[OrenMc]

I recently bought the K20D and after today it is going back for exchange for sure.

I started a thread a few days ago thinking I had a FF/BB problem. Well this is what I shot today. I think there is some sensor noise going on also.
What is going on??? Please let me know if exif is intact. I am not very proficient at this so that is why I need the help.

This first shot I took in raw and resized. I also bumped the exposure ( not the reason for the noise ) what is making it so damn soft. My Sigma is a razor.
ISO 200 f/4 1/350


This second one is what? out of focus? Noisy? ISO 400 f/6.7 1/90 sec


This is just plain terrible!!!! ISO 200 F4/ 1/350 sec


I've never had a problem with my K10D. Maybe part of it it isn't exposing properly. I have been able to get some okay shots with the camera but even then there is always something lacking.

I plan on sending it back Monday but I though I might get some insight before I do.

I can get the camera exchanged but I took some great shot today that should have been keepers. Thats the big bummer.

You can go to my flickr page and see a larger file
Thanks

---

Last edited by OrenMc; 12-13-2008 at 09:19 PM.

[kenyee]

Was that w/ the Sigma 100-300? :-)
The first shot looks great. The 2nd shot might be slightly FF'd (maybe 3') though I suspect you didn't think so when you posted this ;-)

[Wheatfield]

Really sucky bokeh.

[OrenMc]

The first one looks fair at a distance but it just doesn't have near sharpness the this lens is capable of. And if you look at my flickr page at the larger size you will see the noise.

I think your right about the second one FFing. but a closer look at it shows a lot of grain.
The third one, going by the histogram should have turned out better than it did. It looks like underexposure grain. This one confuseses me the most. I had the camera in AV also

---

Last edited by OrenMc; 12-13-2008 at 08:47 PM.

[OrenMc]

Really sucky bokeh.

Maybe thats what I should call it
But thats another thing I like about the Sigma is the Bokeh. It is usually very nice.

And I should have had results more like this on the last one.

[Peter Zack]

before even attempting to assess the shots, what ISO? EXIF data is missing. The final shot is at leat 1 shot underexposed and that will add noise to the shot without question.

[OrenMc]

Thanks Peter,
How do I keep EXIF intact when I am diong this? first and last shot were at ISO 200 , second shot ISO 400.

[Jewelltrail]

OrenMc: I've never had a problem with my K10D. Maybe part of it it isn't exposing properly. I have been able to get some okay shots with the camera but even then there is always something lacking.

I plan on sending it back Monday but I though I might get some insight before I do.

I can get the camera exchanged but I took some great shot today that should have been keepers. Thats the big bummer.

The first shot is not a bad looking shot. I think the highlight clipping in the background, because of the water's reflection of sunlight, is giving you a less-than pleasant bokeh. Compare that with the shot you like. In the shot you like, the background is much darker, close to the bird itself actually, making for much easier metering. Considering the very difficult metering in the first shot and that it was only shot at f4, it's a decent shot I think.

The 3rd shot is very much underexposed. The EXIF data has been stripped, but a histogram is still available in Photome. The histogram confirms what the eyes see; the shot, where your subject is concerned, is very much underexposed. But, if you look carefully, you'll see there is highlight clipping once again due to the water reflecting light. This too is shown clearly in the histogram. Notice how the ripples, in the highlight areas, are blown to the point where the ripples are not visible. I think this is just another case where you are expecting a lot from the metering. You could have spot metered for the bird (your subject) and you probably would have got a nice sharp bird, but the background would have blown out--which may not have been a bad idea in this case because the bokeh would have been much smoother. If you have a difficult to meter scene like this, you really need to meter for your subject (take the metering controls). You can always crop later. You really can not expect the camera to know what you are trying to do here. The camera simply tried to do what it is supposed to do under these circumstances: it tried to get the best exposure, for the whole frame, while minimizing clipping.

The 2nd shot is more enigmatic. It looks like you may have moved a bit when taking the shot--camera shake. The SR is very effective most of the time, but not in all cases, of course. The metering is good in that the details of the light-colored dock have mostly been preserved, with detail in the rest of the scene staying intact--it just is not a sharp pic--was SR on? What focal length did you shoot at?

I find avoiding noise is sometimes tricky in tough metering scenes. You really need to expose to the right as much a possible, while still getting the exposure you want. You clearly are underexposing here, magnifying noise, even if unintentional.

