Adobe Photoshop CS5 for Photographers - Martin Evening [175]
Right now there are certainly a lot of photographers who are interested in exploring what can be done using high dynamic range image editing. For example, by using the Merge to HDR Pro feature in Photoshop CS5 you can combine two or more images that have been captured with a normal digital camera, shot at different exposures, and blend these together to produce a 32-bit floating point, high dynamic range image. You can then convert this 32-bit HDR file into a 16-bit per channel or 8-bit per channel low dynamic range version which can then be edited further in Photoshop.
Basically, high dynamic range image editing requires a whole new approach to the way image editing programs like Photoshop process the high dynamic range image data. Because of this the Photoshop team had to rewrite a lot of the Photoshop code so that some of the familiar Photoshop tools could be made to work in a 32-bit floating point image editing environment. Photoshop now offers a limited range of editing controls such as layers and painting, and these features are available for all current versions of Photoshop CS5 rather than the extended version only (which was previously the case with CS3 and CS4).
HDR essentials
Traditionally, most camera sensors have been designed to record the light that hits the indivi dual photosites. I won't complicate things with a discussion of the different sensor designs used, but essentially the goal of late has been to design sensors in which the photosites are made as small as possible and crammed ever-closer together so as to increase the number of megapixels. Camera sensors have also been made more efficient so that they can capture images over a wide range of ISO settings without generating too much electronic noise in the shadow areas or at the higher ISO settings. The problem all sensors face though is that at the low exposure extreme there comes a point where the photosites are unable to record any usable levels information over and above the random noise that's generated in the background. At the other extreme, when too much light hits a photosite it becomes oversaturated and is unable to record the light levels beyond a certain amount. One way to look at this is to imagine photosites as being like glasses ready to be filled with water. When only a little water is added it can be hard to accurately measure how much water is in there, and if you fill the glass to the brim, the water will overflow and you'll be unable to measure any extra water that is poured (see Figure 7.1).
Figure 7.1 One approach to increasing the dynamic range of a camera sensor is to have two different sized photosites working together to capture the light hitting the sensor. One can liken this to having a wide rimmed glass and a narrow rimmed glass side by side. When water is poured, the wide rimmed glass is able to capture more water than the narrow rimmed glass and can easily detect small amounts of water, but when more water is poured than the wide rimmed glass can handle, the narrow rimmed glass is able to pick up the slack and keep measuring the amount of water poured.
Fuji Super CCD
So far, Fuji is the only camera company to come up with a new approach to this problem. The Fuji-designed Super CCD comprises of two sets of photosites laid out in an octagonal pattern, where you have a standard size photosite that can record the same dynamic range as a normal photosite and a smaller photosite next to it that can record any detail that is brighter than what the standard-sized photosite is able to record. The data captured using this type of sensor can be extrapolated to create a raw capture image which represents a dynamic range that is wider than most other digital cameras. Fuji classify the two photosites as a single photosite sensor,