[dkittle]

+1 to Gooshin and rparmar. My K20D exposes my grey card correctly 100% of the time. Any other 'scene', I blame myself for not checking the histogram when I'm in a hurry.

[Class A]

[url]
The K20D underexposed for all test shots in the same test scene.

I don't think so. Perhaps in terms of putting the exposure where it should be w.r.t. to a 18% grey reading, but I much prefer the K20D exposure to others that I consider to be overexposed.
Notice that the α350 has a very similar exposure. Do Sony engineers don't know their stuff either?

These are crops from the pictures on the site RiceHigh mentioned, I don't know if I could have found a better example myself to prove my point.

Thanks Marc, for posting these crops. This exposure behaviour was in part why I chose the K100D over the Nikon D40. The latter blows out highlights on a regular basis. A friend of mine has a constant -0.7 EV dialed in. Otherwise, his pictures always come out too bright.

So it's not just a D80/D90 "feature".

[pingflood]

discussing exposure in digital cameras in my opinion is a moot point


first, you have an exposure compensation dial that goes +/- 2 or 3 stops (some of the higher model cameras go like 5 stops i think?)

second, you have 3 parameters to chose from, spot, center, and full


with such controls at your disposal its rather bleak trying to argue which camera exposes the best.

we have been blessed with a histogram right there in the camera, you take the shot, you judge it, you adjust accordingly.


in the past you had to learn how to read scenes, and pay attention to really fine details to get the correct exposure.

now days people want a camera that will automaticaly do everything right for you... and you want this camera to cost less than 10 bucks? honestly now guys...


any exposure problems with digital cameras is user error, IMO, unless your camera is broken.

Hell, in the distant past we didn't even have METERS. That doesn't mean we need to go back there, does it?

I have shot with Canons and Nikons in addition to my Pentax gear and from my experience they just do a better job metering. Especially Nikon; even my brother's cheap D40 nails the exposure in damn near every shot, regardless of lighting conditions. Yes, you can argue that any poor results are "user error"; but, I know how to meter well (using Leicas from the 1930s and 4x5s you kind of have to learn) and I would prefer not to have to worry about it if I don't need to. To me, what matters is the image and if the camera can take one step of manual labor away all the better. Sometimes it can be fun to go "old school" and take your time but sometimes you just want to get the image and then it's nice to have a certain degree of automation.

I love my K20D (or I wouldn't have paid the $ to buy it), but I do believe that other brands do a better job metering and autofocusing in lower light. Not sure why people get so wound up about this and rush to Pentax's defense (some of the sillier examples being "just focus manually") at any mention of the cameras being less than perfect. IMO, the K20D has incredible image quality and performs very well at higher ISO. That coupled with the great Pentax glass new and old that's available is something special. So what if it isn't great in every other aspect?

[daleroy]

two points

the canon version and the pentax version of this lens focus differently (AFAIK). Canon's implements an internal motor, pentax relies on its body motor.

it has been somewhat established that Pentax has a very strict focusing algorithem, its easier to focus faster when your error variable is 2-3 times as large.

my friends Canon 30D focuses fast, but many times the photo comes out OOF so..

This lens in Canon mount uses the body motor. And it's not just that lens, the nifty fifty is a lot quicker than my FA 35, especially on low contrast areas.

In regards to accurate focus, the 40D shots do focus accurately, just as much as the K20D from what I can tell.

Don't get me wrong, I love my K20D, it's just that indoors, it's focusing isn't as quick as the competition.

[frank]

Got the K20D the first day when it reached here in Singapore about 9 months back, really love the new sensor and what it's capable of. Slow AF and FPS aren't problems to me 'cause they are suffcient to my needs.

The only problem I found from this camera: there is no way you can turn off noise reduction for long time exposure! On K10D you can set noise reduction ON or OFF. But on K20D, the only two options are ON and AUTO. Luckily I haven't shot any fireworks w/ the K20D camera yet, but I know I'd be pissed if I use it for fireworks. Don't know if this thing can be fixed by firmware, but seem Pentax don't think it's important.

Other than this, I'm pretty happy w/ the camera.

[Class A]

So what if it isn't great in every other aspect?

Absolutely nothing.

But why accept myths/legends about AF problems when the reality isn't always that clear cut?

And why call something an exposure problem when in fact it leads to better results? One can always pull up the shadows to some extent but never recover blown highlights. Even if the Pentax metering is to conservative in terms of the traditional film method of metering and could strictly speaking said to underexpose, isn't that a desirable behaviour for a digital camera which has no headroom for overexposure as films have?

I know that my K100D is no speed king (FPS, AF) and will never claim it is flawless but I also don't see why I shouldn't prefer it's metering over the Nikon D40's which needs a compensation all the time.

The only problem I found from this camera: there is no way you can turn off noise reduction for long time exposure!

