Polarizer and D2 - again, but with a slant...

[rob_rob]

Advertisement (gone after registration)

I got my filter from Heliopan. I have to admit, I don't use the polarizer on my D2 too often.

 

Last Friday 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 in the sunlight was unusually high. I went with a girl to the beach to take some "summer photos" and I decided to use my polarizer.

 

The resulting images are less than spectacular. I expected really deep blue skies and saturated colors. However, nothing like that has occured, unfortunately. The 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).

 

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 D20 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.

 

Is it possible that using a polarizing filter with a digital camera, especially the one with a small sensor, may have a degrading effect on picture quality? Does this kind of combo interact with a sensor or a firmware of a camera in some sort of bad way? Could this problem be aggrevated by extremely high level of UV in the light? Did LEICA know about it and that's why they decided to use a unique filter size, so that people would be discouraged to use filters on the D2?

[philipotto]

I love your avatar

[sgegreen]

I use a Hoya 72mm S-HMC circular polarizer with my Lumix LC1 with no such problems. I'd always thought the Heliopan filters were reckoned to be better than the Hoya ones.

[jlancasterd]

Rob

 

Are you using a linear or circular polariser? IIRC There is a problem with one type not working well with CCDs but not the other.

 

John

 

I got my filter from Heliopan. I have to admit, I don't use the polarizer on my D2 too often.

 

Last Friday 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 in the sunlight was unusually high. I went with a girl to the beach to take some "summer photos" and I decided to use my polarizer.

 

The resulting images are less than spectacular. I expected really deep blue skies and saturated colors. However, nothing like that has occured, unfortunately. The 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).

 

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 D20 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.

 

Is it possible that using a polarizing filter with a digital camera, especially the one with a small sensor, may have a degrading effect on picture quality? Does this kind of combo interact with a sensor or a firmware of a camera in some sort of bad way? Could this problem be aggrevated by extremely high level of UV in the light? Did LEICA know about it and that's why they decided to use a unique filter size, so that people would be discouraged to use filters on the D2?

[rob_rob]

Circular, of course.

[etanguero]

did you use your polarizer 'solo'?

or did you put it on top of you UV-filter/lensprotector (if you do use one)?

 

I could imagine that the UV-filter would have filtered away any of this extra UV-light in the sky and therefore reducing to normal situation.

 

eT

[rob_rob]

Advertisement (gone after registration)

Solo, of course.

[ho_co]

Rob--

I'm unaware of any reason that a polarizer shouldn't work on a digital camera. Sounds as if you are on top of the matter, so this is probably not news.

 

Remember that the greatest sky polarization is at 90° to the sun, so when the sun is overhead, the entire horizon shows high polarization. At other times, northerly and southerly directions will show more polarization than easterly and westerly.

 

Remember also that at its widest 7mm setting, the lens covers 76° diagonally (64° horizontally), so in general no matter which direction you point the camera, the sky will show a fairly large range of polarization.

 

Hope this helps!

 

--HC

[rob_rob]

Exactly, XC. I was expecting a varying degree of darkening of the sky, depending on the angle between the lens and the sunrays. Still, I thought, overall darkening should be little more pronounced. But you've got me thinking because of this remark:

so when the sun is overhead, the entire horizon shows high polarization.

I've been shooting between noon and about 3:00 PM. The sun, indeed, was high and that could have made the entire scene darker, hence lowered contrast and not much of sky darkening... Still, the question of higher than normal level of noise at ISO 100 remains unsolved.

[like_no_other]

Rob,

the level of noise depends on the sensor and signal processing. Therefore it shouldn't be increased by using a polarizer. Did you do any post processing to these pictures that could have made always existent noise more visible, maybe some automatic adjustments are applied by your software you are not aware of?

 

Another story is about the reproduction of saturated colours in sunny pictures. It's not unusual with digital cameras that you may find unreal colour reproduction in .jpgs in such situations. One example is that green tips over into yellow. This alone without using a polarizer.

My experience with polarizers in similar situations is that you may get colours that doesn't match the results you are accustomed to get from film. For example my polarized skies have good contrasts but an unattractive very dark steel blue.

