Ad
Ad
Ad
Pages: [1] 2 3 »   Bottom of Page
Print
Author Topic: Using a polarizer with a digital camera  (Read 42025 times)
iconoclast
Newbie
*
Offline Offline

Posts: 14


« on: July 24, 2006, 01:49:02 AM »
ReplyReply

This is my first post in a LL forum, so perhaps I should introduce myself...

First of all, English is not my native language even though I live in the U.S.A., so bear with me if I don't always make sense. I am involved in photography since 1971. For the last 2 and a half years I am using a digital camera exclusively.

I'd like to ask those of you who are involved in digital photography a question I wasn't able to get an answer to so far, even though I already posted this problem on a couple of prominent forums. It may sound silly and it may be dismissed very easily without even a second thought, but I hope to get to the bottom of this problem. Here it goes:

Previous weekend, we had a perfect sunny day here on Long Island. The sky was blue and almost without a speck of a cloud. As I learned from TV weather report later that weekend, the UV-content advisory was in the purple zone, which was the highest possible. I went with a girl to the beach to take some "summer photos" and I decided to use my polarizer.

I expected no reflections on the water, really deep blue skies and saturated colors. The resulting images were less than spectacular. The reflections on the water were indeed gone. Skies were of various depth of blue, but not much deeper than the ones photographed without polarizer. But that wasn't the biggest of my disappointments. What's worse, overall contrast was lower than normal and my pictures contained visible noise/grain (all pictures were taken as RAW at ISO 100). I use Leica Digilux 2 with a non-interchangeable lens which has an unique filter diameter (69mm). For that reason, there's only one polarizer that fits that lens. And it's been made specifically for this particular lens by one of the best filter manufacturers in the world (Heliopan). So, there's no question of incompatibility or low quality of the filter.

So, I started wondering. In the past, I've heard complaints about results of the D2/polarizer combo. I've never thought much of those complaints until now. Couple of days after I took those pictures, I read a post on another forum. In that post, an owner of Canon 20D wrote of his own experience that was like an exact mirror image of my own disappointment - and it happened exactly at the same time, only he lived in Ireland and I live in NY metro area.

So far, I have been unable to find time for any tests and this weekend weather was not cooperating either. At this point, I am really not sure that I should expect better results from pictures taken in such situation - for the following reasons:
  • The sensor, especially small one like in my camera, the anti-UV filters built around every camera's sensor and the firmware of the camera may all act alone or in combination producing degraded image quality when dealing with light filtered by a polarizing filter.
  • Extremely high UV content in the light may force the camera's firmware to work beyond its normal range of the ability to optimize an image. In that case, I think that even without polarizing filter, I would be able to detect adverse result on the image quality.
  • And last but not least, I suspect that the maker of my camera (Leica) predicted potential trouble - otherwise why would they decide to use 69mm filter size (the only lens in existence with this filter size), if not to discourage people from using filters? Leica is well known for their thorough research and sticking to the very highest standards in photography, which makes me think that they didn't want people to use filters with at least this particular model of their camera.
I still intend to do more conclusive testing, but it will take more time and I don't know whether I'll be able to get similar weather conditions...

Did anyone of you ever experienced similar problem or heard of one like that?
Logged

Regards,
Rob
jani
Sr. Member
****
Offline Offline

Posts: 1604



WWW
« Reply #1 on: July 24, 2006, 08:34:22 AM »
ReplyReply

Are you using the automatic settings, or are you shooting in raw?

Keep in mind that a polarizer reduces the amount of light by at least one full stop, and in some cases as much as two stops.

This means that if you leave a camera in fully automated mode, you risk taking pictures at a higher ISO level than the camera can reproduce noiselessly. Using e.g. ISO 400 will create a bit of noise with a Leica Digilux 2, according to various reviews on the net.
Logged

Jan
Lisa Nikodym
Sr. Member
****
Offline Offline

Posts: 1702



WWW
« Reply #2 on: July 24, 2006, 10:06:57 AM »
ReplyReply

Also, just in case you weren't already aware of it while you were taking the pictures, polarizers do the full extent of their "magic" to the sky only when the sun is at about 90 degrees to the axis between the camera and the subject.  At the time, did you check where the sun was?  If it was close to behind you or in front of you, that would explain why the sky wasn't much darker than without the polarizer.  (If you are already very familiar with using polarizers with film cameras, then apologies for stating what might sound obvious; with new posters here, it's hard to tell what their experience level is...)

