K20D review on dpreview, just posted on Thursday

[Jeanne B]

I haven't read through it myself yet but here's the link for anyone that's interested...

http://www.dpreview.com/previews/PentaxK20D/

 

[Groucho]

It's just a preview, not a review.

DPReview really are a bunch of dingbats. Not only do they love to ban people from their forums for nothing, they made a big stink about this alleged "hot pixel" issue, like it was a total dealbreaker.

This "issue" is that apparently the camera (with original firmware, it's fixed in the current firmware) doesn't do automatic hot pixel removal when using the two-second timer - and with jpgs only of course, as raw hot pixel removal is done by the raw conversion software. Of course, DPReview likes to just throw out a bunch of FUD about "putting the review on hold", just like they did with the K10D's alleged "soft jpgs".

Anyway, point being, it would be trivial for them to do the review and just avoid two-second-timer jpg images. But no... I'm fully expecting them to end up trashing it for one thing or another.

 

[Anewman]

It's just a preview, not a review.

DPReview really are a bunch of dingbats.

Maybe, but even if one does not agree with their opinions... No other review site compares with the quality of comparisons.

I can see the output of countless camera models under identical(as close as possible) conditions, where as I have seen other review sites test one camera with a Kit lens and test another with a prime...

If they compare the jpeg output of comparable models with default settings, the images are there for one to decide if it is and issue or not. You do not have to listen to their opinion on the matter, where as other review sites give JUST OPINION.

 

[Groucho]

The full review is up now. It's actually mostly pretty fair, maybe because it was done by Richard Butler instead of Phil.

But really, overall it's still a bit of a mess. The biggest annoyance IMHO is, in the design section, they compare it to the E3, 40D, and D300 (sensible as those are its prime competition.) Yet when it comes time to do IQ comparisons, they put it against the E3, XSi, and A700? This makes absolutely no sense, especially the Canon. The XSi isn't even close to the same spec and is much cheaper, whereas the 40D is very similarly-priced. IQ-wise, I'd be much more curious how the side-by-side shots look with the 40D and especially the D300. From my own observations, the K20D can certainly match the D300 in IQ and The Online Photographer can come right out and said that it's superior. (Granted, we're talking pretty thin differences here, but those are great bragging rights. ) This is especially true as the 40D and D300, being targeted towards more advanced shooters, are less likely to apply aggressive NR than the others - the E3 and A700 in particular turn a bit "watercolory" at even ISO 800 compared to the K20D. Like Anewman says, this is often the most useful part of a DPReview review, and it's not that useful!

This makes it even more frustrating that, after generally positive comments through the review (apart from some rather bizarre hotpixel conclusions which I certainly haven't seen, after pixel-peeping over 5k images, and very few others have noticed, either), they give it only 8.5 in the IQ rating. Surely it deserves better than that. And why 8.5 for ergonomics and handling, when again, through the review, they generally gave it top marks for such things. (Even in the conclusion, ergonomics are listed as among the best of the best.)

Ultimately, they're still mainly interested in lab tests and only give cursory comments about a camera's real-world functionality, usually buried among laundry lists of every possible menu and option. Still, overall a quite positive review (probably because the camera is capable of doing well in lab tests!)

 

[MarkBarbieri]

The XSi isn't even close to the same spec and is much cheaper, whereas the 40D is very similarly-priced. IQ-wise, I'd be much more curious how the side-by-side shots look with the 40D and especially the D300.

I don't know about the differences between the XSi and the 40D, but historically, the Canon APS-C cameras of any given generation have had extremely similar (some say identical) IQ. The difference has generally been in other factors like the feature set, durability, and usability of the cameras.

 

[Groucho]

I don't know about the differences between the XSi and the 40D, but historically, the Canon APS-C cameras of any given generation have had extremely similar (some say identical) IQ. The difference has generally been in other factors like the feature set, durability, and usability of the cameras.

Certainly true. I've seen a few debates about XSi vs 40D image quality, obviously they're so close that it's difficult to pick a champ, especially with the XSi offering extra megapixels. But the XSi is clearly doing more aggressive NR at the expense of detail as low as ISO 800, I wonder if the 40D may do less, with the idea that the buyer of such a camera is more likely to a) do more post-processing or b) prefer more detail even if it means higher noise.

At lower ISOs, the entry-level Canons certain have IQ as one of their strong points - like you say, the extra money goes for the controls, build quality, etc.

Still, I can only assume that DPReview must not have had a 40D or D300 handy during the tests; no matter which way you slice it, the XSi is certainly not a direct competitor. For comparison, their Olympus E3 review does comparison shots with the 40D, D300, A700, E510, and Panasonic DMC-L10. (Ironically, they claim that their tests were done with lenses that included the Pentax 50mm although no Pentax was included - and they don't list the Nikkor 50mm! This is obviously another error, as the actual D300 comparison pages list the Nikon lens, as expected.) Their 40D review includes comparisons with the D200, K10D, and 30D. Not including the 40D and D300 in their comparisons is just bizarre.