Olympus-OM
[Top] [All Lists]

Re: [OM] A Modest Proposal I [was Digicam for the ages]

Subject: Re: [OM] A Modest Proposal I [was Digicam for the ages]
From: Jez Cunningham <jez@xxxxxxxxxxxxxx>
Date: Thu, 27 Aug 2009 15:18:34 +0200
I like logic.  Good 'article'.
But:

- Is it just partially-silvered mirrors - I don't think so.

- I think there's another topic to be considered re: polarizers and it's the
effect on autofocus (if it's phase-based which also uses light reflected off
the mirror to AF sensors and not contrast-based from the main sensor).
- Also another reason for using a pol is not just specular reflection but
reflection off any non-metallic shiny surface - like shiny leaves.  Makes a
big difference that can't really be recovered in post.

br
jez

On Thu, Aug 27, 2009 at 12:05 AM, Moose <olymoose@xxxxxxxxx> wrote:

> A Modest Proposal I
> Being a recommendation that logic, however flawed a tool in many human
> eneavours, may be usefully applied to technical problems in photography.
>
> 1. Polarizing filters have been used in photography for many decades to
> achieve certain desirable effects.
>
> 2. The advent of TTL metering using partially silvered SLR mirrors led
> to a problem with polarizers. Light passing through the partially
> silvered mirrors is itself polarized. That is, only one plane of
> polarization passes through the mirror, with the rest being reflected.
> The simple result is that placing a polarizer in front of the lens will
> result in metering errors, the amount varying with the angle of the
> polarizer.
>
> 3. The solution was a so-called circular polarizing filter, which has a
> second active layer that 'unpolarizes' the light that has assed through
> the primary filter. Works great for metering and there are no complaints
> during the film era.
>
> 4. Digital sensors largely replace film. Users complain that polarizers
> give subtle, but annoying problems with their digital cameras. It turns
> out that the 1/4 wave plates used to re-scramble polarized light aren't
> linear with wavelength, so the effect varies with color. Didn't bother
> film, causes problems with digital.
>
> So what's the solution? As Mike G pointed out, it appears some filters
> work better than others, the expensive ones being mostly better.
>
> 1. So one solution is filter testing and purchase of new filters, not an
> inconsiderable expenditure of time and money.
>
> The next three solutions differ in that they take advantage of the
> advent of digital capture and processing, rather than assuming nothing
> has changed other than a new weakness. Remember, the problem is with
> metering, not polarizer and film/sensor - AND that's an area where
> digital is VERY different!
>
> Unlike film, one may determine instantly whether an exposure is accurate
> on a DSLR. Hmmmmm. So why not attack the problem where it lives, with
> exposure, using those free, old, LINEAR polarizers we still have lying
> around?
>
> 2. What if I simply take some test shots to determine the angle of the
> linear polarizer with the least effect on metering and mark that angle
> on the polarizer. Then I start any shot with polarizer at that position,
> set manual exposure, and rotate to my heart's content.
>
> That may seem wrong at first, but think it through. Rotation affects all
> unpolarized light the same regardless of the setting of the filter. The
> whole point of using the filter is to vary the balance between polarized
> and polarized light reflected from the subject(s). So the only change to
> the basic exposure as the filter is rotated is to the elements intended
> to become darker.
>
> 3. I can do a more through testing to determine the compensation to the
> meter reading needed at marked angles of polarizer rotation. I could
> even mark the polarizer in stops, rather than degrees. This should get
> exposure close with the first shot. Then, remember chimping and
> histograms? Check exposure and adjust as necessary.
>
> 4. Oly started it and now many have followed - Live View. Well guess
> what? The mirror is up out of the light path during LV, and metering is
> done using the image sensor itself, not some separate thingie through a
> polarizing mirror. LV metering is unaffected by linear vs. circular
> polarizers*.
>
> Last, but not least, for those of us relatively adept in the digital
> darkroom:
>
> 5. Eschew polarizers for everything but "seeing through" reflective
> surfaces like glass and water and perhaps shots with large amounts of
> specular reflections. Those are the only effects that aren't easily
> duplicated in an image editor.
>
> One of the big uses of polarizers is for darkening skis and bringing out
> clouds. This is easily done in post, and actually better in most cases
> with WA and super WA lenses, where the polarizer effect varies across
> the sky.  The increases in saturation and mid-large scale contrast are
> also easily duplicated in editing.
>
> Moose
> * An assumption on my part, as I don't have a LV DSLR to test it on. I
> can't see why it wouldn't be true, though. I'm not sure testing on a
> non-DLSR digicam would convince some people. Somebody with a LV DSLR up
> for a test?
>
> Ken Norton wrote:
> > Rambling Joel wrote:
> >
> >> ... I was actually kind of disappointed with a couple shots on digital
> with PL that I think will be OK on film with PL.  I just think that's kind
> of weird, but I guess we have to remember that a sensor is not the same as
> film.  In practical terms, I'm still learning what that amounts to.  I see
> it working in a more nuanced (i.e., trickier) way with ND grad split filters
> as well as PLs.
> > I'm equally dismayed with my results of polarized digital images.  It's
> something that I've fought and remain flummoxed. Maybe we can figure this
> out this weekend.
> >
>
> --
> _________________________________________________________________
> Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
> Archives: http://lists.thomasclausen.net/mailman/private/olympus/
> Themed Olympus Photo Exhibition: http://www.tope.nl/
>
>
-- 
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz