Scenes with very high dynamic range (containing very bright and very dark sections) are difficult to capture in a single shot without either blowing out the highlights or underexposing the shadows. Although common sensors in current DSLRs have 7 to 8 stops of dynamic range, many natural scenes would normally have a much greater dynamic range, too large to be captured accurately, even though they would look perfectly normal to our eyes. This is because each photoreceptor in the retina can adjust its gain independently (or almost independently) of each other, while an ISO setting in the camera applies equally to all pixels in the sensor. HDR imaging is a way to extend the dynamic range of a capture by taking sequential exposures of the same scene with bracketed settings. High-end DSLR cameras can do this automatically after dialing in the desired bracketing sequence and interval. Typically one shot is taken at the “correct” exposure indicated by the metring system, while additional shots are taken above (say +1, +2 or +3 EV) and below (say -1, -2 or -3 EV). The trick is to get all highlights and shadows correctly exposed in at least one of the exposures in the series. The images in the series are then combined to generate an HDR image using specialized software. Later editions of Adobe Photoshop can do this. Dedicated HDR software, such as Photomatix, is however easier to handle and yields very nice results.
Here is one classic Pompei shot with Vesuvius in the back. Midday sun, crowds all over the place, too bright sky, poor contrast:
Cropping (in Camera Raw) allows focusing on the center temple and eliminates the worst part of the crowd:
Here is the Raw conversion: mainly tone curve for added contrast, graduated filter to pull back the sky and adjustment brush to bring back top of the building and columns that were too dark after the grad filter:
Here are the histograms before (left) and after (right):
One final clone tool in Photoshop removes the unwanted visitors. Voilá:
Posted in POTN here.
I am often asked why I shoot my photographs in RAW format. RAW capture refers to the direct transfer of the information acquired by the sensor of a digital camera to the memory card without any in-camera processing. In the 18-megapixel Canon EOS 7D, this translates into files of 25MB, compared to the 6MB of a high-resolution JPG-compressed file. Why would one like to shoot RAW files? The RAW format contains all the information captured by the sensor and is therefore most amenable to corrections of exposure, saturation, chromatic aberrations and noise during post-processing. Compressed JPG files contain a reduced amount of information and so are much more limited to adjust during post-processing. Why would someone want to adjust a photograph? Shown below are three examples taken during a recent trip to the Otavalo valley in Northern Ecuador. In all cases, the top image is straight-out-of-camera, while the lower image is after conversion in Adobe CS4 Camera Raw (no Photoshop in any of these examples!).