And as Jan says, the polarizer cuts down on the light reaching the sensor, which could increase the noise level if the camera is either upping the ISO or reducing the exposure because of it.

Lisa
Logged

iconoclast
Newbie
*
Offline Offline

Posts: 14


« Reply #3 on: July 25, 2006, 03:38:19 AM »
ReplyReply

Quote
Are you using the automatic settings, or are you shooting in raw?

Keep in mind that a polarizer reduces the amount of light by at least one full stop, and in some cases as much as two stops.

This means that if you leave a camera in fully automated mode, you risk taking pictures at a higher ISO level than the camera can reproduce noiselessly. Using e.g. ISO 400 will create a bit of noise with a Leica Digilux 2, according to various reviews on the net.
[a href=\"index.php?act=findpost&pid=71615\"][{POST_SNAPBACK}][/a]
Jani, those particular pictures were all taken in 'auto' mode, but that does not imply that the camera changes the ISO value whenever it sees less light than it is 'comfortable' with. My camera can only change shutter speed or aperture or both, but not ISO speed. And all pictures were taken at ISO set to 100 (all RAW files). There was more than enough light to take correctly exposed shots at ISO 100 even with a polarizer attached. All shots are sharp and are normally exposed. Only the contrast is lower than normal (which I can contribute to using any kind of filter in front of a lens). The thing that I don't understand is the noise (although very fine) which was visible all over the frame, not just in shadows.

Quote
Also, just in case you weren't already aware of it while you were taking the pictures, polarizers do the full extent of their "magic" to the sky only when the sun is at about 90 degrees to the axis between the camera and the subject.  At the time, did you check where the sun was?  If it was close to behind you or in front of you, that would explain why the sky wasn't much darker than without the polarizer.  (If you are already very familiar with using polarizers with film cameras, then apologies for stating what might sound obvious; with new posters here, it's hard to tell what their experience level is...)

And as Jan says, the polarizer cuts down on the light reaching the sensor, which could increase the noise level if the camera is either upping the ISO or reducing the exposure because of it.

Lisa
[a href=\"index.php?act=findpost&pid=71623\"][{POST_SNAPBACK}][/a]
Lisa, yes, I am aware of that. And no, apologies are not necessary.  As I mentioned in my first post, the sky looked a bit darker then without polarizer. I have been shooting for over 30 years with my analog cameras and I always had darker skies with a polarizer (even when shooting in 'auto' mode).

Of course, I was changing my position in relation to the direction of sunrays, but even taking that into account, overall, I should have much darker skies (to a varying degree - depending on the angle).
Logged

Regards,
Rob
Phuong
Full Member
***
Offline Offline

Posts: 113


« Reply #4 on: July 25, 2006, 10:51:50 AM »
ReplyReply

can you post a few example pictures. it's better "analysing" it that way.
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #5 on: July 27, 2006, 01:22:41 PM »
ReplyReply

I concur with those who have said to post problem images. It's too hard to tell what the issue is without having something to look at. As has been mentioned, polarizers do penalize on between 1 and 2 stops of light, meaning  a wider aperture, longer exposure, or higher ISO is required to get equivalent exposure. A polarizer does not by itself increase saturation; all it does is remove certain specular highlights and reflections that can mask the color of foliage (many leaves have a semi-shiny surface) and the color of water. A polarizer's effect on sky depends heavily on atmospheric conditions and the angle between the camera and the sun; when shooting directly into or away from the sun, its effect will be negligible. The increased noise complaint makes me suspect that the camera may have upped the ISO for some reason, or you may have inadvertently underexposed, which will increase noise levels. But again, it's all speculation until we can see the shots.
Logged

delnerdo
Newbie
*
Offline Offline

Posts: 1


« Reply #6 on: August 03, 2006, 09:31:07 PM »
ReplyReply

I read this discussion while trying to justify the expenditure on this filter.
Is it possible that the poor performance results from "circular" polarization?

The alternative is "linear" polarization.
The difference is how the tiny little lines that cause the polarizing effect
on photons (light particles) are oriented on the filter substrate,
as either concentric circles or parallel lines.

When you look at polarized light through a linear polarizing filter,
then rotate the filter 90-degrees left or right, the contrast
in effect should be most striking.  However, with a circular
polarizing filter, the behavior should be different.
(Light reflecting off an electrical insulator is polarized;
conductive reflectors do not polarize photons... see
Unihedron's EM spectrum poster for more.)

