I think that if the camera were clever enough to apply actual HDR algorithms to the set of exposures, so that shadows were given more, and highlights given less, they would be claiming so in the documentation, since it would be something of a selling point.

Although you may be seeing more shadow detail because of the fact that it's superimposed several shots, it clearly hasn't ignored the burned out highlights in the over exposed shots.

So I don't think it's at all likely to be doing anything at all in the the direction of actual HDR processing.

And if it did, it would potentially ruin the 'ordinary' use for multi-exposures, which is where the individual frames are different, not the same.

I reckon that thinking of the in-camera multi-exposure merge as HDR is barking completely up the wrong tree, interesting thing to try though it was.

Can you still not see it?

From the top right corner, it's dark blue, cyan, light blue, cyan, lighter blue still where it meets the horizon.

The bands are curved - like an arc of a rainbow centred above the scene - except not as many colours!

Come on folks, help me out here - surely I'm not the only one that can see it?

It looks very similar to the vignetting I've had with my 18-55.

But vignetting has the effect of underexposing the corners. So if the centre (of the sky in your case) is overexposed, which it is, there's got to be a point before it gets completely burned out where it's only partially so.

Hence the cyan band, which I have on several of my overexposed skies.

So if you have two shots where the exposure's different, the cyan band will happen in a different place.

And if you merge them, it doesn't seem too far-fetched to hypothesise that you'll get both bands, and there are definitely two.

Interesting experiment.

Interesting to note the banding in the sky. I can see two distinct cyan bands, which I presume is the clipping in the sensor that happens just before a part of the frame is completely burned out. Except that when several exposures are merged, the clipping happens in different places and you get them all.

Thanks John, very helpful.

At risk of muddying the waters slightly, how do the As compare with the FAs and DAs?

Specifically, the A-50 1.7 with the FA 50 1.4 at comparable apertures?

And the A-28 F2.8, say, with the DA* 16-50 at 28mm?

My priorities have been changing a bit recently, from long to fast, so I'm having a bit of a rethink.

Would the DA image circle still be Ok at 1.3?

I think, reading between the lines of what those allegedly in the know have been trying not to drop too many hints about, it's more likely to be a new model, at a significantly higher price point than the K10D.

K10D prices might creep down a little, but I can't see anywhere near as much movement as there's been this year (which has seen a £200+ drop).

But I know nothing. I'm only guessing.

If you're going to get a K10D, it might be £50 cheaper in four months time, but that's nothing compared with the value of having it four months earlier and getting used to it in time for your holiday.

Even this would be potentially useful if the tests were reliable and scientific, but this can be a long way from reality. He writes in a pseudo-scientific style, but don't let that fool you into assuming that what is being reported is all that scientific. It is certainly not impartial.

I think this is true, but in itself I don't think it's an issue. The brilliant creative types that can take a beautiful photo are not likely to be good at doing the boring drudgery of rigorous equipment testing on which we rely.

Sadly though, rigour is not a word that comes to mind in what I've seen of RH's work.

It stands for High Dynamic Range.

It's a cunning way of merging several bracketed exposures of the same thing, so the dark bits are more exposed (so you get the shadow detail) and the bright bits are less exposed (so you don't burn out the highlights).