K10D and that "Pentax look."

[Prieni]

I find the amount of difference puzzling myself, Prieni, and I used to think that the more megapixels the more detail myself.

I was hesitant to step up from the *istD (6 MP) to the K20 because of the high MP number. 6 MP is quite enough for what I need, even had a 2 m x 1 m enlargement made from a 6 MP image without problems. But my concern was noise, mainly. The K20 certainly shows more detail than the *istD did (and noise performance of the K20D is also better). I never owned the K10D, so I might have missed out on the better one...

The difference, I believe, is between CCD and CMOS sensors. Despite the advancements in CMOS sensors, especially for high ISO performance, scientists and astrophotographers still prefer CCD. Detail is most important to them.

Dynamic range is also important to them. And image uniformity, as the light level in many cases is a measure for something else. And uniformity (i.e. x photons giving exactly y volts for any given photo site (pixel)) is better for CCD's than for CMOS.
There is a comparison of both systems from Teledyne Dalsa here.

Most cameras designed specifically for underwater photography use CCD. Panasonic's newest underwater camera reverted back to a 10MP CCD from the previous 12MP CMOS. It's not cheaper, and they think it is better--an improvement in IQ.

I imagine that here the better fill factor is decisive, as underwater you always have lower light levels.

Even the tiny CMOS sensors in P&S cameras now feature 14 or 15 MPs, but I think all this is not driven by function but fashion and general market desire for more megapixels.
It's kind of like fishing lures, with which I'm very familiar. Fancy paint jobs catch many more fishermen than fish.

Absolutely agree (and I like your comparison  ). It's crazy for the P&S cameras, but it might well be too much for DSLRs. No camera manufacturer, however, can afford to opt out of that race for higher numbers of pixels.

I certainly don't think that newer is always better, you just have to look at all the HD-ready trash TV sets that are thrown at you for cheap money. A lot of them are certainly much worse than a quality CRT one. On top of that the digital channels are in many cases horribly compressed so that you can see all sorts of artefacts that you didn't see in the analogue signals of yesteryear.

But back to the original puzzlement: Looking at your tree pics it seems like the K20D has a lower DOF than the K10D.
Hey, that could be the solution...
If your lens is not the limiting element resolution-wise then the K20D, having 20 % more pixels in a line, has a smaller 'circle of confusion'. On second thought, it looks much more different in your sample pictures and after re-sizing that shouldn't be visible anymore... It remains a secret.

Prieni

[Ron Kruger]

I really was expecting sharper detail with the K10D, because I noticed it without testing, but the increased DOF was a real surprise. Both were shot at f-8 and the focus was centered on the tree. I am so surprised at this, I plan to do the test over, just to make sure I didn't do something to cause it. It is a really dramatic difference, and that doesn't make sense to me using the same lens and same f-stop.
If anything, I would have expected the higher MP (K20D) sensor to increase DOF, not the other way around, but this alone would make images much sharper from the K10D.

[Prieni]

Yes, sounds like it's worth to repeat that experiment. I doubt, though, that it will give a different outcome... We'll see.

Prieni

[spyglass]

I hope we are not confusing DOF with Contrast here ..... I do not see a difference an appreciable DOF between the K10 & K20 in the examples offered. What I do see is a significant contrast difference, an slight exposure difference, and a bit of a WB difference.  

Not sure what your exact test parameters are but I feel that manual focus should be employed for both shots, and full manual operation of all settings (same f stop, same shutter speed - all set to manual to ensure as level a playing field as possible).

If you look at the greenery shots I like the K20 version best. The K10 version looks a bit harsh (too much contrast)

Have you tried another version or piece of software? If you want you could redo your tests, burn the files to CD (copied directly from camera memory to CD), and send them both me & Tcom to run through a alternative pieces of software. It would explore and either find or extinguish this possible aspect in the overall question. Hell I will even pay for the CD's and shipping. What version of ACDsee are you currently running?

"We are going to drag this thing out into the bright light of day...... and beat it to death"  


Spyglass

[blackcloudbrew]

Honestly, unless someone is becoming upset with this thread, it's quite interesting and is making me want to go and do a similar test tossing in the K5 as well.

[Ron Kruger]

Spy: I used manual focus, AV, f-8. There was a difference in shutter speed on the weeds, but not on the tree. Auto WB on each. I was trying to see what the cameras produced, and while these results could be altered with PP tinkering, I tried to keep it as simple as possible, keeping processing to a minimum. To me this is not about processing, but photography.
While part of the reason I like the K10D over the K20D is the improved contrast and color richness (which also increases detail), that still does not explain the drastic difference in the DOF. Maybe an ant moved the tripod leg or something when I took the K20D shot. The earth's axis tilted--whaterver. Seems strange, so I'll try it again and we will go from there.

[Prieni]

I hope we are not confusing DOF with Contrast here ..... I do not see a difference an appreciable DOF between the K10 & K20 in the examples offered. What I do see is a significant contrast difference, an slight exposure difference, and a bit of a WB difference. 

Hmmm... I fear the confusion just has begun
Look at the first two crops and the areas left and right of the tree trunk. That is not difference in contrast.

[spyglass]

There is no way that the focal point and or aperture could be the same in the bark/tree shots. A focus difference of this degree would be a test error or major equipment malfunction. Unfortunately the original captures are no longer available to examine the EXIF information. T-man I do not want to ruffle your feathers on this but I think we need a more clinical test. If possible please include all EXIF data.

In regards to DOF and Contrast my comments were directed to the "greenery" plant shots (last 4 photos).  I find the K10 images to be a little harsh in terms of contrast. The K20 images seem more true to expectations. The last 4 images do not offer the same composition so comparing minute focus or details is not a viable.

As for AWB, I think this is one area where there might be a significant difference between the two models. I would be interested in seeing a test comparing the various white balance settings.

Spyglass

[Ron Kruger]

I've been crippled with gout, but I'll reshoot the test shots when I can. I, too, suspect something went wrong.
There is one thing I don't like as well about the K10D. It is very inconsistent with external flash, and I use flash a lot. I'm getting the job done, but it's taking many more exposures and attempts. Sometimes, it just doesn't make sense, but I suspect there is some quirk about it I haven't figured out yet.

[blackcloudbrew]

Actually, I'm glad you said that. I've had trouble with external flash use on mine too but thought it was just me. I do find that if I shoot flash in Tv mode and use the Ev compensation (for the flash) on the camera to control the flash, I get much more predictable results (after a bit of trial and error to get the exposure right). 

[Ron Kruger]

I experimented with various ways of using flash with my K20D, and settled upon AV mode, adjusting the apature according to the distance and ambiant light and became very consistent with it, but the K10D seems inconsistent at best.
I used to see a lot of complaints about Pentax flashes and believe they probably came from users of the K10D and the 360.
I'll try TV mode.

[calsan]

I jumped from the Canon rebel to the K10d.  The plastic look that Ron's referring to, a few posts back, may be the excessive noise reduction on the other systems, that often starts being applied from ISO 100.
At first, I thought the K10 had a lot of noise, as jpegs from the canon had NONE at any ISO setting.  But then I noticed that there was a lot more detail preserved in the K10's photos and realized that I'd been taking photos through an electronic 'Vaseline' on the Canon.

As for the flash, I always start out at f5.6 and then adjust either way after "chimping" the first shot. I don't find the K-7 any more predicable than the K-10.  I think it will get correct exposure in P mode though, but then you hand over control of your aperture to the camera, which I don't like.  I have also given High Speed sync in TV mode a trial run, and it seems like this might be worth further investigation.