You may recall that standard/linear polarizers have a rotating
ring-mount and lever, with which to fine-tune the effect.  
I bet the Heliopan 69mm has no such adjustments, due to its
circular polarization.

It's merely a conjecture (hypothesis), but worth considering.
If wrong, thanks anyway for saving me from spending the $$$!

HTH!!!
Logged
iconoclast
Newbie
*
Offline Offline

Posts: 14


« Reply #7 on: August 04, 2006, 02:26:56 AM »
ReplyReply

I may post a sample... eventually. But right now, I am trying to gather as much input from other photographers as I can.

Just two days ago I had a feedback from a photographer on the Leica Forum, who said he had exactly the same result shooting with the exact same kind of a polarizer (only with a different camera/lens). And his experience was not a random event. Inspired by the discussion, he went out and did test shots with and without the polarizer.

I am not very familiar with the theory of light and polarization effect differences between linear and circular polarizers, but I think there might be some truth to what delnerdo just said. I will investigate that possibility, as well. Thanks!

Oh, BTW! Regarding what Jonathan said about the increased noise - I think that was entirely my fault (I had to increase the overall contrast of a majority of my pictures to get decent prints and therefore, inadvertently, I also increased a perception of noise).
« Last Edit: August 04, 2006, 02:32:27 AM by iconoclast » Logged

Regards,
Rob
jani
Sr. Member
****
Offline Offline

Posts: 1604



WWW
« Reply #8 on: August 04, 2006, 06:49:07 AM »
ReplyReply

Quote
I read this discussion while trying to justify the expenditure on this filter.
Is it possible that the poor performance results from "circular" polarization?

The alternative is "linear" polarization.
The difference is how the tiny little lines that cause the polarizing effect
on photons (light particles) are oriented on the filter substrate,
as either concentric circles or parallel lines.
No, not quite.

See this FAQ on polarizers for an explanation.

Yes, circular polarizers can degrade quality more than a pure linear polarizer, but you don't have much of a choice with cameras using beam splitting for autofocus (most, if not all DSLRs, IIRC).
Logged

Jan
Doug Kerr
Newbie
*
Offline Offline

Posts: 17


« Reply #9 on: August 04, 2006, 07:04:52 AM »
ReplyReply

Hi, Rob,

I assume you rotated the polarizer while composing with the EVF or back panel monitor to see what orientation of its axis produced the desired effect.

Did you see the desired effect when doing that, but it did not appear in the actual captured image?

Or were you unable to produce the desired effect when composing?
Logged

Best regards,

Doug

Visit The Pumpkin, a lbrary of my technical writings:
http://doug.kerr.home.att.net/pumpkin

"Make everything as simple as possible, but no simpler."
Doug Kerr
Newbie
*
Offline Offline

Posts: 17


« Reply #10 on: August 04, 2006, 07:13:46 AM »
ReplyReply

Quote
I read this discussion while trying to justify the expenditure on this filter.
Is it possible that the poor performance results from "circular" polarization?

You may recall that standard/linear polarizers have a rotating
ring-mount and lever, with which to fine-tune the effect. 
I bet the Heliopan 69mm has no such adjustments, due to its
circular polarization.
A circular polarizer has an axis of polarization (with respect to its handling of the "incoming" light) - it just does not have an axis of polarization ofits "output" light (which is why it is safe to use on cameras with semireflective mirrors in the metering path, such as many dSLRs).

Thus a circular polarizer will have a rotatable mount just like a linear polaraizer (and must if it is to be able to perform its function inphtpography - selective attenuation of components of the "arriving" light based on the axis of polarization of the light from the image. Just like a linear polarizer, it must be rotated to produce the desired effect, the selective attenuation of polarized components of the light from the image.

You can find a further discussion of this in my tutorial article, "Polarization of Light and Polarizers ", available here:

http://doug.kerr.home.att.net/pumpkin/inde...ationPolarizers

Best regards,

Doug
Logged

Best regards,

Doug

Visit The Pumpkin, a lbrary of my technical writings:
http://doug.kerr.home.att.net/pumpkin

"Make everything as simple as possible, but no simpler."
iconoclast
Newbie
*
Offline Offline

Posts: 14


« Reply #11 on: August 04, 2006, 11:37:29 PM »
ReplyReply

Quote
I assume you rotated the polarizer while composing with the EVF or back panel monitor to see what orientation of its axis produced the desired effect.

