Linear polarizing filter with canon t-90 in center weighted mode

[Ricardo Miranda]

Go out and and actually take photographs?

 

[MattKing]

If the image is fine the metering is fine too, it seems logical to me even though I noticed that some member s here logic doesn't apply, in a book I have on the T90 the author specifically states p-l filters should not be used because they would "fool" the lightmeter, then if I had to challenge that statement I'd shoot one or two shots, if they come out well exposed these filters can be used, that's it.

One of the problems with this approach is that there is a great likelihood that the problem with using a linear polarizer with a camera that incorporates a beam splitter in the metering path, is that it will yield inconsistent results - correct when the light and/or the positioning of the polarizer are one way, while incorrect when the light and/or the positioning of the polarizer are another way.

There may be a limited set of circumstances where the linear polarizer would not cause problems, but that set of circumstances may vary with the polarization of the light.

 

[cuthbert]

Go out and and actually take photographs?

See? Finally some common sense from a FSU shooter!

Canonikonisti have a too convoluted mind...you want to see if it's a problem? Get out at shoot a test roll!

 

[Ricardo Miranda]

See? Finally some common sense from a FSU shooter!

Canonikonisti have a too convoluted mind...you want to see if it's a problem? Get out at shoot a test roll!

Lol
Yeap.
As Matt said, there might be conditions where a linear polariser will work, others that it won't.
But, a test roll will tell what conditions you can rely on it. It might also be the case that the wrong exposure might actually be the right one for the photographer. I had years ago a Spotmatic that was overexposing, especially noted on slides, but on certain films it was the right exposure for me.

 

[cooltouch]

Anybody here remember their high school or college physics? If you do, you might recall how a polarizer works -- or at least the area in which polarization occurs in relation to a light source. Let's use the sun as an example, since this will almost always be the light source when one is using a polarizer. You can determine the area where polarization occurs by using your hand. Hold your hand out before you perpendicular to the ground, i.e., vertically. Now, point your thumb directly at the sun with it stretched out so your thumb and fingers are at right angles to each other. As you're pointing your thumb at the sun, pivot your hand from side to side. Your fingers will point to the area of the sky in which polarization occurs.

Now, back when I learned this rule, there was no such thing as a circular polarizer, so it is safe to assume that this rule applies to a linear polarizer.

As an example to this, I noticed a few things when I used a linear polarizer with my old F-1. First of all, I would see the meter's needle plunge even though I didn't really detect much of a difference in the changing of the light value when I rotated the polarizer. And when I got my slides back I'd notice odd things like half the sky being a very saturated blue color with the other half being much lighter. After pondering these oddities for a bit, I recalled this hand-rule to polarized light and I began to apply it to my photography. Knowing where the polarized sky would be, I was then able to do a better job of composition to avoid these uneven areas of saturation. And as far as the dip in the meter that occurred -- to be honest I don't recall anymore what I did to resolve this. Bracketing, probably.

The above discussion in the previous posts is mostly about how the meter reacts to the use of a linear polarizer, but a linear polarizer is gonna polarize light just the way it always has, regardless of what the meter thinks. Now, as far as a circular polarizer goes, I honestly don't know. I've never used one. I just know that with modern metering systems it works better. Or rather the meters aren't fooled by it. But I've often wondered if a circular polarizer works as effectively as a linear polarizer does.

I own a T90 and I have an assortment of polarizers I've picked up over the years from buying outfits. I probably have at least one CP in my stack of polarizers. So I guess what I'll do one of these days, just to satisfy my own curiosity, is I'll try both LP and CP on a given subject and see for myself what happens. I reckon I should be able to tell quite a bit just by watching how the meter behaves, though.

 

[AgX]

Cooltouch,
the Canon T-90 has 3 sensors at two locations. The light beams to the two locations are treated differently and this difference is the reason for this whole discussion.

 

[GRHazelton]

One of the problems with this approach is that there is a great likelihood that the problem with using a linear polarizer with a camera that incorporates a beam splitter in the metering path, is that it will yield inconsistent results - correct when the light and/or the positioning of the polarizer are one way, while incorrect when the light and/or the positioning of the polarizer are another way.

There may be a limited set of circumstances where the linear polarizer would not cause problems, but that set of circumstances may vary with the polarization of the light.

When I began shooting with a TTL metering SLR, I'd read of the possible problems, so I just bought a circular polarizer. Problem solved. CPs aren't prohibitively expensive. Just do it!

 

[AgX]

At least in theory a circular polarizer means a further image degadation related to a linear one. There are other issues involved too.

But I doubt the all are significant or even visible in practical photography.

 

[cooltouch]

Cooltouch,
the Canon T-90 has 3 sensors at two locations. The light beams to the two locations are treated differently and this difference is the reason for this whole discussion.

I understand this, but ultimately it's about the amount of light hitting the film plane. As I mentioned before, I wanted to give it a try with my T90 -- using both a linear and a circular polarizer, just to see how the camera behaves with each. I just performed a quick little experiment, and here's what I found.

For this little experiment, I chose a non-metallic subject with a nice amount of glare on it. I chose the roof of our neighbor's car, over half of which was reflecting the sky and surrounding trees. The car is painted black. I pointed the T90 out of my upstairs window at the car's roof to meter things. I had the T90 set to Tv mode, Spot metering, with the lens set to "A". I have some ISO 1600 film loaded in the camera, so keep this in mind with the values I mention.

