Tuesday, October 12, 2004

Use the dynamic range

Photographers talk about Adams’ Zone System when they often mean the dynamic range. His system contains 11 zones of image tonality (0-10), while the dynamic range contains five. The Zone System is cool for art photographers, but is less useful for PJs and the general public. Most PJs use a dynamic range for daily work.

I’m not bashing the Zone System. It’s a wonderful system for those who have the time, knowledge and inclination to use it. However, it’s not practical for PJs who have one hour to drive through snarled traffic to a scene, get the shot and transmit to make deadline.

The Zone System
For those unacquainted, the Zone System operates along the standard dynamic range (see below) with the addition of photographer controlled film exposure, development and additional tinkering while printing on a calibrated system. Frequently, a photographer is required to push or pull film to constrict or expand the tonal range of a single frame. Additionally, a deliberate base fog on the film may be required to open up shadow detail while reducing development time to restrict highlight density.

I don’t suggest PJs try to lay down a base fog on a digital exposure or dunk their dit into a bucket of Dektol or T-Max to control development. It will control the development, but not the way anyone wants.

Dynamic range
A good guide for most photographers is the dynamic range. It consists of five stops of light. The dynamic range shifts light from white to black and vice verse anywhere along this range. Furthermore, the photographer sets a point along the range for each tone by deciding which tones to let fall away.

On the dynamic range, these are the five tones:
1. white with detail
2. light gray
3. middle gray
4. dark gray
5. black with detail

A simple explanation is at the end of most people’s arms. Look at your hand. Imagine your middle finger is middle gray. In stops of light, this example makes your pinkie white and your thumb black, both with detail information. Consequently, if you want to turn your pinkie black instead of white, you must stop down four stops. Conversely, if you want your thumb to appear white, you must open up four stops.

I know, it's a strange example, but wait until the next time you want to photograph a swan and you catch yourself looking at your fingers to find the right exposure. ;-}

This exposure control can be accomplished with the aperture, shutter speed or film ISO. The camera doesn’t care which adjustments are made, but it will require manual control.

We understand how a camera’s reflective meter works. A camera meter tries to make a general scene middle gray. If the camera is pointed at a tire, it will make the black tire gray (overexposed). If the camera is pointed at a cloud, it will make the white cloud gray (underexposed). If the camera is pointed at a small fire at night it will make the fire solid white and most of the rest solid black. Joy.

Some of the high-end professional cameras use complicated algorithms and RGB multi-sector metering to avoid this basic problem, but it’ll do the same darn thing more often than not.

Therefore, it’s important to know how to out-think the meter. With digital cameras, you can "chimp" (look at the display screen) and adjust. However, it’s better to hand meter the scene because the viewing screen can be off by two stops. We’ve discussed several ways to ensure the correct exposure.

Choose what to keep
We’ve metered the light and set the correct exposure. We've even double-checked the exposure with one of the alternative methods. However, the image isn’t getting the color saturation or pop we want. Sometimes, we simply don’t like the tonal shades in which we’re working. What can we do? Shift the tonal range.

In flat, even light, there isn’t a problem. A proper exposure yields blacks and whites with texture. However, when the light is raking the subjects from one angle or the light sources are dim or unevenly distributed, the PJ must make some choices. Something isn’t going to fit within the standard five-stop variance.

The tonal range can be shifted with any of the three methods listed above. Some make more sense than others depending on the subject matter. The point is to find which parts of the tonal range to keep and which to let fall away. The ends of the dynamic range (detailed blacks and whites) are the sacrifices. The whites can either blow out or the blacks can become voids. However, the right decision will save the overall image. The PJ must choose what to save and what to let go.

An extreme example
Most of the time, PJs are dealing with people as subjects. We encounter every possible skin tone and must make quick decisions under often stressful circumstances and incredibly bad lighting.

Consider for a moment the exposure for a person handcuffed on the street at night with police lights flashing behind the subject while a spotlight is on the subject’s face. What’s the correct exposure?

First, switch to spot meter on the camera’s reflective meter. Next, color balance for the tungsten light of the spotlight. Now it gets interesting.

If someone has particularly less-reflective skin, I’ll lean toward overexposing actual tones by one stop (-2 + 1 = -1 from middle gray). I’ll do the opposite if they have overly reflective skin and hair (+2 – 1 = +1 from middle gray). I still get the same five stops of dynamic range, but I have set which five I want to capture.

With less-reflective skin, I’ve moved the black zone to dark gray. I understand this takes all the detail out of everything four stops lighter, but it’s my choice. With highly reflective people, I’ve added some texture to their skin and hair, but black T-shirts lose texture and becomes a dark voids. Again, this is my aesthetic choice rather than a meter reading.

In most cases, the red and blue police lights will accumulate on the CCD or film and become glowing white. Meanwhile the surrounding area, which is not in a spotlight, will drop down to black. So, I’m stuck with a relatively proper exposure of a suspect’s face while everything else in the scene pushes off both ends of the dynamic range.

Since I know I’ll be facing an editor, I’ll also make an overall scene exposure. This increases the saturation of the police lights, but drops the subject’s face back into darkness with the rest of the scene. However, this might be useful to set the mood of the scene or conceal the subject (if a minor). Yet again, this is not setting the scene to middle gray, but setting the lights to middle gray to saturate the color.

