AF is quick, though, on the FA, for the same reason.

I've been very pleased with mine, and not having an 18-250, I occasionally still take it out if I can only carry one lens.

Something else I just noticed...

Point a polariser at your computer screen (assuming it's an LCD monitor, and if it's a CP, ensure that you look through it in the same direction as the camera would).

Rotate the filter.

The screen goes almost completely black.

I never knew that the light coming off an LCD was polarised. But I guess it is.

Well, well... you learn something new every day. All this time I've thought that the exposure is supposed to change as you rotate the filter.

Naively, I always assumed that since less light was getting in as you approach the ideal angle, you'd need more exposure, but it seems that it's not supposed to be dependent on the angle of the (linearly) polarised light.

I've done a bit of digging, and this seems to be quite a good article.

Here's how I'm now understanding it - perhaps anyone that really understands the detail would tell me if I've now got this right:

- A LP allows only light through if it's polarised at a particular angle.
- Reflected light tends also to be polarised at a particular angle
- So if you rotate the polariser so that its 'allowed' angle is at right angles to the angle of polarisation of the reflected light, you cut out most of that reflected light.

So far so good, hopefully.

- The light meter measures the intensity of light which is itself polarised, since it's reflected off a partially reflecting mirror.
- If the light hitting the mirror is polarised only at one angle, then the intensity of the light hitting the meter will depend on what that angle was - if it's same, it'll be unchanged, but if it's at right angles, it'll be less.
- Since the sensor doesn't care what angle the polarised light hitting it is at, it's not subject to this variation in polarisation angle.
- So you can get variable metering results, for an amount of light that would be the same by the time it hit the sensor.

Ok. There's one thing that I still don't understand - why is the light coming off the mirror polarised? I thought reflections from metal surfaces retained whatever polarisation they already had. Is it because the mirror is only partially reflecting? If so, why does that make the difference?


Anyway, all very interesting though it all is, if the potential error is only a stop, as suggested in the Bob Atkins article, given all the other variables that contribute to getting the exposure right, I'm still in no hurry to go out and replace my LPs with CPs.

I've used my LPs ever since getting my K10D, and never once noticed a significant problem. Which isn't to say that it's not there, just that it's not large compared with the other things I get wrong

Tested what, specifically?

With the K10D, or with the later ones?

As I've posted on a number of occasions, first establish that you actually need a circular polariser.

You haven't said which camera you're using, but linear polarisers work perfectly well with the K10D, and maybe with the later Pentax dSLRs too.

They're a lot cheaper than the CPs.

For fast jets, even a monopod is no real use in my experience. Following them all over the sky, at the rate they move (especially in the vertical plane), on any kind of pod, is completely impractical.

If you're getting blurred results, it can be one or more of three things:

- Focus inaccuracy
- Movement blur
- Camera shake

Especially for the fast jets, you often need very fast shutter speeds to avoid movement blur unless you're panning smoothly on the aircraft in question. Take this shot, for instance:

[IMG]http://i162.photobucket.com/albums/t244/chris5gd/Aviation/Farnborough%202008/IMGP0627.jpg[/IMG]

You can see that 1/400s was plenty fast enough to freeze the one I was panning on, but nowhere near fast enough to freeze both, since panning effectively doubled the speed of the one going from left to right. I should think that blur would still have been evident at 1/4000s in that shot.

Camera shake won't be an issue at 1/1000s or shorter, even with long lenses, so you probably just need to use faster shutter speeds, and practise panning accurately.

I bought one of the original Sigma F4-5.6 EX DC 10-20s almost two years ago, and I've been very pleased with it. However, some have had difficulties getting a good copy, so a little caution may be warranted.

I've never wished it was any faster, and the extra 2mm at the wide end compared with the 12-24s is quite noticeable. There is significant barrelling at 10mm, which can be removed in software these days, of course, and goes away by 12mm.

At the time, it was only about £300 IIRC, which was a lot less than the Pentax 12-24. It was an obvious choice for me at the time, and I haven't regretted it.

With the price rises since then, and the availability of the other extreme wide angle zooms that have appeared since then, and the Samsung 12-24 being a lot cheaper than the ludicrous price of the Pentax one, the choice isn't so obvious any more.

I'd try and handle a couple of the alternatives if you can, especially if you have no experience of anything wider than 16mm.

The Pentax SMC one? It's great isn't it, although something of a beast. Were you on a tripod and pre-focused, or were you holding it?

Also, how do you get the caption into the EXIF?

It's the cloud shots that give it away. What a place to have your office, eh?

I love the Kelvin–Helmholtz one.

Certainly an example of amazing dedication.

It could easily be an artefact of any conversions and resizings that went on, but it's interesting to note that the pictures in the OP's link don't appear to be particularly sharp.

Especially the first one, of the two lionesses, appears quite soft to me.