With the LP, I got a reading of 1/125 @ f/11 with the polarizer rotated so that it was not affecting the scene, that is, maximum reflections were seen on the car's roof. When I rotated the polarizer to its maximum effectiveness where no reflections were visible on the black expanse of the roof, the meter read 1/125 @ f/16.Yep, that's definitely a strange reading.

With the C-PL in place, I got a reading of 1/125 @ f/19 with the polarizer rotated so that it was not affecting the scene, that is, maximum reflections were seen on the car's roof. When I rotated the polarizer to its maximum effectiveness where no reflections were visible, the meter read 1/125 @ f/11.

The C-PL definitely provides the more convincing reading. In fact, the L-PL's reading is nonsense. It shows that the meter should open up the lens when the scene is brighter and to stop the lens down further when less light is entering the camera. So all it takes is for a person to be aware of the light values being recorded to see that the L-PL is causing the meter to give a nonsense reading. Overexposure for a bright scene and underexposure for a dimmer scene.

I was originally thinking that, if I knew what an L-PL did to the meter, I might be able to use one on my T90 anyway, but after performing this little experiment, I am now convinced that I should not even think about using the L-PL when using Spot metering, since its readings are far enough off to ruin exposures with slide film, and would provide barely passable if not substandard results with print film.

 

[AgX]

Cooltouch,

You still did not get the point (different sensor locations). You still did not try the "Spot" metering versus the other modes.


However, in a complete different set-up you did something different:
comparing the two different filters against each other.

 

[cuthbert]

AgX, Mike the Cooltouch, let's resolve the question by real photographers: get out and take some pics with a T90 and a P-L filter.

Then duel at high noon on the board.

 

[cooltouch]

Cooltouch,

You still did not get the point (different sensor locations). You still did not try the "Spot" metering versus the other modes.


However, in a complete different set-up you did something different:
comparing the two different filters against each other.

It seems to me that comparing the two filters is the entire point of this whole dialog. Which works and which doesn't and under which conditions.

In retrospect, I realize I should have tried both Spot and Partial (Averaging and Partial should agree in principal if not by the same exact amount), but after reading your comment, I took another look at my T90, and I realize I came to an incorrect conclusion. I thought I'd set my T90 to Partial Metering, but when I picked it up just now to take a look at it, I saw that it was set to Spot. Crap. So I did a quick test again, this time with the meter set to Partial, and I got completely different results.

I didn't try Averaging, but since it uses the same silicon photo cell as Partial does, it should give similar results. So anyway, when you read my note above, just substitute "Spot" for Partial, ok? In fact, I think I'll change it just so people won't come to the wrong conclusion.

So anyway, when I set my T90 to Partial and metered a subject with a reflective surface, it didn't matter whether I used a linear polarizer or a circular polarizer. Both gave me identical values. My neighbor's car was gone so I couldn't meter off it anymore. Instead I angled this laptop in such a fashion that it cast a pretty good reflection of available light. With both polarizers I was getting 1/125 @ f/4.0 when the polarizers let through a maximum amount of glare and 1/125 @ f/1.8 when they choked down the glare to virtual nonexistence.

So, in conclusion, I'm gonna state that the spot meter, because of its extra elements and sub-mirror, requires a C-PL, but if you're doing plain old Averaging or Partial, it sure looks to me like it just doesn't make any difference. Sorry about the misunderstanding.

You know, in the scientific community, whenever a hypothesis with empirical data is presented and a conclusion is drawn from that data, it is customary for it to be peer reviewed. In other words, it would be nice if somebody else here would take the time to confirm or refute my results.

 

[wiltw]

No, the light that comes through the polarizing filter that is reflected up off the mirror from whatever sensor has to go through the beam splitterhttp://en.wikipedia.org/wiki/Beam_splitter located at the bottom of the focusing screen that effects the meter reading which is why the T90 according to the instruction manual needs a circular polarizing filter not a linear one to meter correctly as it also does with the New Canon F1.

Sorry Benji, this is not correct, the T90 beam splitter is NOT at 'the bottom of the focusing screen' as you state, but the reflex mirror is semi silvered in one section, allowing a secondary mirror to reflect light downwards to the spot meter photosensor in the base of the camera (like in my OM-4). Here is a drawing of the T90 metering in various modes, and you can see what I have described. In other modes, the light merely bounces from the reflex mirror to the focusing screen and photodiodes read the screen (my OM-1 is like that, and it uses a linear polarizer). The T90 adds the Partial and Centerweighted capability not present in my OM-1.

http://www.google.com/imgres?imgurl...zVdisIsyzoQTGnoPgAQ&tbm=isch&ved=0CFkQMygyMDI

In the OM-4 user manual Olympus very specifically mentions the impact of linear polarizing on spot meter accuracy (because of the semisilvered spot transparency of the reflex mirror), but gives no cautions about other metering modes.

Cooltouch's results posting in post 37 reflects what the camera design characteristics would have one anticipate about meter accuracy in different modes with linear vs circular polarizers.

 

[kingbuzzie]

And a newbie triggers a four page thread! Thanks for confirming. I will probably pick up a circular eventually... but good to know it's not needed if I don't use the spot.