Will a Linear polorazing filter allow metering and AF on mirrorless.

[monte12345]

I expect that circular became so essential was due to the use of semi silvered mirrors in a DSLR for behind the mirror metering and AF modules. Mirrorless cameras don't use these features so can the older linear filters be used. One immediately noticeable feature of the circular polarizing filters I have are lighter in terms of light loss and just don't seem to be as effective as the linear filters I had when I was shooting with an F2Sb.

 

[alcelc]

No reason a mirrorless can't do metering and AF properly with a linear filter, or any filter.

I don't use linear filter, CPLs only.

As per my understanding on the working principal of mirrorless: metering and AF is working on the sensor data.

A good example is on the Auto Exposure system of my M43 camera: G1 which is the first MILC. I adapted vintage film lenses on it with a dummy adapter. Upon the actual amount of light (through the lens) sensed by the sensor, the AE system of G1 is able to suggest the required shutter speed and ISO for A mode shooting.

As long as the sensor can see any light, it can meter for the AE system to work. If the amount of light passing through the filter could meet with the AF requirement (some can do darker than others), no reason AF couldn't be used.

 

[ThrillaMozilla]

I expect that circular became so essential was due to the use of semi silvered mirrors in a DSLR for behind the mirror metering and AF modules. Mirrorless cameras don't use these features so can the older linear filters be used. One immediately noticeable feature of the circular polarizing filters I have are lighter in terms of light loss and just don't seem to be as effective as the linear filters I had when I was shooting with an F2Sb.

Where shall I begin?

1. You should be able to meter all right, but I wouldn't use a linear polarizer on most digital cameras, because the anti-aliasing filter interacts with polarized light.

2. Some polarizing filters absorb light than others, and none of them is completely color-neutral. It just depends on the polarizing material used. A circular polarizer has an additional quarter wave plate, which is completely transparent. The light loss has absolutely nothing to do with whether the filter is circular or linear.

3. They are all equally effective. They all reject virtually 100% of light that is polarized in one direction. The "effectiveness" of a filter depends on (1) whether the light from the subject is polarized (it may not be polarized at all); and (2) the orientation of the filter. That's why the filters rotate. Read the literature on how to use one.

 

[bjn70]

Do some internet searching and you can find some past discussion. I did a quick search and found an older thread on DPR and an older thread on Fred Miranda, as well as discussions other places. Most of the discussion said that linear polarizers worked ok with mirrorless cameras but there were some buts. The comment about the mirror being a part of the focusing system makes sense. The circular polarizer is supposed to transmit only light polarized a certain direction but then re-orient it so the light transmitted is no longer polarized. How it does it is beyond me but that explanation sounds like a reason why a CPL works with a DSLR.

 

[MediaArchivist]

3. They are all equally effective.

That depends on how you define "effective". Generally speaking, they will all filter based on the polarization of incoming light. So in a binary sense, they are all effective. But some have much higher transmission than others.

Lens Rentals

 

[ThrillaMozilla]

3. They are all equally effective.

That depends on how you define "effective". Generally speaking, they will all filter based on the polarization of incoming light. So in a binary sense, they are all effective. But some have much higher transmission than others.

Lens Rentals

That's the OP's term, not mine. They all have virtually complete extinction. I believe I already said that some have higher transmission than others. You don't necessarily want the ones with the highest transmission. You also need to consider color, since transmission is not constant across the spectrum.

I can't view the Lens Rentals article right now, but I think I remember Roger Cicala wasting a lot of time to find out that yes, indeed, they all have virtually perfect extinction.

 

[monte12345]

Comparing the two by eye alone the difference is easy to see but that may be due to hand holding the two filters and not getting an exact match on alignment. However on camera the difference is a lot more subtle. So subtle I really can't decide which is better. Below are both images and yes I didn't have the Pan axis locked down on my tripod and bumped it. I did have the camera set to manual exposure with the lens at f5.6 and shutter at 1/60. Lens used is the 40mm f2 Nikkor Z. IBIS was On and I can't see any negative effect of this choice. I'm curious to see what image others prefer.


Circular Polarizer


Linear Polarizer

 

[monte12345]

On a Nikon Mirrorless camera. When I got out the camera and started taking some pics there was no real difference in the exposure required and no effect on image quality. There was a subtle difference in the color rendition which I believe is due to much different coatings between a 50 year old Linear and a Circular purchased recently. See the pics posted in an earlier reply.

 

[mreynolds767]

Very similar Yes.

I think the circular polarizer is a tad sharper in your 2 images. I assume that is a little IQ loss with the other filter when compared.

I think I would choose the linear one for most pleasing color but there is not a big difference.