Did you see the desired effect when doing that, but it did not appear in the actual captured image?

Or were you unable to produce the desired effect when composing?
Of course, I did change the angle of rotation of the polarizer to get the maximum effect. Also, I changed rotation as I was switching framing from vertical to horizontal and vice versa.

I did see changing effect while rotating the polarizer during composing, but the effect was not as strong as I expected. Same was showing in resulting pictures. Perhaps my expectations were not entirely realistic, since my Digilux 2 gives me very nice, dark skies in sunny conditions even without any polarizer at all.

Doug, thanks for the link. There's a wealth of information many of us can use (I know, I certainly can).
Logged

Regards,
Rob
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #12 on: August 05, 2006, 04:14:04 AM »
ReplyReply

There is no practical difference between linear and circular polarizers, except:

Circular polarizers reduce light transmission by about 1.5 stops, compared to a little over a stop for a linear polarizer.

Linear polarizers can cause problems with the AF and metering of SLR cameras when the polarizer is oriented in certain directions. Circular polarizers do not have this problem.

The effect of the polarizer is clearly visible in the camera viewfinder; that's how you adjust it to the proper orientation.

ElNerdo is way out in left field; all polarizers have rotation adjustment.
Logged

stede66
Newbie
*
Offline Offline

Posts: 2


WWW
« Reply #13 on: August 15, 2006, 05:33:01 PM »
ReplyReply

I've had the same problem Iconoclast had.
Bright blue sky day, used the circular polarizer, and in RAW-Adobe Bridge the sky was "texturized" with niose.  
The problem is I used a 20D at ISO 100, (setting: contrast --, sharpness --, saturation -, color tone 0), a Canon EF 70-200 used at 105,  f2.8 at f/10 1/400 sec., id est the best conditions to shoot.
Same problems with other lens 20-35 f/2.8 at various aperture and speeds.  

Problem solved with Magic Wand to select the sky, Noise Ninja, then Inverted Selection, Unsharp Mask on the building.  
This time I was lucky the building border had color tone very different from the surrounding sky, but if I had to shoot a forest on a lake with the blue sky?

Any further explanation about the phenomenon?

Thanks,
Stefano

(the shoot attached was not modified, no sharpening, no Noise Ninja, only reduced to 8 bit/channel, Jpegged to 8, and cropped the to 25%)
« Last Edit: August 15, 2006, 06:01:22 PM by stede66 » Logged
Naude
Newbie
*
Offline Offline

Posts: 1


« Reply #14 on: September 02, 2006, 02:59:34 AM »
ReplyReply

I have the exact same problem on my 350D with the polarizer. it creates noticeable noise at iso100 and at iso200 with clouds in the sky it's simply not worth using a polarizer.
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #15 on: September 02, 2006, 09:37:34 AM »
ReplyReply

Quote
I've had the same problem Iconoclast had.
Bright blue sky day, used the circular polarizer, and in RAW-Adobe Bridge the sky was "texturized" with niose.  
The problem is I used a 20D at ISO 100, (setting: contrast --, sharpness --, saturation -, color tone 0), a Canon EF 70-200 used at 105, f2.8 at f/10 1/400 sec., id est the best conditions to shoot.
Same problems with other lens 20-35 f/2.8 at various aperture and speeds.  

Problem solved with Magic Wand to select the sky, Noise Ninja, then Inverted Selection, Unsharp Mask on the building.  
This time I was lucky the building border had color tone very different from the surrounding sky, but if I had to shoot a forest on a lake with the blue sky?

Any further explanation about the phenomenon?

Without looking at the original RAW, this is just an educated guess, but since polarizers act as a 1-2-stop ND filter, when shooting at ISO 100 your effective ISO is actually between 200 and 400 because of the light loss from the filter. Increasing exposure time will increase noise; a 2-second exposure will have more noise than a 1-second exposure at the same ISO setting given identical lighting and a 1-stop aperture adjustment to maintain identical overall exposure. In many instances the difference is not very significant, but when noise levels are on the edge of acceptable, sometimes the lengthened exposure times required by the polarizer will increase the noise level to undesirable levels.

