Will it do very low frequencies? At a guess you'd need less than about 10 Hz.

Nice one, Don. That was the kind of precision engineering I had in mind

Not, sadly, that I have either of those.

I've been following this thread with some interest.

I can think of a way of testing it, but not without additional equipment.

What it needs is the camera on a shaking platform, where the amount of shake is consistent, at a level where pictures taken with stabilised SR are obviously more shake-free than SR-off pictures taken on the same platform.

Then all that's needed is to take pictures with the camera on the shaking platform, with SR on, but before 'SR-lock' has been obtained.

Then compare the pictures with the two control sets to see if they're better, worse or neither.

The amount of shake from the platform would ideally be large compared with the amount of finger-induced shake (to allow influences from the latter to be discounted), but not so large that stabilised SR can no longer cope.

I suppose in the absence of the shaking platform, maybe sitting the camera on a washing machine (on something soft, probably) would do if it was quite a shaky one (possibly on the wash cycle rather than the spin cycle if the shake frequency in the latter case is too high).

Then if the pre-SR-lock results are grossly different from both the SR-off and the SR-lock cases, it may still be possible to draw some useful inferences from the experiment. But in the absence of the precision engineering solution that Don might provide, proving that the amount of shake was consistent enough might be difficult.

If it's the same quiet "slide thunk" as you get with the K10D, then the noise is the sound of the sensor as it flops from one side of its unrestrained travel to the other.

I investigated this, with the lens off and the mirror up, and watched as the sensor flopped to the side, making the slide-thunk noise as it did so.

No longer any question about this, AFAIC.

As others have said, it's completely normal.

I recommend leaving the manual in the bathroom.

And a reduced intake of dietary fibre.

For someone called 'laxy', this should be all you need to have a bit more time for essential reading.

In this context I found this short and quite readable article interesting - I hadn't previously understood the difference between longitudinal and lateral CA. And longitudinal CA can look exactly like purple fringing.

I don't really know enough to judge the arguments in it (and in ref 16) against PF being caused by the sensor, but they seem plausible at my level of knowledge.

I do agree with this. I'm not trying to put off anyone that's happy with the trial and, if not error, shall we say refinement , approach.

But forewarned is forearmed, as it were. Always good to be able to make informed decisions about purchases.

I'm exactly the same. I had a Super A and later an ME Super (that I preferred, strangely), since the mid 80s, and no other SLR till the K10D.

As I say, I was initially very disappointed, but my experience is that a small mindset adjustment is all that it takes.

I don't miss the aperture ring at all now. And my photos are a lot better than they ever were with film.

It is probably obvious, but bears repeating, that of course, even A-series lenses have to be used set to A, to avoid the overexposure errors.

If they're not set to A, they behave just the same as M lenses.

Using the aperture ring to set the aperture is the issue, because then the camera cannot find out what the aperture is, so cannot meter wide open and work out what the stopped down shutter speed should be.

If this is true, then you're one of the very few people with a K10D that don't experience the progressive overexposure with M- and K- series lenses as you stop down, using the green button in manual to make it use stop-down-metering.

Most people's experience is that stop-down-metering appears to work, but gets the exposure progressively more wrong as the aperture gets smaller (F-number gets larger).


I bought the K10D largely because of the K-mount compatibility - I had several M series lenses that I was hoping to use.

After the initial disappointment of discovering the now-well-documented issue with stop-down-metering, I decided that rather than persevering, I'd embrace the whole newer lenses thing, and stick with a couple of As, but mostly the later AF lenses.

It's not a move I've regretted.