SRS seems to have at least one of these in stock at the moment.

"Previously enjoyed" shouldn't be interpreted as quite the same as "second hand". Have a look at the SRS definition of "refurbished".

"Previously not enjoyed", would seem a bit more accurate, actually.

I can certainly attest to this.

I've had the K10D for a little over a year and a half now, and my "in-focus" hit rate is considerably higher than it was at the beginning.

I photograph dance events from time to time, and these are usually quite dark, with moving subjects. Possibly one of the worst AF tests possible, yet my results are hugely better than they were, even with the original 18-55.

I'm not even sure what I'm doing differently, although sometimes I notice myself instinctively pointing at higher contrast parts (faces, often) of the frame to focus, and reframing, sometimes only a little, to take the picture - all within a very short time.

Now it may be that other brands have better AF, and if I had one of them I wouldn't have had to develop as much skill to compensate for the electronics.

But fortunately I seem not to need the better electronics now.

Maybe the camera has just realised that I'm not going to give it an F2.8 16-50 or 50-135 to look through yet, so it had just better get used to seeing through the slower lenses - and shape up anyway.

I've never owned an SLR other than Pentax, so I can't comment on other brands.

However, I've read many people criticising Pentax AF, and all the reports I read compare focus speed (ie, time-to-lock) without also comparing focus accuracy.

Somehow it doesn't seem to me to be a valid thing to exclude from such comparisons.

Have you ensured that you're metering with the viewfinder covered up when you check the top LCD?

To test for back/forward focusing, the pictures of the test chart should be with the lens wide open, to make the depth of field as narrow as possible. Then you will clearly see where the actual focus point is.

Hmm. I think this is overstating the issues with Pentax AF.

Evenly lit, contrasty landscape?

There's nothing at all wrong with the K10D's AF in such conditions.

My understanding is that the K100D doesn't suffer the same stop-down metering issues that the K10D does.

I don't have one, though, so I can't confirm this.

Even on the K10D, if you meter with the lens wide open, and then calculate the shutter speed yourself for some other aperture you set with the aperture ring, it'll be fine.

Just less convenient, as you say.

What puzzles me is that it's just some lenses.

The 16-50 probably has the worst rep of all the current ones, followed, possibly by the 10-20 (albeit not necessarily in the same category). I haven't heard even a murmur against the 50-135.

Not at all including Simon's 10-20 experience in this, but in general I find it difficult to separate out the anecdotal garbage from the real worries.

Not here so much, but elsewhere, people rubbish a lens based on a couple of bad photographs, reduced down to 800 pixels, from which they believe they can tell anything at all about the lens' sharpness. It may be a problem lens, or it may not, but oftener than not there isn't enough information captured to allow a sensible conclusion.

Well it does a bit.

Either that or you've used nicely-balanced fill-in flash with a guide number of about 100,000.

But I prefer the restrained look too. I occasionally like impressionism, but in small doses, and in a kind of "hmm", sort of way.

________
PS... I'm curious as to why some of the sensor spots look like they have ghosts down and to the right a little.

Ah, so it's another new sky. Shows how much I notice...