Maybe such effects vary from camera to camera due to different sensors and different .jpg engines.

Regarding the sensor you can't do much. Regarding the .jpg engine shooting in raw format maybe helps a bit (I don't know if the digilux offers raw output). Using the largest available colour space, post processing the raws, ...I'm sure there are metohds to improve the results.

Particularly I think that you won't find the lack of image quality with polarizers in every picture. Different colours and different light will result in more or less good or bad picture output of your camera.

 

But there was another annoying polarizer effect that attracted my attention. It is a visible lack of sharpness compared sensor to film with the same lens and the same filter. I discovered this with the best Helioplan polarizer (slim multicoated version) and a good 20mm prime lens. The lens alone without polarizer is able to produce very sharp pictures on the digital camera as well as on the film camera.

Together with the filter there is a visible lack of sharpness in the digital picture, not in the analogue one. The polarizer must degrade the optical system in a way that only gets visible with the sensible sensor of a digital camera. Same with medium grade lenses which are ok with film and bad with digital cameras.

 

I don't know if there are any digital optimized polarizers available from Helioplan or B&W?

 

Best regards

 

 

Exactly, XC. I was expecting a varying degree of darkening of the sky, depending on the angle between the lens and the sunrays. Still, I thought, overall darkening should be little more pronounced. But you've got me thinking because of this remark: I've been shooting between noon and about 3:00 PM. The sun, indeed, was high and that could have made the entire scene darker, hence lowered contrast and not much of sky darkening... Still, the question of higher than normal level of noise at ISO 100 remains unsolved.

[rob_rob]

Philip,

Thanks for the insightful response.

 

I don't know whether you've noticed my remark about extremely high UV content in the light at time of my shooting session. I know I am just trying to find a shred of logic in connection to a seemingly irrelevant fact, but maybe that intensely UV-saturated light has pushed the camera firmware's capability of capturing an optimal image to the limit?

 

I don't even go as far as converting my images to JPEG format. I print an image converted from RAW to a Photoshop's '.psd' format, or directly from Lightroom after converting my RAW image. But I did notice the "steel blue" color of the sky you are mentioning. This is due to relative insensitivity of all digital sensors to a red content in the blues.That's why blue color is often rendered by digital cameras as cyan. BTW, I just bought a specialized Photoshop plug-in that maximizes the natural spectrum of colors in digital images. It is especially good at restoring purples and blues (DCF Full Spectrum from the TRIBECA IMAGING LABORATORIES - http://www.tribecalabs.com).

Together with the filter there is a visible lack of sharpness in the digital picture, not in the analogue one.

See, there has to be some dependency between the quality of image from digital cameras with and without a polarizer. You've noticed that in regard to sharpness and I found out the same about contrast and noise. That reinforces my original thought that Leica researched that problem well enough to know that their top quality lenses will NOT give top quality results with filters attached and that's why they chose to use a unique filter size on their lens, so that there will be no filters available in that particular size for people to use (69mm)...

[ho_co]

... The lens alone without polarizer is able to produce very sharp pictures on the digital camera as well as on the film camera. Together with the filter there is a visible lack of sharpness in the digital picture, not in the analogue one.

Philip--very interesting. Have you tested the 20mm on both cameras but with a different filter?

 

US Leica tech taught at one time that putting any filter in front of a lens could reduce image quality, most especially with wide-angle lenses. The reasoning: Each glass surface of a filter causes some refraction of the light rays striking it. Since wide-angle lenses accept light from a higher angle of incidence, the amount of image degradation is greater there than with normal and long lenses. Of course, a polarizer would be worse yet because it has 4 instead of 2 glass surfaces.

 

I don't know whether the argument is accurate. You might try testing: 20mm lens alone; 20mm lens with UVa; 20mm lens with polarizer; each on both cameras.

 

And then for confirmation, do the same thing with a normal or tele lens on both cameras.

 

You've seen the problem comparing digital and film with the 20mm. Making the test I've described would confirm that the polarization itself is the culprit and not the combination of polarizer and WA lens.