Also, you have stripped EXIF data--this is often the result of software used to process pictures. Try to use software which does not strip. The pic which you posted with the easy metering scene, from your K10d, still has EXIF data in place. I notice you used a lower ISO in this pic (100). In the K20 shots you have used one and two stops higher--that, coupled with the underexposure and tough scene metering would give you these results--I think. I am no expert, but I hope I have been of some help here.

I remember how hard it was when I first got going with the K20d, metering in particular. In your K10 shot, you used "pattern" metering. If you also used this in the K20 shots which you display here, the results are, I think, to be expected--except for number 2.

I wish you the best. I would not give up on the camera yet. The K20 can take some time to get comfortable with though it is, I think, well worth the effort.

[Peter Zack]

Jewelltrail has covered all of this very well and I think he's correct in the assessment. I can't add much to that great post but I'll make an assumption that you have the D range on in the K20D. I'm not a fan of that 'enhancement' in most shots if there are dark or deep shadowed areas in the image. I find the noise is increased and if you couple that with 1-2 stops of underexposure, you have a noisy shot. Turn it off 90% of the time IMO and exposue to the right.

Pentax metering is conservative anyway and will tend to underexpose 1/2 a stop or so most of the time. I use spot metering always with very few exceptions. If the scene is complicated, I'll check metering in a few parts of the scene before deciding on the settings. When in doubt, bracket when possible. I'd much rather Pentax metering than Canon's approach where everything is blown out in bright or contrasty conditions. At least under exposure can be recovered and worked with. Blown out is a whole lot of white nothing.

Put the main point is if you underexpose, you'll have noise. Happened with film and happens on a DSLR. It's a touch more in the K20D than the K10D if the shot is not exposed properly. Use spot metering.

Don't return the camera, there's nothing wrong with it. It might look like the K10D but it's not. Learn to shoot with it and practice more with it's differences. It will reward you and from my own experience, it took a couple weeks to learn how to use what is a really great camera.

[OrenMc]

You beat me to the post Peter, I was writing this up before I seen your post

Is th emetering system that different in the K20 as the K10?

I have been shoting the k10D for around a year now and have developed some habits of shoting. I basically treated the camera and conditions that same. I guess what I am trying to say, I feel if I would of had the K10 and took those same shots I wouldn't have had the problems. The K20 loves the new 77 limited but if I shot with any of my other lenses I seem to have trouble of some kind.

I had the SR on in the second shot and I can understand why this would be a thought. I took other shots that turned out pretty much the same. I have never had a problem handholding the Sigma. To me, #2 looks like it has something else going on other than motion blur.

I will try to get out today and and watch for the problems that you have suggested. I was lucky yesterday and had a short break in the clouds and a little sun. There is snow on the ground this morning in Seattle, a rarity.

I am looking forward to the use of the K20. I got my K10 back out and swaped out the scratched focus screen with a split screen, it was suppose to go in the K20 but I am going to wait until I figure this out so I don't screw up the warrenty. I may just end up keeping it as a back up if the resale prices keep dropping.


Thank for the suggestiions.

[Peter Zack]

No the metering system is not different a far as I know. Although there could be some minor changes to adpt it to the new sensor. All I'm saying is the metering is set to underexpose a bit more than other systems.

But the sensors are very different and take some adjustment to your shooting style. I will often shoot at +2/3 (+.7) with this camera.

[OrenMc]

Thanks Peter,
I will most definately keep that in mind today when I get out.


Heres a few more examples from yesterday. I really think it is at least FFing.
This one was underexposed also and I bumped up the exposure in PP.




I had the focus one the gal in the red jacket with this one.

[ghelary]

I had a look at your flicker page, those images looks to me like 100% cropping of the original image. Such a crop would have been impossible with the K10D because it would make something like 130% magnification. You would see a lot of pixels.

When PP, I often use a lot of chrominance noise reduction, while leaving luminance untouched. This gets the image some grain a bit like 35mm film.

[OrenMc]

The heron in both sets is a crop from K20. But the other bird in the series is a crop from the k10. The other photos are only resized not cropped. I went out today and took some more photos with both cameras and will post examples. I haven't had a chance to even look at the yet.

[Peter Zack]

Well if it's really a FF issue, then you can adjust that within the camera and that's not really a problem. But to be sure you need to test it on sbjects that are not moving and use a tripod.