Westminster fountain at sunset

by Jorge Córdoba

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

When processing sound, one can use a Compressor to reduce the dynamic (DR) range of the signal. In photography, we often run into the problem of scenes with DR larger than the sensor's capture ability. For example, shooting a person on a white snow background. This leads to either the subject is underexposed, the snow is overexposed or both.

A Neutral Density filter is used to reduce the amount of light received by the sensor. However, a uniform filter does not (theoretically) reduce scene DR. Graduated ND filter can be helpful in some cases, but the use is very specific.

Is there a device (material) that can reduce the received DR? Preferably, a passive device (like a lens filter)?

Obviously, in the sound compressor example above, the DR of the signal chain has to be able to accommodate the original DR in order to prevent unwanted clipping and distortion. In Digital Photography, this is equivalent to a high DR sensor, and the compression itself is equivalent to HDR tone mapping.

share|improve this question
1  
engadget.com/2011/06/25/… :D –  Dreamager Sep 14 '11 at 15:05
    
A filter wouldn't be able to do it, since all of the image light passes through all of the filter. The device would have to resemble a "sharp mask"; something that adds a variable neutral density to a focused image just ahead of the sensor. One could, I suppose, envision a system that focuses the image on a masking element (passive, like variable sunglasses, or active) that you would then shoot through, but then the primary lens would have to be mounted on that device and a secondary lens would be needed between the camera and the device. Sounds like money and weight to me. –  user2719 Sep 14 '11 at 16:10
    
For reference, this question photo.stackexchange.com/questions/15363/… deals with why a Neutral Density filter does not reduce DR. –  Sean Sep 14 '11 at 17:18
1  
@Dreamager - thanks for reminding me why I hate HDR... –  ysap Sep 15 '11 at 1:11
    
@ysap No kidding - the video starts out by showing us what "properly exposed" looks like for the dark areas and light areas. No, "properly exposed" means black is black, and white is white, not middle grey! –  Evan Krall Sep 15 '11 at 6:14

3 Answers 3

up vote 7 down vote accepted

Tiffen ultra contast filters were developed for this. Look it up; they got an Oscar for it. (My bad, wrote Low contrast. Edited)

share|improve this answer
    
Some example images I found: forums.dpreview.com/forums/… –  Sean Sep 14 '11 at 23:42
    
Wow, thas is unexpected and slick! I wonder if, due to their claim that it works "by spreading the ambient light throughout the entire scene to lighten all shadow areas", there is no effect of softening of the image (inspite of their claim). Unfortunately, I could not find on their website more detailed examples, but thanks @Sean for the forum link. –  ysap Sep 15 '11 at 0:57
    
So, further examining the sample images, it seems that one definitely does not want to use it were the scene is not problematic in the first place (like the 1st image). Then, using the filter were large portions of the image include non-problematic subject (like the 2nd image) does no good either (although some details in the sky are recovered, the whole image is degraded. The last image is a total mess. –  ysap Sep 15 '11 at 1:09
    
That said, it may allow for recovering more details in post-processing, where playing with the tone curve and/or adding local contrast will make the images look more pleasing. –  ysap Sep 15 '11 at 1:11
1  
You could probably get the same effect for a lot cheaper if you just buy the cheapest UV / NC filter you can find. Alternatively, breathe on your lens a bit. –  Evan Krall Sep 15 '11 at 8:53

A polarizing filter does this in a very limited way. If the strong light (highlights) are strongly polarized the polarizing filter can reduce the dynamic range.

share|improve this answer
    
+1, although, I guess it is true only for specific situations. I can imagine the opposite, where the shadows are polarized and thus increasing DR. –  ysap Sep 15 '11 at 1:00
    
Actually, that may not be a problem, ysap. If the shadows are polarized, adjust the pola filter to let them straight through with minimal light-loss in the filter. Unpolarized highlights are cut by approx 2 stops by the pola. Instant DR reduction. (I know they do something like this to emphasize reflections sometimes) –  Staale S Sep 15 '11 at 10:04
    
@Staale S - are you saying that the 2 stops cut in highlights will not affect the shadows? Polarizers do cut the light intensity of the image, but AFAIK it is done uniformly across the intensity spectrum. The polarization filtering is on-top of the regular light loss. –  ysap Sep 15 '11 at 15:19
1  
That is exactly what I am saying, ysap. Polarizers are like window-blinds, in essence, and cut off light that is not aligned with the slats in the filter. Normally this is used to cut off reflections, by rotating the filter so that the slats do not match the polarization of the reflections. Most "normal" light does not match the slats either, which leads to the 2-stop light loss. But if the reflections match the slats... they are let through almost unimpeded. And non-reflections are still cut by 2 stops. –  Staale S Sep 15 '11 at 16:12
    
@Staale S - OK, I see what you mean. It makes sense but I need to grasp it... –  ysap Sep 16 '11 at 21:26

No such thing has been invented yet.

Although Fuji has made several attempts to solve this problem including the famous SuperCCD SR used in the S5 Pro. The fourth-generation version places small low-sensitivity photosites between standard sensitivity ones, essentially capturing exposures simultaneously. Then software in the camera blends these two into one.

Following that, they introduced the SuperCCD EXR which has a uniform array of diamond-shaped pixels and can read half of them throughout the exposure to catch them before they saturate. The second half of pixels are read at the end of the exposure and blended with the first half to produce an extended dynamic-range image. As you can imaging this version, while easier to fabricate, imposes lots of restrictions on sensitivity and flash use.

share|improve this answer
3  
Unless I'm mistaken, it sounds like Fuji's devices are attempting to extend the capturable DR - not reduce the incoming. Which is a better solution to the same problem, but not really a 'reduce device', right? –  rfusca Sep 14 '11 at 15:43
    
Well, it depends the way you look at it. It captures more DR in order to compress it just like in the sound-compressor example, you need a component that can accommodate the entire range in the first place. –  Itai Sep 14 '11 at 19:27
    
+1 Probably the other side of the coin, but my question really aimed towards a filter-type solution, which surprisingly does exist. –  ysap Sep 15 '11 at 1:03

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.