Handling less extreme dynamic range choices
Sometimes, a subtle decision must be made about the dynamic range. Light raking over a subject will make the PJ decide between highlight and shadow detail. Only one can be kept if the light has more than five stops of exposure latitude.

Each subject will typically dictate the best decision. If there is something interesting within the shadows, shift the range to keep the shadows and let the overall image become high-key. If something interesting is in the highlights, keep the highlights and let the shadows drop.

Dramatic dynamic range shifts
Occasionally it’s interesting to give a scene a little push. In these cases, PJs may deliberately set a tone to something other than the standard. Digital cameras allow immediate feedback at least. If it doesn’t work, there’s no harm done.

Some common examples of dramatic use of the dynamic range include silhouettes and high key portraits. When combined with a strobe or flash, it gets even more interesting.

A silhouette is nothing more than setting the shadow below the 1st step of the dynamic range. When middle gray or dark gray are set on the light source (the sun for example), the color saturation will increase while dropping other subjects and environments into two-dimensional blackness.

To make silhouettes pop, a primary subject with a strong shape is in sharp focus and normally a shallow depth of field. This defines the outline while minimizing competing elements within the frame. A giraffe in a barn would make a good silhouette.

High key
High key images are commonly used with fashion and bridal portraits. Like the silhouette, it's frequently shot toward the light source (an open window for example), but middle gray is set for the subject’s shadow detail and the background is allowed to blow out to white.

High key images are frequently "soft" because the subject’s edge detail is blocked up. Typically more depth of field is used to make these images and creates slight movement in the subject to further soften the overall appearance. A bride in a white gown would make a good high-key subject.

Combo dynamic range
Once the general principle of the dynamic range is understood, it can be combined with stobes, flash or even reflectors to make stunning results. The PJ can prepare a portrait for a silhouette scene and selectively light the subject within the silhouette framework. Likewise the PJ can prepare a high-key portrait and still maximize foreground detail.

Silhouette combo
The image in my mind is by Tom Fox. He made a portrait of a jockey on horseback inside a barn. The horse and jockey are silhouetted against a well-saturated blue sky and framed in the doorway of the barn. Instead of leaving it as a featureless silhouette, Tom snooted a strobe and lit only the face of the jockey to match the (probably -2) exposure of the sky.

In other words, he completely ignored the actual dynamic range of the scene, moved the range to have a well saturated sky, and then placed the jockey’s dynamic range within this exposure framework. The end result was an excellent image which used the power of a silhouette with the detail of the jockey’s face. It immediately told the readers who this person was in a dramatic way.

High key combo
This light combination can also used with the high-key scene. To get a soft white background and still retain maximum detail in a white wedding dress, PJs will use a strobe to balance the front light.

Measure the foreground light at middle gray (properly exposed). If the subject is a bride or some other subject wearing white, the PJ might want to set the white to light gray for extra detail.

Once the primary light is established, boost up the background's available light to blow out one or two stops above white. In other words, remove the background texture so it doesn’t compete with the subject. This is typically done by increasing exposure time because it’s the only loose variable once the strobe exposure is set. The result is a bride with full detail in her dress with a pleasant white glow engulfing her.

Enough for now,


E said...

what an epic post! Something one becomes capable of at 2 am, I suppose. :). Hope you're well!

Mark M. Hancock said...

Photo 102 in a few paragraphs. This is the crux of most photography. This was the most condensed I could get it. I'll write fluffier posts for a while until this one absorbs. :-)

Michael Rubenstein said...

That is a lot of information, your blog is an increadible source for everyone who needs to take great images.

Mark M. Hancock said...

Thanks. Now if I could only follow my own advice...

rannie said...

Awesome info.

Haven't commented lately, but I have been really enjoying and benefitting from the info you have posted lately.

Alot ofwhat you posted here I already kinda knew, but you definitely cleared things up.


Mark M. Hancock said...

Thanks Rannie. I know this blog isn't as useful for really advanced PJs, but every now and then it may remind them of a barrier they wanted to push.

jinut said...

hi, this is wonderful blog and it gives me lots of information.
But i am a bit confuse about a statement under "AN EXTREME EXAMPLE" where you said:
"I'll lean toward overexposing actual tones by one stop (-2+1= -1 from middle gray. I'll do the opposite if they have overly reflective skin and hair (+2 - 1= +1 from middle gray)."
i was confuse because you said OVEREXPOSING, while you mentioned -1 from middle gray. i thought overexposing should be +1 or +2 since you move the middle gray to dark gray so you have more white part on all dynamic range.

Mark M. Hancock said...


The first number is the skin reflectivity (-2 skin in proper light).
In this example, it's overexposed, but the meter says it's still underexposed. The meter makes everything 18 percent reflective, so you must trick the meter by telling it to do the opposite to get the right amount of light.
So, -2 skin reads as 0 (a correct 18% exposure) on a reflective meter. Instead of letting the meter overexpose by two stops, switch to manual controls, subtract 1 stop and then shoot. The -2 skin is now -1. Technically it's overexposed by one stop, but the meter still thinks it's underexposed by one more stop. Otherwise, the camera will overexpose by two stops.

jinut said...

wow! i got it! thanks a lot, keep posting, especially about light. thanks a bunch! great blog, i'll keep reading