Saturday, June 13, 2009

Watch out Ansel Adams, My first HDR photo


Ansel Adams is like the Michael Jordan of photography. His technical brilliance combined with his keen artistry was what made him legendary. One thing that he understood was that film had the ability capture a greater range of light intensity than could be reproduced on the print. In the darkroom, he could dodge and burn different areas of the image to reduce overexposed areas or bring out more details in dark areas. Today most of us are more than happy to turn the switch to full auto and let the camera do everything for us. This is unfortunate, because with just a little work and a little equipment, we can do things that would make Adam's jaw drop.

What is HDR? HDR stands for High Dynamic Range. In this case we're referring to the range of the light intensities across the image. We call light the intensity that passes through a specific area luminance. Think of a sheet of wire mesh being held up in the air. If you measure the light going through one of the squares in the mesh, then divide that by the area of that square opening, you get luminance. Now imagine that you're looking at a shiny motorcycle at high noon w/ lots of chrome. The glint and glare reflecting off the chrome is blinding you but you can still see the texture of the street in the shadow caused by the bike. Roughly speaking the difference in luminance between these extremes is about 1 to a billion. HDR is basically the ability to measure, store, and display this wide range of light information.

So what's the problem? Everything. Our cameras, file formats, and displays are all limited in terms of dynamic range Our cameras will let you capture highlights while losing the shadows or vice versa. The digital image files we normally use only store about 256 different levels of intensity. Finally, our monitors and prints can't reproduce the same level of contrast that we experience in real life

How do I use HDR then?
We can solve each of these problems individually. Using a tripod, you can take the same photo at different exposures (I like to vary only the shutter speed). Some cameras even have a setting for this called "auto exposure bracketing". I usually do 3 shots at -2, 0, +2 EV. With these photos (in RAW format) I combine the luminance data into a single HDR friendly file format using a free program called Qtpfsgui. This solves the first 2 problems. Now we need a way to view these HDR images. If you had $49k to spare you could buy one of these or we can try to approximate this information on our LDR devices using a technique called tone-mapping. It makes some compromises to bring out the details in areas that would have been lost. With Qtpfsgui, you can do this part too.