Contrast and Dynamic Range

[marcg]

Advertisement (gone after registration)

I've been taking photographs for a long time (note, I don't call myself a photographer) and I was suddenly struck by a thought/question.

 

Is it possible to say that dynamic range is the opposite of contrast in some way?

 

It seems to me that dynamic range is about all the steps between maximum dark and maximum light and the more steps there are then the more of those steps there will be and of course the more subtle will be the gradations.

 

On the other hand it seems to me that contrast is designed to emphasise differences in light and dark and to reduce the number of steps.

 

I'm amazed that I have come up with this question – and I really should know or understand the answer by now.

 

If people would like to contribute to this thread, please do so in easy language – not highly technical geek-speak.

 

Ta

Edited October 31, 2017 by marcg

[hoppyman]

But the dynamic range has fixed lowest and highest values yes? 11 or 14 stops or whichever. That’s just a function of the sensor and processing, not to do with higher or lower contrast due to the lens, light, subject and exposure decisions. . A lower or higher contrast image will fit somewhere in that range and the distribution of the values will vary but the theoretical dynamic range will not.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[pico]

A couple of terms might help. Subject contrast and micro contrast.

 

Subject contrast describes the overall contrast. Successful analog color photography, usually commercial, limited subject contrast so that it fit within the impoverished range of color film. Micro contrast concerns detail, and today's better digital equipment can exceed what you see in a print at normal viewing distance, and it has an effect we enjoy as we move closer, or print larger.

 

A side thought - today's LCD displays will one day seem rather primitive and we (not me) might wonder how we could ever be happy with the pixel density we accept today. (By 'not me', I mean that some high resolution imaging is disturbing. I mean I don't want to see the edges of poor makeup, beard stubble, all that. Imagine seeing a close-up of Phyllis Diller in super high rez.)

[Jeff S]

Imagine seeing a close-up of Phyllis Diller in super high rez.)

In her current state, I'd rather not see even a low rez pic.

 

Jeff

[marcg]

Thanks, that is extremely helpful.

 

Especially the explanation below has pretty well nailed it for me.

 

But the dynamic range has fixed lowest and highest values yes? 11 or 14 stops or whichever. That’s just a function of the sensor and processing, not to do with higher or lower contrast due to the lens, light, subject and exposure decisions. . A lower or higher contrast image will fit somewhere in that range and the distribution of the values will vary but the theoretical dynamic range will not.

[UliWer]

Certainly the dynamic range of a sensor does not change by a lens or the contrast of a subject.

 

Though hoppyman's description that an image "fits somwhere in that range" may give an idea of a certain interdependance of contrast and dynamic range. 

 

If your motive shows a lot of light contrast and you use a lens with high transmission (often called "contrast" though it is something else than micro contrast described by pico above)  the dynamic range of your sensor may be too small. The image does not fit into the dynamic range of the sensor. Outblown highlights are the most frequent result where the image does not "fit" - or the dynamic range is too small.

 

Now you use a lens with low transmission ("low contrast") for the same subject with the same sensor - and it fits. Old uncoated lenses may help a lot in certain situations where the subject's contrast is very high. You can easily see the differences in the histogram: it doesn't hit the left and right borders with a "bad" lens. This gives a lush impression on many images with normal or low subject contrast - but it may save your image from exteme differences between light and shadow.

 

It may even help to open up your lens fully - it is less "contrasty" wide open than stopped down. Opened up the inherent optical imperfections of a lens may help to cut off some extremes of the contrast of certain subjects - so they may fit better into the  dynamic range of your sensor. 

 

The "glow" of some older and/or wide opened lenses may result from this effect.

Edited November 1, 2017 by UliWer

[jaapv]

Advertisement (gone after registration)

Excarbated by the fact that the high and low extremes of the Dynamic Range may not be the photographic limit of the sensor If the falloff at either end, but especially in the shadows is at a low angle, i.e. just a small bit higher than the noise floor for one or two stops, that will be unusable  for photographic purposes. If, however an exposure curve on a sensor with a limited DR is near-linear at its ends, one may end up with a larger exposure range than with a sensor with a huge theoretical DR.

[pgk]

Is it possible to say that dynamic range is the opposite of contrast in some way?

 

I'm amazed that I have come up with this question – and I really should know or understand the answer by now.

