Jump to content

Lens Contrast vs Dynamic Range


Recommended Posts

Advertisement (gone after registration)

For long time, I was asking myself a question - how lens contrast affects dynamic range.

More contrast lenses = less details in shadows or bright areas. And as we all know - producers fight for better dynamic range of their cameras.

 

So is once against the other?

 

I've found quite interesting test as start for discussion:

Lens contrast for digital (1) – using 28mm rangefinder lenses | photostream

Lens contrast for digital (2) – using 28mm rangefinder lenses | photostream

 

He says differencies are suprisingly not huge, but less contrast lenses give more details in shadows.

So how it is?

Link to post
Share on other sites

I was just shooting today with a range of new and old lenses in bright low-angle sun, and looking at the shots tonight the new and contrasty 35 Summicron ASPH clearly had harsher tones that required a lot more "fill light" in post-processing to bring up the shadows - usually to the detriment of the overall brilliance of the picture. Compared to my sweet old 21 pre-ASPH.

 

I'll concede that comparing a 35 to a 21 is not ideal, so I don't make a "scientific" claim for those results.

 

I would also say that the real difference was not so much the shadows per se, but that the old 21 has a very long shoulder (highlights) - which meant I could give more exposure before blowing highlights - which meant more real exposure (not just fogging or flare) in the shadows.

 

I would also note that the tests you link to are using a Voigtlander 28 Ultron as the "old, low contrast" lens - and I don't know that that is a representative sample of ALL lower-contrast lenses (and is certainly not especially "old", since it was designed since 2000). He also does not say WHICH older Elmarit he was using, and the v.4 1992 lens is not especially different from the 'cron or current ASPH Elmarit in contrast, whereas I find the 1980 28 Elmarit to be very close to the 1980 21 in smoother tonality.

 

Finally, I'm not sure shooting a 2-dimensional grayscale is a good way of measuring this effect. A series of gray patches with no internal detail is not the same as shooting a 3D object lit from one side with shadows on the other, and with a texture casting its own microshadows within each area of tone. And with bright specular highlights that may or may not be equally controlled for.

 

Here's what I got measuring two images shot close enough together in position to show what I saw. 35mm 'cron ASPH full-frame, and a crop from a 21 pre-ASPH image (the one with the labels) covering the same vertical area.

Both shots were given identical RAW processing settings. Exposure was normalized for point "A" (17% density). I then measured 9 highlight and shade areas (labelled A-I) with the eyedropper in Photoshop.

 

..........21...........35

C.......9%...........9%

F.......9%...........9%

A......17%.........17%

H......70%.........75%

G......76%.........81%

E......79%.........85%

I.......89%.........91%

B......90%.........94%

D......91%.........94%

 

The highlights are equal in brightness, but the 3/4 tones get darker faster with the contrastier modern 35. Which leads to compression of the deepest shadows: samples B and D are indistinguishable with the 35 but still slightly separated with the lower-contrast 21. (Again, ideally this would have been 35 vs. 35 or 21 vs. 21, but I think the lighting and framing and such are close enough to be suggestive)

 

I leave it to the subjective eye of the beholder whether there is more visible separation or openess in the shadows of the colored foil garlands in the 21 or 35 shot.

 

But I guess my conclusion would be that there is a difference in the pattern of tonality, even if both lenses get to "black" about the same time.

Link to post
Share on other sites

Opening my heart to the benefits of lower contrast lenses has been one of the great discoveries in my photographic journey. Brought up to believe in 'contrast is good' and that max dynamic range was essential and my Wacom Cintiq pen was always ready to drop a few points on a curve, I was horrified to see the mess I was making when I started doing this to the Noctilux f/1 images on my M9. It just made every one of them worse and I ended up dragging the adjustment layers into the bin.

 

I then realized part of the beguiling character of the lenses was the sense of airiness the lower contrast affords an image, which got me to look around and notice that often the world is not afforded steep S curves. The image of one of my Beagles here works so well in this low contrast, low exposure zone; but became a harsh mess when I gave it a proper black and white point.

 

Even the image of my daughter making waffles has such on open centre that higher contrast would rob its inherent ease.

 

I am glad this was brought up because I thought I was alone.

 

Ian

Link to post
Share on other sites

My personal understanding of this: The used dynamic range is connected to the image contrast at time of image processing. The scope for these two variables are the total available dynamic range (white to black in stops) and the bit depth of the recorded values. The bit depth is limited by the noise floor. This is where the noise skips the bits, so that the recorded information is random, rather than actual information.

 

The problem is, that many people confuse those values, e.g., used and total available dynamic range. Actually i'm not even sure that the manufacturers want to improve the dynamic range. Recent cameras introduced higher bit depths for raw images, which implies they'd rather provide the image quality for higher contrast images.

 

Now how does lens contrast benefit to this situation? Depends on what you want to improve.

Link to post
Share on other sites

Thank you Steve for starting an extremely useful thread. Thanks, also, to Andy Piper for posting a truly insightful analysis.

