good.

But the K20D is a big surprise; will it have >5fps.. it will be mine.

*John Kovarik* · 02-08-2008, 06:56 AM

Oustanding! Assuming some confirmation of these pix as representative of K20 IQ and not unfair to the competition, it is clear to me why the Product Manager in NY last October just smiled and said, 'Something wondeful is coming.........'

I think I'll order one now............

Tbear · 02-08-2008, 07:49 AM

Thanks for posting. Very, very interesting comparison! Looking better each day for the K20D sensor. Even though I am not getting it, I am very happy for everyone that is.

Ted

Ishpuini · 02-08-2008, 08:20 AM

In all honesty, the pictures make comparison of the signal-noise-ration impossible. Mainly since the K20D shots are overexposed compared to the other shots. Or the others underexposed compared to the K20D. I'm not judging correctness of exposure. Since noise comes from pushing the shadows, relative overexposure helps reduce noise.

Quality of noise and relative increase in noise between stops can be judged better here. That looks great to me.

Wim

lol101 · 02-08-2008, 02:27 PM

Originally Posted by Ishpuini

In all honesty, the pictures make comparison of the signal-noise-ration impossible. Mainly since the K20D shots are overexposed compared to the other shots. Or the others underexposed compared to the K20D. I'm not judging correctness of exposure. Since noise comes from pushing the shadows, relative overexposure helps reduce noise.

Quality of noise and relative increase in noise between stops can be judged better here. That looks great to me.

Wim

I frankly could care less about comparing the K20 to other cameras since I am not in the market to replace a D300 but my K10 but what I conclude from this "test" is this:

The K20D exposure is better than the other and it didn't get fooled by the white dress as much.

The noise characteristics are looking good an I think that this is with NR "off" on the K20 while the competition all have NR on.

Still, there is room for better performances since it could have been paused a little bit more to the left, giving even better noise figures without touching anything on the impressive resolution.

To me, it just spells outstanding IQ. Is it better than the competition? I don't know... but since they are all excellent... who cares?

d.bradley · 02-08-2008, 03:32 PM

I think the differences in exposure might be caused by the relative differences in exposure calibration for ISO values. That would assume that each camera had the same ISO, shutter speed and aperture, which may or may not be the case.

I'd like to see the same shots done from DPreview....at least they're a reputable tester...

lol101 · 02-09-2008, 01:31 AM

Originally Posted by d.bradley

I think the differences in exposure might be caused by the relative differences in exposure calibration for ISO values. That would assume that each camera had the same ISO, shutter speed and aperture, which may or may not be the case.

I'd like to see the same shots done from DPreview....at least they're a reputable tester...

exif show that they have different shutter speed.

d.bradley · 02-09-2008, 02:20 AM

yah I was too lazy to go check the originals. Perhaps it is metering then....in which case it could be argued that the camera that can meter to expose properly with least noise would have an advantage. :P This would not however suggest that under ideal conditions, this would be the least noisy camera...

thibs · 02-09-2008, 06:28 AM

Originally Posted by d.bradley

I'd like to see the same shots done from DPreview....at least they're a reputable tester...

Hum... oh not in my book at least ...

d.bradley · 02-09-2008, 08:19 AM

Originally Posted by thibs

Hum... oh not in my book at least ...

I didn't say they were unbiased, but they don't have testing problems like poor focus or differing exposures when they put the cameras through their paces.