K20D vs. K7 in High-ISO RAW Noise

hikenhi wrote:

The "Overall" score kept declining. The K-10D got a 66, the K-20D got a 65, and the K-7 got a 61, so yes it was time for a change. The K-5 got a 82, which is the highest mark for of this standard for any semi-pro and it also beats many pro DSLRs including the D3. What more do I have to say?

It probably would have been better to say nothing at all on this subject. Your previous input as to the better exposure metering system of the K-7 as compared to all previous cameras are valuable input though as were your thoughts on relative body sizes.

The OP's question was related to whether to choose the K20D or the K-7 as a used camera, indicating that budget is limited.

The K10D used a Sony CCD sensor and its DxOMark overall score of 66 was in line with or slight better than other cameras using the same sensor such as the Nikon D40X with a DxOMark score of 63. Firstly, since all other cameras considered are CMOS sensor cameras and the K20D and K-7 both use CMOS Samsung sensors, I don't know how one can draw any kind of trend line inference from this model. Secondly, the OP doesn't indicate any interest at all in this model, which is relevant to the discussion only because it shares the same body and overall user interface as the K20D.

The K20D has an overall DxOMark score of 65 and the K-7 an overall score of 61, with the K-7 having a lower score due to higher noise due to the addition of higher frequency video read out capability, but according to DxOMark scoring documentation it takes a difference in scores of about 5 points in order to notice it in real images. The reason that the K-7 score is much lower than the others is a drop in low ISO Dynamic Range (DR) and not everyone will use this. In compensation, the DR that the K-7 has is completely usable with very little pattern noise or banding whereas the K20D have a little more pattern noise and banding as well as "border colour fringing", something that the DxOMark scores and graphs do not show. Practical high ISO comparisons as to images probably confirm that the K-7's noise may be just slightly higher in the deep shadows than that of the K20D but it is also natively (before additional Noise Reduction) less "blotchy" so this partly cancels the amplitude of that noise.

I'm quite certain that Greg, the OP, very well knows how good is the image quality of the K-5 (with a new technology Sony CMOS sensor), and would dearly love to have one if his budget permitted, so this is again irrelevant to the OP.

Regards, GordonBGood