Originally Posted by zac510
I feared this might be something to do with my addiction to my polarising filter..
Is it a linear of circular polarizer? (That refers to the way it polarizes not the shape of the filter. Just in case you were not aware that there was a difference.)
I spent some time yesterday with my elderly Canon Pro1 and a circular polarizer and there are certain observable differences in aaent noise levels for shots taken when the sun went in.
Now the problem is likely that the poor old Pro1, great though it can be, is not exactly the latest technology and is naturally quite noisy in some situations. The polariser will cost it somewhere between one and two stops of exposure and the metering and focus systems are probably compromised as well, especially if the light dips as a cloud passes.
On a Pro1 compensating by upping the ISO is not really an option, though as andrewc has pointed out, at typical print sizes (and provided one is not cropping the original too much before printing) you should be able to create something where the noise does not show. Unless you have had to go for an extreme adjustment somewhere along the way, as I mentioned before.
The circ-pol on my Canon 400D + 70-300 DO lens also introduced some dubious results, notably on motor sport focus matters, but then that would not be a surprise really given that the focus systems like to have something like at least f5.6 worth of light and that is the DO lens at full stretch BEFORE the filter is taken into account.
My conclusion to date is that the safest way to use a polarizer and get predictable results is going to be to set it up for a single shot and manually focus if necessary. That way you get consistent polarization effects and a known focus point and save some battery power as an added bonus.
Or it could be that my Cokin filter is not very good ...