Originally posted by Class A Thanks for the link. I'm not sure I'd call the RAW images from the K-7 better than the JPEG images. In the past dpreview gave a camera credit when it retained detail despite showing more noise. I wonder why the K-7 RAW images look as mushy as the competition. Must be the RAW converter applying some heavy handed noise reduction. Whether the JPEG images by the K-7 really have more detail or just look as if they had more detail, I just don't like the clean but washy look of the heavily noise reduced images of the competition.
FWIW, the K-7 will apply some RAW NR starting at ISO 3200 - this cannot be disabled.
Originally posted by Class A I've read a review of the Canon 7D where the users where rather deeply disappointed with the IQ.
Could it be that the images by this camera measure better than they look?
On the "other" forum, GordonBGood had some interesting findings regarding the 7D recently:
"Revisiting the Canon 7D review as to raw noise comparisons only leads to one "aha" that can't be explained as above: Why is the noise performance of the Canon 7D so much better than that of the Canon 50D right across the ISO range? In order to get this kind of improvement, the 18 MP 7D would have to have almost twice as many maximum electrons in the electron well as does the 15 MP 50D, and while Canon has reworked the layout of the new sensor to improve this, they haven't been that effective. At this point I refer you to the results of my work above, that the 7D has unequal alternating green photosites and the news from a few months ago that many raw converters were being updated to support cameras that had such unequal greens, including ACR!
[...]
This also explains the slight softening in appearance of the ACR produced noise gray patches of the 7D as compared to the 50D; when processed identically they have an almost identical appearance. The softening is due to the effective averaging between adjacent green photosites that takes place in demosiacing in order to avoid maze patterns due to the unequal adjacent rows, which reduces noise by about 30% just as having about double the maximum effective electron well capacity would.
I hope some of you find this interesting. I would post this in the Canon Forum but can't be bothered defending it against Canon fanboys. In fact, this doesn't really detract from the Canon 7D's use in any practical sense anyway, it just means that the superior raw noise results from the DPR Canon 7D review aren't real. In fact, I was looking for signs of chroma noise reduction in the K-x and discovered this."
In case you are interested in following this in more detail, here is the link to the post:
Re: 7D raw noise filtering unlikely reply (Conclusion)...: Pentax SLR Talk Forum: Digital Photography Review