Sunday, February 24, 2008

A Quick but Full ISO Test by a New K20D User (Vs K10D too)

A Hong Kong Pentax user has acquired a new K20D at the street of Hong Kong and he has immediately carried out a quick test for checking the image quality and camera performance (variations) at all different ISO speeds and here is his post:-

(Text in Traditional Chinese, use Babelfish to translate if needed. Btw, the captions in the posted test images are self-explanatory and are in English.)

The first thing to note is that the K20D image starts to appear grainy/noisy at ISO 800 and in my opinion ISO 1600 is already very marginally usable. But if you compare the K10D images which are shot at the same time for the same object with the same setup, the ISO 400 image is somewhat more or less the same as the ISO 800 image of the K20D, which means that there is a considerably obvious improvement in the noise performance of the K20D when it is compared to the K10D, especially considering that it has nearly 50% more pixels than the K10D. Having said that, it should be noted that the ISO 400 noise performance of the K10D is in fact sub-par, which is also quite some K10D users and I have been talking about for a while now.

The second thing is that the user reported that the (Auto) White Balance is far more accurate than the K10D, even under tungsten lighting. The reproduced color balance is more faithful to the original subjects as he could see so far (way better than the K10D, as reported). Well, it might be that it is really true or it is just that he is still enjoying his honey moon period with his new toy, anyway.

But the third thing which I can observe is the dark and gloomy pictures, which actually nearly becomes an unshakeable symbol or even icon for Pentax DSLRs, *can* *still* be seen. The K20D pictures look just a little bit brighter, though (but it still looks dark, muddy and dull).

Last but not least, since he purchased his K20D body (only) at near HK$11,000 (about US$1,400), which I would consider simply too expensive to consider objectively as the much more stronger (in specifications) and faster (for its true camera performance, e.g., AF and shutter lag, etc.) Nikon D300 is also selling just at a price which is dearer by very little bucks.

Do note that Nikon's system is undoubtedly more comprehensive and complete than the Pentax's one and Nikon's flash system is renowned to be the most accurate and versatile one, too, whereas Pentax's P-TTL has been reported from time to time rather frequently on the Internet for its inaccuracy and inconsistency. Although the K20D can be different but I have no high hope as long as exposure accuracy department is concerned for the Pentax DSLRs, judging from the past experiences of the Pentax DSLR users (and now the new samples published by the beta and production K20D so far). Besides, as long as the AF system of the K20D is yet the SAFOX VIII (at least as it is called), which is infamous for unsatisfactory low light performance and accuracy, I don't think choosing the K20D is a wise choice when one can actually get a far superior and better Nikon D300 at nearly the same price.

In fact, I have just tried a D300 yesterday (not at shop, one of my friends has purchased his new D300 so that I could try it out at the field), the mirror action is so gentle but yet crisp and I am really amazed with its lightning fast short shutter lag which is noticeably even shorter than my Canon 5D (around 78 ms as measured by CAPA Japan and the Imaging Resource, best timing) which is again noticeably faster than my MZ-S (which is again just faster than *any* Pentax AF DSLRs ever made!). So, I just wonder except the old Pentax DSLR user base who are particularly eager to replace their old Pentax DSLRs (which the users might be rather unsatisfied with something), for those who are new, WHO will really choose the K20D in view of the comparatively poorer price/performance ratio? So, here is my bet: Unless the price of the K20D goes down significantly, the product and Pentax won't succeed. Or, people just go to get the Samsung GX-20 which should be (much?) cheaper and with the original SilkyPix RAW software package (and better user interface than the Pentax Photo Lab), too.


nlx said...

For once, i can agree with you. I think the price of the K20d (unlike the K10d) is slighty too high. To me, it seems a very good camera but the price range make people compare it to Nikon D300 and Canon D40, thing that could be a good thing if the Pentax price were clearly better thanthe one from those competitors.

But anyway i think the Pentax price will drop down very quicly : faster than Nikon and Canon price. I think Pentax know that but that they want to enjoy the bucks of the consumers that just can't wait.

If the price drop down just a bit, it will be a killer.

Anonymous said...

I think the price is too high as well, since the biggest reason to use Pentax is their value for money.
But you can't judge a camera on price based on its RELEASE price, its always inflated to catch early adopters, as previous poster said.

RiceHigh said...

I just think Pentax should just make a better body if they wished to sell at that high price OR they should sell at a (much) lower price for what the K20D actually offers in its package.

Anonymous said...

I think if AF speed were improved and AE were a bit more consistent, this camera would be a winner even at this price.

But in-body SR which is much improved (supposedly) over K10d, which itself was much improved over k100 makes this camera uniquely valuable.

For people like me who don;t bring a tripod everywhere, this feature alone makes Pentax the best choice... when compared to Sony's offerings, Pentax is much, much better in every way.

Jaka said...

I'd be prepared to pay 1000€ for a K20D but I think it is wise to skip this year and think about buying a new body in 2009.

I'm not saying the K20D is a bad camera just that it's not that much of an improvement from the K10D.

Then again I'll have to wait and play with it some to come to a conclusion.

RiceHigh said...

I hope there will be a K1D in 2009, which should be a 135 Full Frame DSLR.

But since Pentax is still unable to introduce new AE and AF systems in their K2X line, I doubt that if they can make things right (suddenly) even some days later when the K1 flagship comes.

Dancel Pangilinan said...

I second Jaka! its not enough for to update.
1) the live view system on dslr's will improve by the next models for all camera for sure.
2) Noise control is gonna be better
3)I different body for the new pentax.

my k10d is doing fine no need to upgrade.

Jaka said...

I just played with the K20D and the 200mm f/2.8 lens yesterday at a fair in Zagreb and I'm not that much impressed. The noise level at ISO3200 is impressively low tho.
I tried the 21FPS mode too and it's done totally digital so you don't even get a fell of how many shots you've shot.
Live view is not impressive at all. 40D does it better IMHO. The mirror locks up, the image on the screen is great and responsive (grid display can be added in the menu), focusing is a pain in live view ... you don't even get a focus confirmation beep or anything. I have to say that I'm not a fan of live view in the first place.

Other than those quick observations it is a good camera if you don't own a K10D already. I'll skip this year's purchase and rather buy a lens.

Oh, the 200mm f/2.8 feels slow to focus (my 50-135mm f/2.8 does it faster - not measured just a quick test and feel). And we all know that testing equipment should take a longer period of time in different environments.

Just my opinion based on half an hour with the camera and lens.

Anonymous said...

The D300 ($1700-1800 on March 8, 2008) is hundreds of dollars more expensive than the soon-to-drop-in-price K20D ($1300-$1400 on March 8, 2008). Add to that a decent Nikon lens that fully takes advantage of the D300's superior capabilities and matches the K20D image stabilization, and there's no question that the K20D offers enthusiasts incredible value.

Post a Comment

Related Posts

Creative Commons License
RiceHigh's Pentax Blog by RiceHigh is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.