 

That is, the problem might be simply that the way three individual RGB pixels pick up light produces a worse problem with the 20mm on digital as compared to the way a three-dimensional film emulsion does. If the above assumption is correct, the problem would be reduced with a longer lens. And if the above assumption is correct, there would be some degradation of image quality with a different filter on the digital compared with the analog camera.

 

--HC

[rob_rob]

I spent some time yesterday comparing the pictures I was talking about to the pictures taken without polarizer (same camera). There's not enough evidence to say that there's significantly more noise with than without polarizer. Perhaps increased graininess in shadows made an impression of a genarally higher noise... But the contrast is definitely lower. I noticed that during conversion from RAW, I had to increase contrast signicicantly to get decent print. There's also a shift in overall color rendition, but that doesn't surprise me. And, as HC suggested, the wider the angle, the more image degradation (which also doesn't surprise me).

 

I came to the conclusion that, however small and hard to define precisely, there are differences between analog and digital image rendering with a polarizer versus no polarizer. Perhaps the reason, by analogy with wide angle versus normal or tele lenses with film, may be the fact that my Digilux 2 has relatively tiny sensor in comparison with a film camera frame size.

 

Despite my conclusion, I would still love to hear what everyone thinks about this problem or... "problem".

[rob_rob]

P.S. Also, whatever increase in the noise visibility, may be and probably is a result of my increasing contrast.

[ho_co]

Rob--

You got me interested.

 

I went out today with a thin Heliopan CPol on a Nikon D200, and was very disappointed with the results. Used the 18-70mm 'kit lens,' 135 equivalent of 27-105. Shot pairs, one with and one without the polarizer.

 

I haven't yet compared sharpness, but otherwise:

1) in the early afternoon daylight, I saw virtually no effect of polarization

2) obvious effect of polarizer was warmer, desaturated color (blue sky became steel-grey)

 

Curiouser and curiouser.

 

--HC

[pierovitch]

The polariser may be transmitting Infra Red in higher proportion. Imaging chips are more sensitive to these wavelengths.

See http://www.photo.net/equipment/filters/

For a well documented comparison on UV control.

 

Cheers Pierre

[rob_rob]

 

I went out today with a thin Heliopan CPol on a Nikon D200, and was very disappointed with the results. Used the 18-70mm 'kit lens,' 135 equivalent of 27-105. Shot pairs, one with and one without the polarizer.

 

I haven't yet compared sharpness, but otherwise:

1) in the early afternoon daylight, I saw virtually no effect of polarization

2) obvious effect of polarizer was warmer, desaturated color (blue sky became steel-grey)

 

--HC

Exactly my point, HC! Same polarizer I've been using with my Digilux 2. So, is this a problem inherent to this specific polarizer, or is it a wider issue having to do with a small digital sensor in combination with any circular polarizer? I would bet my money on the latter...

 

Perhaps we should start asking the manufacturers this same question? I will wait for some more potential feedback from other photographers and after a while I will start corresponding with Heliopan and Leica regarding this topic. Anyone else has something more to add to this thread?

[cbretteville]

Rob,

I may be reaching, but could it have something to do with where on the globe you live? My impression of Helipoan's pola on my D2 is favourable. When I get home I'll shoot some comarison shots to see if it is for real or just placebo.

[rob_rob]

Any input you can give me (us) will be much appreciated, Carl. BTW, I live on the East Coast of the USA (NY state).

 

Personally, I suspect that the technology is to blame rather than geography (provided that we don't do something stupid and obvious that we shoudn't do, in the first place, although nothing like that comes to my mind... ).

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[ho_co]

The polariser may be transmitting Infra Red in higher proportion. Imaging chips are more sensitive to these wavelengths.

Pierre--

Could be. It's an interesting idea that I hadn't considered.

 

But the Digilux 2 has almost no IR sensitivity, while the D200 has a fair amount; and Rob and I are getting similar results from the two cameras.

 

Thanks for the link--there's a heck of a lot about digital that I can stand to learn:).

 

--HC

 

PS--OOPS!

I just checked the link, and it is concerned with UV, not IR. UV is indeed a higher-probability culprit than IR.