This has been mentioned a couple of times. Could it be that the images wouldn't really be useful without it? Perhaps the standard procedure of substracting a lens cap shot afterwards doesn't work so well with the Samsung sensor which may have a more dynamic noise pattern?

[laissezfaire]

I found that if you are focusing in low light, just simply rotate the auto focus control (it surrounds the directional control) from Auto to SEL or center focus. You can do it w/o taking the camera off the eye and viola! SUPER FAST FOCUS LOCK (center, of course)!

[Gooshin]

This lens in Canon mount uses the body motor.

REALLY??!

which canon dlsr model has a built in focusing motor in the body, please tell!

[frank]

This has been mentioned a couple of times. Could it be that the images wouldn't really be useful without it? Perhaps the standard procedure of substracting a lens cap shot afterwards doesn't work so well with the Samsung sensor which may have a more dynamic noise pattern?

I don't know if the results are useful w/o NR turned on, but I do know that I would miss quite some shots when shooting fireworks w/ K20D that's for sure. It takes too much time for the camera to process NR after each long time exposed shot

[heliphoto]

I don't know if the results are useful w/o NR turned on, but I do know that I would miss quite some shots when shooting fireworks w/ K20D that's for sure. It takes too much time for the camera to process NR after each long time exposed shot

Frank, I'm with you... I love to shoot lightning storms, but at night that often means one or two minutes with the shutter open waiting for a strike. To then wait another two minutes waiting for this NR process. It will be maddening (but do I really want to carry my K200d around all summer too, expressly for lightning?... we'll see).

[k100d]

Now that's funny and the best post in this thread full of BS.

why is this thread full of BS? does the K20D have no flaws?

Feel free to post sample shots. It's usually quite simple to demonstrate how the exposure chosen was the brightest that would avoid clipping, or that the average luminance is right where ISO standards say it should be.

i can find plenty of examples where "not blowing the highlights" is not the right "excuse". i always find i need severe EV compensation in the shade or especially indoors. if i forget, i try to +1EV PP and then i got lots of noise and i'm not happy. in bright sunlight, usually things are fine. or maybe i prefer my photos bright.

centerweighted. i would add +1 to this


this restaurant always fk's up the meter. this could use +2EV, look at the histogram, it's awful, it's like the camera decided to expose to the left. excuse the WB, quick conversion.

[Marc Sabatella]

i can find plenty of examples where "not blowing the highlights" is not the right "excuse".

These are great examples of pictures that might be darker than you personally wanted, but *not* represent underexposure in the photographic sense. The histograms average a little left of center, just as it is supposed to. These illustrate the more conventional reason why people erroneously perceive a camera as underexposing a scene: in the absence of bright highlights that would force the camera to render a scene even darker, the main goal of the meter is to try to represent the scene as being somewhat darker than gray card on average. Meaning the histogram should average a little left of center. And that's exactly what happened in these cases.

Any book on photography will tell you that when shooting an object or scene that is lighter on average than a grey card, you need to apply expose compensation. A white wall is the classic example, and the picture of the cake you posted is in the same ballpark. *Of course* that's not the exposure you wanted, but it's absolutely the correct exposure for a meter to suggest. As always when shooting essentially white objects, it's up to you to override this to make the object *look* white, and not a little darker than a gray card. This is how SLR meters have worked for decades.

[jeffkrol]

why is this thread full of BS? does the K20D have no flaws?




this restaurant always fk's up the meter. this could use +2EV, look at the histogram, it's awful, it's like the camera decided to expose to the left. excuse the WB, quick conversion.

First all cameras have flaws or nobody would need to produce another one......
As to your cake, shooting that is not much difference then shooting snow (I assume the cake is whitish). Lots of very light colors which the camera wants to make !2% grey.. (does nobody pay attention )
Here is the image and histograms. 1)Base exposure (about what I, note I expect. A mean of around 100ish.).
2)CONTRAST adjustment not exposure adjustment which would just have slid the whole thing over to the right. At least that's what "exposure adjustment" is supposed to mean. What I did is a "contrast" adjustment. Just drag the whitepoint to full 255, stretching the CONTRAST not exposure...


What the heck color is the cake???

[RiceHigh]

what "exposure adjustment" is supposed to mean. What I did is a "contrast" adjustment. Just drag the whitepoint to full 255, stretching the CONTRAST not exposure...

Funny that your this "drag the whitepoint to full 255" action is already and purely just pushing up exposure for the "white point" you selected as a reference point (to the brightest) for changing *both* the exposure and contrast of the whole picture, despite that it is just named "contrast adjustment". Afterall, this function *stretches* the histogram to full span and thus change *both* the exposure and contrast, nothing more, nothing less.

Btw, do you know what you are doing actually? :-)

[kenyee]

I'm puzzled by that contrast adjustment too (I don't think it should have affected exposure like that). What software is that? I don't recognize the dialog boxes...