Another possibility is that you simply underexposed because you were trying a bit too hard to keep detail in the building wall.
« Last Edit: September 02, 2006, 09:40:59 AM by Jonathan Wienke » Logged

opgr
Sr. Member
****
Offline Offline

Posts: 1125


WWW
« Reply #16 on: September 02, 2006, 01:52:16 PM »
ReplyReply

The noise looks very significant. More than the usual culprits would explain. At first I was thinking it could be that situations that require a polarizer are usually situations with considerable higher ambient temperature. But the resulting noise would not be in this order...

Could possibly be a problem where the Red channel registers lower than expected due to the polarizer, introducing artifacts during debayering because the expected levels for cross channel reproduction are not met.

Does changing White Balance during RAW conversion make a useful difference? (As in something really outrageously low or high?)

Does using the Canon RAW converter make a difference? (Or Leica converter in OP's case).
Logged

Regards,
Oscar Rysdyk
theimagingfactory
iconoclast
Newbie
*
Offline Offline

Posts: 14


« Reply #17 on: September 02, 2006, 10:35:00 PM »
ReplyReply

Quote
Could possibly be a problem where the Red channel registers lower than expected due to the polarizer, introducing artifacts during debayering because the expected levels for cross channel reproduction are not met.

Does changing White Balance during RAW conversion make a useful difference? (As in something really outrageously low or high?)

Does using the Canon RAW converter make a difference? (Or Leica converter in OP's case).
[a href=\"index.php?act=findpost&pid=75322\"][{POST_SNAPBACK}][/a]
In my case, processing a Leica RAW file in ACR does improve some qualities of a picture (mainly contrast), but overall quality is visibly degraded in comparison to shots taken without a polarizer.
Logged

Regards,
Rob
John Sheehy
Sr. Member
****
Offline Offline

Posts: 838


« Reply #18 on: September 03, 2006, 04:10:36 AM »
ReplyReply

Quote
(the shoot attached was not modified, no sharpening, no Noise Ninja, only reduced to 8 bit/channel, Jpegged to 8, and cropped the to 25%)
[a href=\"index.php?act=findpost&pid=73456\"][{POST_SNAPBACK}][/a]

Under-exposure combined with a lack of red light in the sky.  With the contrast set at -2, the JPEG review image should be barely flashing lighty in little patches in the brightest part of the white paint for optimal exposure.  In fact, when you know a deep blue sky is going to be problemic, you might want to take advantage of some RAW converters' ability to render greyscale highlights from just one channel and blow that bright white paint by 2/3 stop.

Judging from the noise, you exposed this image about 2 - 2.5 stops lower than I would have.  That means 4 to 6 times the noise in the shadows (the red in the sky is a shadow; the shadow noise is much higher than it would have been, exposed better, at ISO 1600.  If you shoot Canon models that came out in 2004 or later, it can not be overstated that under-exposure is *FAR FAR* worse than using a higher ISO with the same f-stop and shutter speed.  The only danger of using a higher ISO at a given f-stop and shutter speed is clipping, if you're shooting RAW.  I do all my bright sun, non-telephoto work at ISO 200 with my 20D, by default.  

Oh yeah, BTW, the 20D is really ISO 115 native, so ISO 100 is not really on my list of ISOs to use.  It has 1/4 stop of RAW highlights missing, and Canon stretches the data in the highlights to hide this fact.  Why they couldn't have just had ISOs 115 230 460, etc, is beyond me.  They cripple cameras to stick to useless standards.  Even 125 200 400 ... would have been fine.
Logged
John Sheehy
Sr. Member
****
Offline Offline

Posts: 838


« Reply #19 on: September 03, 2006, 04:14:28 AM »
ReplyReply

Quote
Without looking at the original RAW, this is just an educated guess, but since polarizers act as a 1-2-stop ND filter, when shooting at ISO 100 your effective ISO is actually between 200 and 400 because of the light loss from the filter. Increasing exposure time will increase noise; a 2-second exposure will have more noise than a 1-second exposure at the same ISO setting given identical lighting and a 1-stop aperture adjustment to maintain identical overall exposure. In many instances the difference is not very significant, but when noise levels are on the edge of acceptable, sometimes the lengthened exposure times required by the polarizer will increase the noise level to undesirable levels.
[a href=\"index.php?act=findpost&pid=75303\"][{POST_SNAPBACK}][/a]

These types of shots won't result in any time-related noise.  When you do get it, it appears as isolated hot pixels, not readout noise banding artifacts as in the image of the white building against the deep blue sky.
Logged
Pages: [1] 2 3 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad