Where is the mark on the polarizer supposed to be?

Started 2 months ago | Questions thread
petrochemist Senior Member • Posts: 2,637
Re: Some technical info on polarization.

ThrillaMozilla wrote:

petrochemist wrote:

ThrillaMozilla wrote:

Gerry Winterbourne wrote:

No filter removes all polarised light.

In the right orientation they do remove virtually all polarized light. Photographic polarizing filters are almost perfect in that respect. (Footnote: someone once tested a horrible, cheap filter that transmitted a few percent of polarized light, but that was an exception.)

However, most light that you see in nature is not 100% polarized.

Virtually all is IMO a rather unhelpful term.

Note that this is the Beginners Forum, and the OP seems like a very basic question. Hence the basic answer. But since no good deed goes unpunished, check out some extinction measurements:

Polarizing fillter transmission and extinction tests

The small amounts of transmission occur at the extreme ends of the visible spectrum, where the sensitivity of the sensors is at a minimum.

For those who want a simple statement, all you really need are Roger Cicala's words:

If you are buying a circular polarizing filter because you want some circular polarizing, it doesn’t seem to matter much which one you choose; they all polarize like gangbusters.

I'll check out your links later. I think I've already read the lens rentals one.

Update (after checking the links) I think I'd actually read both of these some time in the past. The extinction tests are better than i can achieve with the spectrometers at work which have unknown polarization (I know both sample & reference beams are at least partially polarized) Like my results with crossed polarizers they show significant transmission of the 'blocked' polarization above 700nm, some show like the Fomei Digital CPL quite a bit in the 400-500nm region too - I note the passed piolarization has greater transmssion in this case so it may be a high transmission type

In some cases 0.002% transmission could be significant, in many the few percent transmitted by that cheap filter will not matter.

Considering that the sky is what, maybe 30-50% polarized at most, and very few surfaces are conveniently arranged at the Brewster's angle, are you really worried about 0.002% transmission? No filter at all will give you 100.000% transmission, so I'd consider 0.002% pretty good. That's 15.6 stops. Keep in mind, this is the Beginners' Forum.

IMO the sky is one of the least useful application of polarisers. perhaps that just that the sky's so often overcast when I'm out with the camera. I like to use them either for reducing reflections or on occasion for boosting them (the second application being missing from many articles on polarizers)

Most of the time 5% transmission of the blocked polarization wouldn't bother me at all, but if I'm trying to block a reflection of the sun & shoot into a darkened interior 0.002% reflection would still be a pain.

When reflections are an issue the photographer can often change their position to get the reflections at Brewster's angle. At least if they know it will be useful.

Polarizers will indeed remove the bulk of visual light that is polarized in the opposite orientation, but very few are effective with Near infrared which many digital cameras have some sensitivity to.

The question is not about how to implement IR photography. Photographic polarizing filters are intended for visible light use. I don't know what kind of cameras you are using, but as far as I know, most consumer cameras have IR blocking filters. Older video cameras are an exception.

Check it out here.

Yes consumer cameras have IR blocking filters, but the strength of these vary considerably, my old K100d was capable of hand held IR shots with a fast lens & the ISO turned up. I believe the Nikon D70 is very similar.

The Panasonic G1 I had was less sensitive to IR but could manage higher ISO so also managed handheld IR when I tried it. I've not really tested many of my newer bodies for IR having got several that have been 'full spectrum' modified for IR. I have however shot long exposure IR with my G5, specifically making use of it's lesser IR response to get a 60s exposure. I don't shoot video

You can buy polarizers that are effective over the entire wavelength range silicon sensor can record, sadly they are quite expensive so I haven't got any.

I don't know of any natural source where light is 100% polarized & I think lasers may be the only man made sources that actually achieve 100% (even the very best polarizers leak a few ppm of the polarization they are designed to block). Most natural sources have photons of random polarization but fortunately this can be treated as vectors the proportion in the direction of the polarizer being blocked & the remainder passed. Refection etc will only polarize a portion of the reflected light with the maximum portion when the light is incident at the materials brewster angle (63 degrees for water IIRC).

I'm trying not to be sarcastic here, but keep in mind, this is the Beginners Forum.

We were all beginners once, but not everyone who reads the beginners forum is a beginner.. Some beginners will have a good grasp of technical aspects, any that aren't interested are free to ignore any parts of my post & the title gives them warning.so they can skip it without effort.

Using two crossed normal polarisers gives a fairly good indication of how much light of each polarization is removed it not hard to get down to around 0.5% transmission in the visual with this approach. With all of my polarisers this raises to around 50% transmission by about 750nm - easily seen with several of my unmodified cameras and not dissimilar to a R72 filter on my modified bodies.

That sounds high for a modern photographic filter, and even higher for a camera. Check the references above. Can you make a case for inadequate extinction for one of those reviewed filters in ordinary photographic use?

The figure quoted was from memory, checking the data directly it was less than I quoted but it did start to increase before 750nm

From the actual data: Visual transmission below 700nm was consistently just showing noise (<0.02%) yet transmission had reached 1% at 730nm, and was at 20% by 800nm. 50% transmission wasn't reached til 843nm, by 900nm it had reached over 70% - My unmodified K100d is certainly capable of recording these wavelengths as it's hot mirrors transmission tails out to ~1000nm.

Those results were from standard photographic filters (perhaps from film days so not necessarily modern) They were measured on research grade spectrometers at work, and I was surprised how soon after the visual band the transmission started.

I made no reference to how much light at this wavelength a camera can see - it varies significantly by model but is unlikely to be more than 5% (unless the camera is converted.) from the data I've seen, i don't have any way of measuring that myself.

 petrochemist's gear list:petrochemist's gear list
Pentax K100D Sigma SD14 Pentax K-7 Panasonic Lumix DMC-GF2 Panasonic Lumix DMC-G5 +15 more
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow