[dafiryde]

AF on the K20 locks faster, especially in low light conditions, than my K10 does. .

high ISO noise (visible noise). With the K10 I used 100 or 200 ISO without hesitation,
The K20 is different. I can use any setting up to ISO 400, without hesitation.
Ben

Ben , sorry for editing your post, but i broke down what you wrote, as these two issues, on the focus speed/lock in low light
and much clearer pictures at higher iso levels, are my 2 most basic concerns.
the #1 concern , being high iso levels, i would like to know that i will see little to no grain/noise in 100,200,400,800 iso settings ( with K10 400 is max setting )

Dave

[marcdsgn]

At least one advantage the K10D has over the K20D:

It currently sells new at about half the price of the K20D, and its technology is still comparatively impressive for that price margin.

[Marc Sabatella]

the #1 concern , being high iso levels, i would like to know that i will see little to no grain/noise in 100,200,400,800 iso settings ( with K10 400 is max setting )

I'd suggest going to a site like dpreview and finding their sample galleries (I'm assuming all the review sites, such as Steve's Digicams or dcresource also have this). Also see the Pentax Photo Gallery. There are tons of samples around for you to come to you own decisions, which is the only way to get the answer that is right for you. Everyone is going to have their own idea of what the maximum acceptable ISO is. Some say 400 for the K20D, some say 800, some say 1600, some even say 3200. Depends entirely on your own tastes and needs.

[Ben_Edict]

Ben , sorry for editing your post, but i broke down what you wrote, as these two issues, on the focus speed/lock in low light
and much clearer pictures at higher iso levels, are my 2 most basic concerns.
the #1 concern , being high iso levels, i would like to know that i will see little to no grain/noise in 100,200,400,800 iso settings ( with K10 400 is max setting )

Dave

Dave, as Marc points out, it is highly subjective, which noise-level is acceptable. I personally find, the K10 very acceptable at ISO 400 and good for snapshots at ISO 800. The K20 is in my eyes noiseless at ISOs 100 and 200, has hardly any noise at ISO 400 and is still very acceptable up to ISO 1000. Only at even higher ISO settings, noise gets visible to a degree, which makes it only acceptable as a last resort - better some noise, than camera shake or motion blurr... BUT even at ISO 1600 or 3200 you can still make 4x6 (inches) prints off the K20 pics, which many people will find nice.

In terms of noise I generally find, thatboth cameras have a noise-pattern, which is very film like. Thus they retain more detail in the image, than most other DSLRs , because Pentax doesn't smooth out the noise too much. I very much prefer that approach. If you remember film: THe old 3M Scotchchrome 1000, which was the first available highspeed slide film, had at ISO 1000 much(!) more noise than the K20 at ISO 3200. And the colour rendition of the K20 is much better, too. Up to ISO 800 I would rate the K20 equivalent to Kodak Ektachrome 400, which was a good slide film.

Ben

[kurt]

I agree with Marc. It is matter of choice, taste and feeling. For some shots the 'old filmstyle' high ISO grain is the thing which gives the image something.

[Ben_Edict]

Does the long-exposure noise reduction also work on RAW files? I'm using an older Olympus DSLR currently that can only apply the dark-frame subtraction when shooting jpeg, so I have to go spot out the hot pixels by hand if I shoot RAW for long exposures. Does the K20D fix that?

As I shoot only RAWs, I would say, dark-frame substraction also works in RAW-mode. I cannot guarantee that (there is no printed info on that by Pentax), but the K20 always takes a second, dark exposure automatically. And as thefinal RAW images show no obvious hot pixels, I assume, the dark-frame substraction has been taken place...

Ben

[dafiryde]

I'd suggest going to a site like dpreview and finding their sample galleries (I'm assuming all the review sites, such as Steve's Digicams or dcresource also have this). Also see the Pentax Photo Gallery. There are tons of samples around for you to come to you own decisions, which is the only way to get the answer that is right for you. Everyone is going to have their own idea of what the maximum acceptable ISO is. Some say 400 for the K20D, some say 800, some say 1600, some even say 3200. Depends entirely on your own tastes and needs.

in film days, my favourite was the fuji NHG asa 400. reason was this film gave bright ,clear , fine grains .the same as a 100 speed.
in comparison with the k10 this is about setting at 200 iso, so i am counting on someone out there who also knows this film, if they can safely say that the k20 is on par with the NHG or better or not as good , at the same asa.

Ben i have only now read your post after , and noticed you did give a comparrison with the kodak ektachrome 400 film, which is a comparisson i can understand, as this was a great film, but i personally found back then, that the fuji NHG gave much tighter grains.