 

[ThrillaMozilla]

I prefer #1, the circular polarizer. The contrast is better, and there are other differences. In particular, the blacks are blacker and more numerous. Did you remember to rotate the polarizers to exactly the same position? You didn't mention that. Is the linear filter coated? Here's a histogram.


Histograms of combined RGB values. Circular (left) and linear (right). Notice the pronounced differences in peak heights, with slightly darker and much more numerous grays in #1.

From the midtone peak positions, the exposures are surprisingly close, but #2 appears to have slightly more exposure overall than #1. You used the same camera settings for both, but different filters transmit different amounts of light.

There is a lot of clipping in the blue layer. That is due either to overexposure or to printing too light from the raw file. That shifts the color of the sky toward green, especially near the horizon of #2.


Histograms of the blue layer. Circular (left); linear (right)

Compared to the circular polarizer, the linear polarizer is strongly biased toward blue, and slightly biased toward green. Here are the histograms from an area of concrete.


Histograms of an unshadowed part of the concrete. Circular (left) and linear (right). From top to bottom: R, G, B. Notice the shifts in modes, which signals a bias toward G and especially B.

There are several possible reasons for the differences.

1. In the Lens Rentals article (see post above) there is a picture of some filters, and you can easily see the differences in transmission and color. Different coatings could also have an effect.

2. Are you sure the linear polarizer is coated? This could artificially lighten the shadows and reduce the contrast.

3. As mentioned previously, you may not have the filters at exactly the same rotation.

4. As mentioned previously, the exposures are slightly different, probably because of a difference in transmission.

5. The sky is overexposed or printed too light.

As far as I know, none of these effects has anything to do with whether the filter is circular, so please do not tell people that a linear filter has more "pop", because nothing you have done is representative of any filters but yours.

 

[ishwanu]

monte12345 wrote:

One immediately noticeable feature of the circular polarizing filters I have are lighter in terms of light loss ...

Later:

Below are both images and yes I didn't have the Pan axis locked down on my tripod and bumped it. I did have the camera set to manual exposure with the lens at f5.6 and shutter at 1/60.

Funny, the shot with the circular filter wanted ISO 60 vs. ISO 90 with the linear. A minor difference, but it implies that the circular one transmitted less light.

Circular Polarizer


Linear Polarizer

 

[monte12345]

Unfortunately this forum won't allow me to edit posts once they have been responded to. When I wrote the initial post all I had done was an eyeball comparison. Once I saw the results from testing both filters I agree with your evaluation about "Pop", there is effectively no real difference in terms of Pop.

I also want to thank you for your use of the color histograms, all I've ever done with the histograms in the past has looked at the overall. Now that I've learned how useful the individual color histograms can be I will start using that feature. I do have to wonder about that clipping in the Blue channel.

Concerning coatings on the filter, I really don't see any reflected color casts from either side of the filter so I'm going to say it's not coated. Knowing my budget 50 years ago when this was purchased I am certain it was the least expensive Tiffen filter available. The newer circular is also a budget level Tiffen and also appears to be uncoated. BTW, if coated I suspect it's a single layer coating on either filter. BTW, I have considered getting multicoated filters but the cost increase is significant.

 

[ThrillaMozilla]

monte12345 wrote:

One immediately noticeable feature of the circular polarizing filters I have are lighter in terms of light loss ...

Later:

Below are both images and yes I didn't have the Pan axis locked down on my tripod and bumped it. I did have the camera set to manual exposure with the lens at f5.6 and shutter at 1/60.

Funny, the shot with the circular filter wanted ISO 60 vs. ISO 90 with the linear. A minor difference, but it implies that the circular one transmitted less light.

My mistake, not looking at the ISO settings. It's just the reverse. Since the linear one required a higher ISO setting for about the same lightness, it must have received a slightly lower exposure. You should be able to see it in the filters.

Circular Polarizer


Linear Polarizer

 

[ThrillaMozilla]

Concerning coatings on the filter, I really don't see any reflected color casts from either side of the filter so I'm going to say it's not coated. Knowing my budget 50 years ago when this was purchased I am certain it was the least expensive Tiffen filter available. The newer circular is also a budget level Tiffen and also appears to be uncoated. BTW, if coated I suspect it's a single layer coating on either filter. BTW, I have considered getting multicoated filters but the cost increase is significant.

Typically Tiffen filters are not coated, or at least they were not coated years ago. They're (were?) often used in the film industry, though.

 

[ishwanu]

One immediately noticeable feature of the circular polarizing filters I have are lighter in terms of light loss ...

Later:

Below are both images and yes I didn't have the Pan axis locked down on my tripod and bumped it. I did have the camera set to manual exposure with the lens at f5.6 and shutter at 1/60.

Funny, the shot with the circular filter wanted ISO 60 vs. ISO 90 with the linear. A minor difference, but it implies that the circular one transmitted less light.