 

Older photographers probably grew up with lower contrast lenses and accepted them as 'state-of-the-art' products for their time. I have been agonizing whether to trade my 1960s 50mm Summicron for an allegedly superior modern version. I am convinced that the reason I haven't is the realization that older lenses have an unique quality impossible to find in a modern counterpart. Aren't we lucky to have such a wide choice for our craft?

 

I have recently bought a near-mint 135mm Elmar lens for my M8. While in performance terms it almost certainly cannot match its two successors, I am finding it perfect for some of my work and aim to capitalize on its relative lower contrast. This discussion helps to reinforce that judgment.

Link to post
Share on other sites

Advertisement (gone after registration)

In the past century, when trying to manage a N+++ scene (remember the Zone System terminology?) with a view camera and modern multicoated lenses, one possible way of 'pushing up' the shadows was to partially cover the interior of the bellows with a white material, in order to increase the bouncing of light. I don't have the text at hand, but Ansel Adams wrote a nice explanation in one of his handbooks. In it he mentioned also "older and newer lenses", being the older, non coated ones, better in some cases for those N+++ situations.

Another method for coping with those contrasty scenes was pre-exposing the film...

Link to post
Share on other sites
Guest maddoc2003jp
In the past century, when trying to manage a N+++ scene (remember the Zone System terminology?) with a view camera and modern multicoated lenses, one possible way of 'pushing up' the shadows was to partially cover the interior of the bellows with a white material, in order to increase the bouncing of light. I don't have the text at hand, but Ansel Adams wrote a nice explanation in one of his handbooks. In it he mentioned also "older and newer lenses", being the older, non coated ones, better in some cases for those N+++ situations.

Another method for coping with those contrasty scenes was pre-exposing the film...

 

That would explain why the 35mm Summilux-M pre-ASPH with lots of coma (and therefore diffused light) is quite useful at night photography.

Link to post
Share on other sites

Thank you all for comments!

 

I am new to "M" world, and still realising that sometimes it may be desired to have more than one lens for single focal. Classical vs clinical. Depending on needs. Sometimes also fast and slow. I didn't have such matters in dSLR world

Anyway, in practice I try to minimize number of lenses.

 

Maybe one more question, what you think.

 

Having less contrasty lens - you can always add contrast in RAW developer soft.

But from opposite direction - not so straightforward. Not always you will be able to add fill light enough, and doing this - increases noise.

Link to post
Share on other sites

I have asked related question to our colleague, who shared his photos taken with Voigtlander lenses here: Depo

 

Forum thread is here: http://www.l-camera-forum.com/leica-forum/leica-m9-forum/116754-some-cv-lens-samples-m9.html

 

My question is here:

http://www.l-camera-forum.com/leica-forum/leica-m9-forum/116754-some-cv-lens-samples-m9-2.html#post1265262

 

My colleague observes, sadly to him, that most samples he can see from M9 - are too contrast, look like from compact camera JPG program (in terms of color and contrast only of course)...

No details in shadows. Black whole...

 

He says, that most samples he can see from M8 and X1 - do not look like...

 

Any idea why that could be? Is there anything specific to M9?

Or owners are different group, with different taste?

Is it difference between C1 and LR?

 

PS: Or look at below samples:

Google Tłumacz

 

Do you think it is natural how we observe the World?

Edited by Jerry_R
Link to post
Share on other sites

Small update.

 

I have analysed several RAWs in LR and observation is: DNG after opening in LR is very very contrasty. Of course it is matter of taste and very often it is desired effect. But first sight is not reminding photos from analog time. Is not reminding reality.

 

Changing tone curve from Medium to Linear, or keeping Medium and changing default contrast 25 to 0 - brings more natural, raeal look like.

 

I didn't had to do that with other cameras. Often, if I would do this with other camera, result would be dull. With M9 is... Natural.

 

How it looked with M8? I know you probably used different RAW developer - C1.

Don't you feel, that too many users do not realise over contrasty "starting point" in LR?

 

I also will have a look at JPGs soon, so far, I only used RAWs.

Link to post
Share on other sites

Been my experience lower contrast lenses do not capture more DR. In fact the opposite is true. The zone 1 or 2 dark shadows come up the scale and end up being rendered as say zone 3. Everything above zone 8 blows out.

 

Take a coated lens and uncoated lens and shoot the same full scale subject, same time, same place, same light, and compare. I have and the above is what I got.

 

A Adams used to downrate film speed for his uncoated lenses so he evidently found the same thing to be true.

Link to post
Share on other sites

Here is a comparison of two extrema - both with M8, 1/4 sec., UV/IR-Filter, JPGs from LR

no further processing:

 

1.75mm Summicron ApoAsph f:2 (don't care for the estimation of aperture in LR...)

 

 

 

 

2. 7,3cm Hektor F:1,9, uncoated:

 

Link to post
Share on other sites

Now again with M8, 1/3 sec, no UV/IR, JPG from LR, no other processing:

 

 

 

Summitar 1939, uncoated, f:2:

 

 

 

 

 

Summitar 1950, coated, f:2:

 

Link to post
Share on other sites
Been my experience lower contrast lenses do not capture more DR. In fact the opposite is true. The zone 1 or 2 dark shadows come up the scale and end up being rendered as say zone 3. Everything above zone 8 blows out.