 

The answer to your question above is no.

 

But, you need to define precisely what you mean by contrast (scene contrast, micro contrast, lens contrast or what?).

[marcg]

The answer to your question above is no.

 

But, you need to define precisely what you mean by contrast (scene contrast, micro contrast, lens contrast or what?).

 

Aaaarrrrrggggh!

[pgk]

I see you've figured the problem.

[SilentShutter]

The dynamic rage ist the maximum a film or sensor can "catch" without loosing "details"

The contrast is the maximum of tones you have in your "motive" caused by light

If boths factors fit together your picture comes close to "reality"

 

But the missed point is "can a print" handle all this ?

Most of the time this is the bottleneck........

[adan]

Higher dynamic range, by including more "stops" of brightness that can be captured, will give the impression of lower contrast in the mid-tones. As marc's intuition tells him.

 

This was a "perception" issue when the M240 first came out, and its images (especially .dngs) straight from the camera seemed flat and dull (and lower-saturation in color, saturation being related to contrast) compared to the previous M9 images. Shock! Horror! (Go back and find the threads after the M240 issue for yourself).

 

I.E. a - device - that has zero stops of dynamic range will posterize the world into either black or white. The ultimate in contrast. Kodalith (for those who don't remember Kodalith: http://joefaraceblogs.com/wp-content/uploads/2011/08/K7.3.jpeg )

 

A device that has one stop of DR will produce black, white and one gray 1/2-way in between.

 

A device that has 10 stops of DR will produce 10 gray steps, all separated by 1/10th the distance between B & W. And thus closer together.

 

HOWEVER - the missing point is that the higher dynamic range means that what constitutes "black" and "white" are also farther apart with higher DR - in a low-DR image the sky, for example, ALL goes to white very quickly, whereas in a higher-DR image, the sky does not blow out as quickly, and the shadows do not block up to pure black as quickly. The grays of the higher DR reveal more shadow and highlight detail.

 

FIlm has always had a non-constant contrast. Due to the realities of physics and chemistry, one stop of brightness does not create a consistent difference in density or final-image brightness, especially at the darkest and brightest ends of the scale. This is shown in a given film's "characteristic curve", whih is not a straight line (constant contrast) but an "S"-curve (which happens to usually have a nearly straight mid-section. https://www.intechopen.com/source/html/39070/media/image1.jpg

 

Most digital camera these days apply a "tone curve" in the camera (even in raw/.dng images) to whatever dynamic range they can natively handle, to better simulate the S-curve of film. Enhancing mid-tone contrast to avoid the pictures looking too grayed-out. The M10 does this (perhaps a bit too much) so that its high-DR images look "contrastier" than did the default M240 images, at least straight from the camera.

 

The mathematical application of a tone or contrast curve "breaks" the link between high DR and lower contrast - one can have both.

Edited November 2, 2017 by adan

[hoppyman]

The dynamic rage ist the maximum a film or sensor can "catch" without loosing "details"

The contrast is the maximum of tones you have in your "motive" caused by light

If boths factors fit together your picture comes close to "reality"

 

But the missed point is "can a print" handle all this ?

Most of the time this is the bottleneck........

“Down the rabbit hole”

Yes a print can only reproduce a fraction of the contrast range of a monitor. But then you are discussing how the tonal range of the image is fitted into that by all of the processes after capture.

Edited November 3, 2017 by hoppyman

[hoppyman]

Thanks, that is extremely helpful.

 

Especially the explanation below has pretty well nailed it for me.

I tried to answer just exactly what you asked but of course everything from the actual subject to an image of that is a very complex and large topic. I just deleted a draft paragraph where I was going to talk about more of that!

 

As a practical matter, taking the example of the S system that I use for nearly all of my studio photography, the system has I would think the largest possible dynamic range of Leica cameras, high contrast and micro contrast and high resolution .

In my experience, Anecdotally the SL lenses characteristics are as good, while the dynamic range is significantly less. Have heard quoted 11 stops vs 14. Remember that half of all tonal values are contained in the brightest stop captured. So the height of the ceiling matters, not just the space between there and the floor !

http://www.pbase.com/hoppyman/image/165881941

http://www.pbase.com/hoppyman/image/164851715

Edited November 3, 2017 by hoppyman