Monday, October 8, 2007
Exposing For RAW
There are special considerations to take into account if you're shooting RAW and you want to be sure that you're getting a proper exposure
On a recent photo tour down the Amazon River with a group of photo students, I was rather surprised to see that several had set their exposure compensation for -1 stop because the color and tone rendering on the camera LCD “looked better.” These students, all shooting RAW, were conducting the opposite of the ETTR technique, introducing more noise and less data, to produce a better-appearing LCD preview.
During this expedition, I decided to shoot a number of bracketed exposures and see if there was any correlation between proper exposure, based on ideal highlight data, and the appearance of the images on the LCD. Unfortunately, at least with my Canon EOS 5D, the “best” exposure often produced the worst-appearing LCD previews. Other than the possibility of truly clipping highlight data in RAW, this ugly LCD preview is one of the biggest disadvantages of utilizing the ETTR technique. You may have some camera controls that affect how the histogram and clipping indicators are plotted on your camera's LCD. Many cameras have “picture styles,” which alter the JPEG rendering and therefore, the LCD previews. Testing is necessary to adjust these picture styles while viewing highlight clipping on your camera's LCD to more closely match the resulting RAW data clipping if present. Such settings may provide closer clipping previews, but don't expect a match to the real clipping of the RAW data. Plus, these picture styles will affect JPEGs you may shoot, but not the RAW data—not useful for those who wish to shoot RAW+JPEG.
When I set my 5D Picture Style from the default to -4 contrast, the clipping of highlights on the LCD didn't appear until I had “overexposed” 2⁄3 of a stop. This is better than the original default settings, yet in actuality, I was able to overexpose 1½ stops beyond what my meter suggested while fully retaining highlight data in my RAW file. The clipping indicators are still far from correctly describing what's happening to the RAW data.
Testing ISO And Exposure
Figure 2 shows a number of exposure and white-balance targets, evenly illuminated by two Balcar strobes within 1⁄10 of a stop in all four corners. A Sekonic L-758DR handheld flash meter was used to measure incident light for a base (normal) exposure. Images were shot first using this “normal” exposure for ISO 100, 400 and 800, then -1 stop and +2 stops overexposed in half-stop increments. I bracketed exposure by altering the strobe power rather than the ƒ-stop, which introduced too many visual variables caused by depth of field and apparent lens sharpness. I was unable to test at ISO 1600 or underexpose at ISO 800 because of the power of the strobes. However, the three ISO settings do provide some useful information about ETTR on the RAW data.
Once I had a series of bracketed exposures, I examined each in my RAW converter of choice, Adobe Lightroom. The image exposed as recommended by the flash meter did appear properly exposed using the default Lightroom rendering settings. It was no surprise that the over-exposed images looked, well, over-exposed (Figure 3).