 

Take a coated lens and uncoated lens and shoot the same full scale subject, same time, same place, same light, and compare. I have and the above is what I got.

 

A Adams used to downrate film speed for his uncoated lenses so he evidently found the same thing to be true.

 

Contradiction here - if the problem with uncoated lenses is blown highlights, then downrating the film (using a lower effective ISO) will increase the exposure and blow them even more.

 

It is certainly true that when I switched from older, low contrast Nikkors (60's-era) to higher contrast Canon lenses (70's-era), I had to downrate my film for the HIGHER contrast lenses to get back shadow detail and then lower development to keep the highlights under control. That was the first time I really became aware of the effect of lens contrast.

 

Shooting with the Contax G lenses, I had real problems with chalk-and-charcoal images, and lowering the overall contrast through exposure and development, and/or using a reverse-S curve in photoshop (raise shadows, lower highlights) muddied the mid-tone tonal separation as the tone curve flat-lined across the middle. One of the main reasons I switched to Leica - trial negs with 21, 28, and 90 pre-APO/ASPH lenses practically scanned themselves, with detail everywhere.

 

Edit: you are correct about Zone 1 and 2 coming up the scale with lower-contrast lenses, though. That's the whole point of preferring them. I don't want the lenses clipping my shadows, as the 75 APO does in Uli's first example. That wall of tones up against the left side of the histogram represents lost shadow details. Sharpness aside, I want a base histogram more like the 73mm Hektor - then I can decide where the clipping will take place - if at all.

Edited by adan
Link to post
Share on other sites

I will clarify, zones 8/9/10 merge into one and all detail is lost.

 

Don`t hold me to exact zones because i did this film years back. PS give more ability to measure and clarify. The pics posted tell the story.

 

Generally there is a big jump in contrast from non coated to single coated and only a none slight improvement with multi. It is hard to compare as multiple coating and singles are rarely done to the same lens design so other factors come into play.

 

The two Summitar shots are perfect examples.

Link to post
Share on other sites

My biggest observation is that after opening RAW in LR - 1st action can be changing tone curve from Medium to Linear. Picture taken with recent lenses looks more natural the.

 

In past, with other cameras I had - it would cause dull, not natural photo. But not with M9.

 

I wonder if:

- the M8 RAW in LR also is so contrast by default - or Adobe profile for M8 is less contrast?

- the same M9 RAW opened in Aperture is also so contrast?

- how it works with JPGs - but this I will test myself soon

Link to post
Share on other sites

Lens contrast and how it relates to dynamic range is often dependent on lighting ratios. Lighting ratios are often the key factor to determining a photographer's preference for higher contrast or lower contrast lenses. Lighting ratios might also have a lot to do with why certain photographers feel the need for more or less dynamic range.

 

I'll try and explain a real life scenario that you should all be familiar with to try and help illustrate the point:

 

First, lets assume that a photographer is taking a portait of a human subject. The subject is standing on a green grassy meadow and it's a cloudy day. The main light source for the subject is coming from the cloudy sky above, and the fill light for the shadows is actually coming from light that is bouncing off the grass below the subject. If the green grass of the meadow is very dark, then it is actually subtracting light which is going to make for a darker shadow on the subject. This type of lighting ratio will probably makes deep shadows under the subject's eye sockets which would often be considered unpleasant for portraiture. In this particular type of lighting ratio scenario, a photographer might want more dynamic range in order to fill in detail in the shadows around the eyes. Or he may also want to use a lower contrast lens in order to prevent the shadows from going too dark. However, his choice of lens contrast or dynamic range and their relationship to each other is mostly dependent on the lighting ratios present in the scenario.

 

Next, imagine the exact same portrait scenario that I described above with just one difference. Imagine that the subject is standing on a meadow covered with white snow instead of green grass. The main light source for the subject is still coming from the cloudy sky above, and the fill light for the shadows is still coming from lighting bouncing from beneath the subject. However, the pure white snow is actually bouncing a lot of extra light into the shadows. In fact,, the white snow might actually bounce so much light into the shadow portion of the scene that the subject doesn't have any deep shadows at all beneath his eye sockets. In this type of cloudy day/ snowy meadow lighting ratio scenario, the photographer might not feel the need for more dynamic range or less contrast from his lens in order to balance out the shadows on the subject's face.

 

I've heard a lot of people around here describe rainy day lenses and sunny day lenses. I think what they are actually describing is lighting ratios. Whether or not a photographer feels the need to use a higher contrast lens or might feel that additional dynamic range is necessary probably has more to do with the lighting ratios that he's photographing more than anything else. This doesn't mean that lenses and DR are inconsequential, it just means that their choice often has more to do with lighting ratios than anything else. This is just my experience and others are certainly free to disagree.

Edited by Gentleman Villain
typo
Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...