Dave

---

Last edited by dafiryde; 11-06-2008 at 08:52 AM.

[rfortson]

A picture is worth a thousand words.



This shot illustrates two of the main improvements of the K20D over the K10D. This was shot with the DA*50-135 in AF-C. Captured at f/2.8, 1/180s, 135mm, ISO2000. Forgetting that the K10D can't actually be set at ISO2000, the noise in this shot is pretty minimal (you can see a 100% crop on my Flickr stream). Also, given the exposure conditions, you can see how dim it was (this was an evening, covered small-time rodeo with poor lighting), yet the K20D managed to lock focus dead on.

The K10D is a fine camera, but the K20D just trumps it in every category (save the dark frame subtraction mentioned earlier). It does everything the K10D does, except better. Plus it does things the K10D can't do. Did anyone mention the ability to map out hot pixels? That and dust mapping are two things I've actually used on the K20D that the K10D doesn't do.

[kenyee]

Did anyone mention the ability to map out hot pixels?

The negative is some K20D sensors exhibit random pixel noise so they can't be mapped out. Seems to happen w/ longer exposures from what I've heard.

[Ben_Edict]

in film days, my favourite was the fuji NHG asa 400. reason was this film gave bright ,clear , fine grains .the same as a 100 speed.
in comparison with the k10 this is about setting at 200 iso, so i am counting on someone out there who also knows this film, if they can safely say that the k20 is on par with the NHG or better or not as good , at the same asa.

Ben i have only now read your post after , and noticed you did give a comparrison with the kodak ektachrome 400 film, which is a comparisson i can understand, as this was a great film, but i personally found back then, that the fuji NHG gave much tighter grains.

Dave

I rarely used NHG, more NPS... But as I wrote, the K20 at 800 ISO is very near Ektachrome 400. So at 400 ISO, the K20 is less grainy, but in terms of colour and contrast quite as pleasing. Up to 400 ISO the contrast range is still quite good. Starting at 1000 ISO the dynamic range starts to crumble, but this is the same with slide film and all the other DSLRs too, except perhaps Nikon's low pixel count, full-format sensors in the D3 and D700 (which are in a very different price range, then).

Ben

[Ben_Edict]

The negative is some K20D sensors exhibit random pixel noise so they can't be mapped out. Seems to happen w/ longer exposures from what I've heard.

Those random pixels are very different from typically stuck hot pixels. One problem the K20 exhibits is, that it tries to repair noisy pixels by interpolating from neighbouring pixels. That can lead to strange artefacts if your image has hard contrasts and the hot pixels are just "across the border" from those pixels, the correction algorithm chooses for its magic. That way I once (only once in exactly one image of a series of perhaps 150) got a white letter inside a printed word, which should have been completely black. (That was the spnonsor's name printed on a hot air balloon). Obviously, using the typical targets (USAF opr whatever) for measuring resolution with their hard edged black and white patterns, really provokes that misbehaviour - and such tests are not indicative of real world images.

Ben

[dafiryde]

[QUOTE=rfortson;387008]A picture is worth a thousand words.



i am impressed

back to cream puff listing

#1 never had that problem
#3,7 no problem with, as i shoot RAW
#2,4,5,6,8,9,12 will be welcomed
#10,11 i dont need

i would think that 7 out of 5 , especially those 5 justifies an upgrade from k10 to k20 is worth it.
waiting on what the k3 whatever would not be reality until next summer, so until then, i will worry about those improvements then.
as for now i am convinced that i will enjoy the improvements from the k20

Dave



Dave

[kenyee]

One problem the K20 exhibits is, that it tries to repair noisy pixels by interpolating from neighbouring pixels. That can lead to strange artefacts if your image has hard contrasts and the hot pixels are just "across the border" from those pixels, the correction algorithm chooses for its magic.

Thanks, Ben. Interesting observation.

[Ben_Edict]

Thanks, Ben. Interesting observation.

Not my credentials, Pentax Japan officially explained the stray pixels this way, some time ago. In my personal (surely limited) experience, visible problems arise only very rarely indeed, as I don't shoot test targets that often...

Ben

[rfortson]

The negative is some K20D sensors exhibit random pixel noise so they can't be mapped out. Seems to happen w/ longer exposures from what I've heard.

I believe the firmware update fixed that, though I've never seen it on mine, even when I tried the "recipe" for making my camera show hot pixels.

Not saying I haven't had hot pixels, just no more than any of my other cameras. And when I used the hot pixel mapping, they "dissappeared".