It's just the reverse. Since the linear one required a higher ISO setting for about the same lightness, it must have received a slightly lower exposure.

D'oh! Of course that's right. Apologies to all.

 

[Marehan]

Yeah, you’re right about linear polarizers not messing with metering and AF on mirrorless cameras. Since they don’t have the mirror setup like DSLRs, you can totally use linear filters without any issues. I’ve noticed similar things with circular polarizers—they do seem to lose a bit of effectiveness compared to linear ones, especially if you’re used to that older gear. Just keep in mind that while you can use linear filters, circular ones are still pretty common for flexibility, especially if you decide to use an older lens with a DSLR later on.

 

[ThrillaMozilla]

Yeah, you’re right about linear polarizers not messing with metering and AF on mirrorless cameras. Since they don’t have the mirror setup like DSLRs, you can totally use linear filters without any issues. I’ve noticed similar things with circular polarizers—they do seem to lose a bit of effectiveness compared to linear ones,...

Are you and the OP sure you understand how a polarizing filter works and how to use it?

Here's a good article on how to use it: https://www.the-digital-picture.com/Reviews/Circular-Polarizer-Filters.aspx . You need to understand that (1) if the light from the subject isn't polarized, the filter will act like a neutral density filter, but otherwise it will do nothing for you; and (2) you need to know how to rotate the filter.

If you read the Lens Rentals article, you will discover that Roger Cicala tested 6 CP filters, including a Tiffen. He wrote: "They were all at least 99.9% efficient at doing their job, polarizing light."

Any filter substantially worse than that would be quite unusual. If you think you have a filter that is significantly less "efficient" than that, I think you might want to tell us how you tested.

 

[GeoffRG]

I expect that circular became so essential was due to the use of semi silvered mirrors in a DSLR for behind the mirror metering and AF modules. Mirrorless cameras don't use these features so can the older linear filters be used. One immediately noticeable feature of the circular polarizing filters I have are lighter in terms of light loss and just don't seem to be as effective as the linear filters I had when I was shooting with an F2Sb.

Essentially you are correct, a linear polarising filter causes problems due to the use of semi-silvered mirrors.



Now for the fun bit, a circular polariser and a linear polariser are actually exactly the same polarising element. The difference is that the circular polariser also has a second layer that randomises the polarised light transmitted by the polarising layer. Thus, there is no reason why a circular or linear polariser, from the same manufacturer, should have any different effect in terms of light transmission or colour rendition, assuming that the additional randomising layer does not reduce transmission or cause a colour cast.

Thus, in the absence of anything in/on the sensor that is affected by polarised light, a linear polariser and a circular polariser should behave in the same way with a mirrorless camera.

 

[Gloomy1]

I am tacking this onto Geoff's post not because I disagree with his post but to put it at the end. In Summation. All polarizers linear and curcular from all manufacturers polarize the light 99% effectively. All polarizers block some of the light and therefore work as low value ND filters. The amount of light blocked is not a function of the type of polarization, it is not the case that linear ones lett more light through or vice versa. All polarizers have colour casts good ones have subtle casts. It is likely that any polzrizer above £50 will not have an objectionable cast. Lee filters sell a polarizer that has a deliberate warming cast although that "effect" is somehat lessened if we use auto colour balance. Again the colour cast is not dependant on the type of polarization. Anyone who wants to buy the very exoensive polarizers because they gsve minimal casts should perhaps think about what happens when we twist the filter and start to polarize the light. The blues in the sky start to darken, foliage becomes more saturated, water in rivers especially becomes less blue and more brown as the sheen is taken off and the river bed shows through. The changes in colour balance as more polarization is applied is far greater than any colour cast inherent in the filter, this again is irrespective of the type of polarizer. Any colour cast that is there can be fairly easily neutralised. There may be good reasons to buy expensive polarizers but how much they polarize or lack of colour cast or even how little they block all light unless you use a polarizer when shooting fast action nature. With a modern mirrorless system there are no reasons why anybody should choose a polarizer because it is linear or circular. In the end it is tge results that matter anf if people are happy using a £20 Amazon or a £250 branded polarizer good but often I think that people look at their images and use poor equipment of any type ( polarizers included) as an excuse for less than stellar images. Ken

 

[ThrillaMozilla]

Thus, in the absence of anything in/on the sensor that is affected by polarised light,

That's just the thing. An antialiasing filter consists of one or two layers of an anisotropic crystal that most certainly does interact with polarized light.

I think one would have to look very closely to see the effect, however. I haven't seen a test of this effect, so I don't really know about this part. For my money, however, it's circular polarizers, which can also be used with my SLR.

a linear polariser and a circular polariser should behave in the same way with a mirrorless camera.