[drerka]

... to tell you that I love the K20D, after having used the K10D for about two years. The price has come down a lot lately and that made me pull the trigger.

And here is a pic I love, showing the low light capability of my new K20:

1/8 sec, thanks to SR
iso 1600
F 1.8
FA 77 Limited




Greetings,
Roger.

[Clicker]

Yes, you're right but that is a cute puppy

[drerka]

Thanks, his name is Joep (Dutch short form of Joseph) and he's eight years old. A Blenheim Cavalier King Charles Spaniel, to be precise. And he's very adorable, and he knows it

Roger.

[Luciferase]

Well, lucky for U. I am still eagerly waiting the k20d to drop more before I can swing the bat.

It does seem the low light handling is great, the background at ISO1600 are almost free of noise as compare to the k10d.

[drerka]

You're right, iso 1600 is at least as usable as iso 800 on the K10D, imo. Even iso 3200 can be used if necessary. Above all I like the built quality, which has remained on a high level.

I hope the prices in Australia come down too in the near future, so you can have yours. A thing I noticed, is that a couple of European sellers are out of stock for the K20 body. Maybe the low price has atrtracted quite some buyers.

[Vormulac]

Look fantastic! I have a K10 and lust after a K20, one question: Does the K20 sport active dynamic range handling like recent offernings from Olympus and Nikon?

[drerka]

The K20D has a feauture called D-Range (if I remember correctly). It adjusts the tonecurve somewhat and offers expanded dynamic range. So yes, Pentax offers a similar dynamic range expansion, allthough I'm not very familiar with the Nikon and Olympus features on this point.

[Vormulac]

Ah, right. The dynamic system on the Nikon and Olympus adjusts the DR for different areas within the frame, so if one area is bright and one very dark (say for example a staircase in bright sunlight and the area under the stairs naturally very dark) instead of the whole frame being exposed low to avoid blowing out the stairs, resulting in detail free black in the dark area, the bright area is exposed low, and the dark area is exposed high, so you get a more eyeball-like overall scene, with detail in each area. At least that's what I've gleaned from a few reviews and a chat with an Olympus-toting tog. It sounds great.

[kenyee]

Yes, it does have D-Range, but that disables ISO100. It also shows more chroma noise in the shadows.
I think I get similar results tweaking RAW files, so I haven't used it much, but if you shoot JPG a lot, it's useful.

I think ISO up too 1100 is ok but after that, it's noisy. Indoors w/ flash, I usually set my K10D to autoiso400....w/ the K20D, I set it to autoiso800...

[Marc Sabatella]

Ah, right. The dynamic system on the Nikon and Olympus adjusts the DR for different areas within the frame, so if one area is bright and one very dark (say for example a staircase in bright sunlight and the area under the stairs naturally very dark) instead of the whole frame being exposed low to avoid blowing out the stairs, resulting in detail free black in the dark area, the bright area is exposed low, and the dark area is exposed high, so you get a more eyeball-like overall scene, with detail in each area. At least that's what I've gleaned from a few reviews and a chat with an Olympus-toting tog. It sounds great.

That's a reasonable description of the effect of Pentax system too. Of course, none of these systems *actually* expose different parts of the scene differently - the whole sensor is exposed for the same amount of time, and with the same sensitivity. It's physically impossible without a complete redesign of how the shutter works for any camera to expose different parts of a scene differently. But the exposure curves can be tweaked in the in-camera processing to bring out detail in the shadows without clipping the highlights. In other words, precisely the same thing you might do yourself anyhow in PP via "Curves" or "Local Contrast Enhancement", except you turn control over the process over to the camera. Useful if you don't shoot RAW and aren't comfortable with PP.

[Vormulac]

Hi Marc,

I'm sure you know better than me (most people do, I'm very new to all this), but if the scene has been metered by the camera for the light levels, wouldn't the microprocessor be able to tell the sensor that this area over here is reading very bright, so underexpose that bit, and this bit here is dark, so up the gain there etc effectively meaning the sensor is utilising variable sensitivity across its area and not just shifting levels in PP?

I could well be talking out of my hat here, I'm just trying to think through the process logically. Anyway, I still want a K20D

[Marc Sabatella]

I'm sure you know better than me (most people do, I'm very new to all this), but if the scene has been metered by the camera for the light levels, wouldn't the microprocessor be able to tell the sensor that this area over here is reading very bright, so underexpose that bit, and this bit here is dark, so up the gain there etc effectively meaning the sensor is utilising variable sensitivity across its area and not just shifting levels in PP?

Actually, you're probably right - it *could* be done by pumping the gain in the shadows before the a/d conversion, *if* you had a separate processor that read the voltage levels to figure out which pixels *needed* the extra gain, and designed so your sensor so it was *possible* to adjust gain on a pixel-by-pixel basis. I'm not sure why they'd bother introducing a whole new component (the processor that determined which pixels to boost) and potentially a new sensor design (one in which pixels can be boosted by individually different amounts) to do that, when doing it after conversion is so much easier. It's also not clear that it would actually actually work much if any better.

EDIT: I realize you mention using the meter inthis process. Well, the meter is pretty coarse, dividing a scene up into 16 segments only. That's far too inexact to use in an operation like this - each segment might well contains shadwos as well as highlights. Results of boosting pixels on a segment-by-segment basis would be far worse than doing it pixel-by-pixel, even if the pixel-by-pixel processing is done post a/d conversion.

Anyhow, regardless of what might or might not be technically feasible, no one does this - they all do it post a/d conversion - and usually only in the JPEG processing. I haven't tried it with Pentax, but I know the Nikon version doesn't affect the RAW data at all, except in that the RAW data reflects the underexposure that is used to prevent highlight clipping in the first place.

[Vormulac]

Fair enough, Marc, I haven't used this function only heard about it so I didn't realise it was only the jpgs effected. I wouldn't be of much use to me then as I tend to shoot in RAW anyway. Interesting stuff though

[drerka]

AFAIK, the Pentax D-Range function *does* work with raw, so it does have its benefits.

[Marc Sabatella]

AFAIK, the Pentax D-Range function *does* work with raw, so it does have its benefits.

I suppose I should test this rather than guess, but since I know this to be the case with Nikon, I'll at leas offer it as a conjecture:

What may be happening is that D-range is like other in-camera settings - it doesn't actually *affect* the RAW data, but the fact that you used it may be recorded in the file for use by a RAW converter. So just as cranking up the saturation doesn't affect the RAW data, but *will* cause PPL to crank up the saturation by default when it displays / converts your image, it may be that PPL is simply noticing that you set D-range and reproducing it's effect internally. If you, you'd find the image looking correct in PPL but "wrong" - dark, basically - in most other RAW processors. If you're inclined to test this, be sure the other RAW converter really is processing the data and not just displaying the preview.

As I said, I *know* this is the case with Nikon - a Nikon image produced with D-lighting appears dark in any RAW processor other than one specifically designed to reproduce that effect itself (and as far as I know, Nikon's own software is the only one that does). I'd be kind of surprised if that didn't turn out to be true with Pentax also. But I've been known to guess wrong before.

In any case, what's actually happening is not really in question here - the camera processing is simply fiddling with curves after the a/d conversion, just as is the case with Nikon and the others. The only question is whether it does this in its RAW processing or just for JPEG.