High Dynamic Range #1 (HDR)

How digital convert the world into numbers.

This is the first in a series of articles that address the fascinating topic of High Dynamic Range or HDR.  HDR not only refers to situations you face when photographing in the field but also the digital darkroom techniques for processing HDR images.  This first article will lay the groundwork necessary for understanding HDR.

How often do you look at your photographs and are dissatisfied because the images appear washed out, or at least parts of then.  Maybe you took an outdoor family shot at a gathering or picnic.  The people may look great but the sky is washed out.  Or perhaps you were on vacation and something similar happened – part of the image was washed out.  The opposite could also happen.  The image could be too dark.

Normally modern digital cameras have very sophisticated built in light meters that give you excellent exposures.  But often the conditions of the scene you are shooting are simply beyond the capability of camera to capture, regardless of whether you’re shooting digital or film.

If you’re collecting pictures for the family album these defects may not be a problem at all.  It’s the memories that the photograph conjure that are important, not the technical merit of the image.  These are photographic records of important events in our lives.

But if you’re intention is to create a work of art, a washed out image is one sure way to frustrate your efforts.  If you’re shooting RAW you have some ability to recover washed out highlights or black shadows.  However, once you’ve completely lost your highlights and/or shadows there’s nothing you can do to recover them.

Colors are Numbers

Modern digital cameras are really computers.  They have internal memory, processors, electronic sensors and the firmware that gives them their extremely sophisticated functionality.  Some digital cameras such as the Canon 1D Mark III and 1Ds Mark III are so sophisticated that they actually have two processors.  They’re more powerful than desktop computers from not all that long ago.

We refer to the resolution of the camera’s sensor in terms of millions of pixels.  A pixel is a dot that records two things – the intensity of the light (referred to as value or luminance) and its color (also referred to as hue).  So the images captured on your camera’s sensor is really millions of dots.  But dots are used in a lot of places to make images.  For example, the images on the computer monitor on which you’re reading this post are composed of dots.  If you look at the screen with a magnifying glass you can easily see them.  A very common screen resolution is 1280 horizontal dots and 1024 vertical dots which gives a total resolution of about 1.3 million pixels.  Boy, if you were shooting with a camera whose sensor had a resolution of 1.3 million pixels you’d be pretty dissatisfied.  Most of us are shooting digital cameras with 8, 10 and more million pixels.

But these pixel numbers are actually understated – way understated.  We can easily get light intensity from a single pixel but to get color we need three of them – one for each of the three primary colors of red, green and blue.  So in truth, a pixel is actually composed of three much smaller pixels, one red, one green and one blue.

With three pixels – red, green and blue – it is theoretically possible to make all the colors of the spectrum,  just by blending them in different proportions.  For example, if we blend equal parts of red and blue we have magenta.  Equal parts of blue and green give cyan.  We can also blend unequal parts (more of one than another) to end up with different shades of color, different hues.  Also, if we have very small amounts of red, green and/or blue we end up with very dark colors.  Likewise, if we have large amounts of the three, we end up with very light colors.  In fact, if we have no red, green or blue we  have no light; in other words, we have black.  Equal parts of red, green and blue produce white if the three colors are all very bright or gray if they’re not so bright.

Now, since digital cameras are really computers, they store data in the only way a computers knows how to store data – as numbers.  The three red, green and blue pixels (by the way, from now on we’ll refer them by their shorthand designation of RGB) are represented as numbers.  These numbers can range from 0 to 255.  Zero is the total absence of the color and 255 is the brightest the color can be.  So the scale from 0 to 255 specifies the brightness of the color.

Computers represent colors with three numbers such as (127,64,191).  The colors are in RGB order so the first represents the brightness of red, the second that of green and the last is for blue.  Pure red would be (255,0,0).  Magenta, the combination of equal parts of red and blue might be (128,0,128).  Pure black would be (0,0,0) and pure white would be (255,255,255).  Gray would be anything between (0,0,0) and (255,255,255) such as (128,128,28).

That’s enough for one sitting.  In the next post we’ll look at how this numbering system relates to the real world and define clipping in terms of these numbers.  Don’t worry, it will all relate back to HDR by the time we’re done.

Other Postings in this Series

High Dynamic Range #2

(931)

Author: doinlight

Ralph Nordstrom is an award-winning fine art landscape photographer and educator. He lives in Southern California and leads photography workshops throughout the Western United States.

3 thoughts on “High Dynamic Range #1 (HDR)”

  1. Pingback: HDR Tip #1

We look forward to your comment.

This site uses Akismet to reduce spam. Learn how your comment data is processed.