You say this is 'fact', but do you have an authority to cite for it?

I've now tested two K10Ds, bought at different times and places, both apparently functioning normally, and they both exhibit the thunk during long exposures, but only if SR is on.

Clearly, unless they are so rushed off their feet that they have to resort to such fob-off techniques*, the Pentax support department is run by morons.

________
* to get rid of the undoubted 75% or more that don't realise you have to actually charge the battery before it will work

Update:

I've just popped over and tried out the same experiment with my Dad's K10D, and the behaviour seems the same as mine.

Well you say it's no mystery, but CallumJ seems to be getting 'no clunk' during any exposure long enough to do the experiment, SR on or off. Whereas, I seem to be getting 'no clunk' only during exposures with SR off.

There has to be a reason for this difference, and without knowing what's supposed to be happening, it's impossible to know whether to be concerned, or whether one if us is just not being careful enough with our observations.

This is my point: without specific, accurate information, it's impossible to tell whether the camera is working correctly, and we are left either floundering around trying to deduce what should be happening from what appears to be happening, or resorting to "well does it seem to be producing good photos?" and muddling on regardless.

This is an example of how hard it is to get a definitive answer.

Power off? I don't have to have my camera switched off to get the noise, and neither, it would seem, does CallumJ.

Here we have two apparently identical, correctly functioning cameras (which firmware level is yours at, Callumj? Mine's at 1.3.), one which exhibits the behaviour at all times except during an exposure, and one which exhibits it at all times except during an exposure with SR off.

At best, "sensor moving when the power is off" is only a partial answer. Although I'm convinced now, that it is indeed the sensor.

Not really worrying (it's yours that's bust, mate )

Joking apart, it is strange if there's genuinely different behaviour going on.

The trouble is, getting definitive answers from Pentax about anything seems to be a lot harder than it should be. So I think people tend to try and work it out themselves.

If you get any joy from them, let us know

You mentioned rocking from horizontal to vertical.

I'm rotating mine from more than vertical one way, to more than vertical the other way, so more than 180 degrees.

The thunk happens right at the extremes.

With SR off, it doesn't at all. I'm in a place quiet enough to be able to hear the thunk as well as feel it.

I just tried this... during a 10s exposure, rocked the camera from portrait one way to portrait the other. Got the same slide thunk. However, I failed to get it with SR switched off, which I hadn't tried before.

Whatever the actual mechanics, I think this confirms that it's the sensor moving. Nice one.

It's perfectly normal, as is alarmed new K10D owners asking about it

Note to the not-so-new...

Do we yet have a definitive answer as to whether it's the sensor shifting or the orientation sensor?

I keep imagining that I recall K100D owners reporting the same quiet 'slide-thunk', and that the K100D doesn't have an orientation sensor... is this true?

If you think it's back-focusing, stop taking pictures out in the field, and photograph the test chart, available free here, under controlled conditions.

This is one of the easiest things to test, since it is very easy to eliminate technique errors.