Color Bit Depth

Traditionally, each channel in a color image is an 8-bit grayscale image. Therefore, an RGB image has a bit depth of 24, or 8 times 3 (see Table 1.1). If you do the math by raising 2 to the 24th power, you will get more than 16 million possibilities of color for each pixel in the image.

That sounds like a lot, and it is a lot because our human ability to perceive differences in tonality fails in the range of a few million possibilities per pixel. In other words, 24-bit images have more gradations of color than our eyes can perceive.

For a long time in the history of digital imaging, 8-bits/channel has been sufficient for all but the most discerning professionals. However, when you manipulate images you lose some of the data in the process (see Chapter 3, "Retouching Photos," for more on this subject). The differences in tonality post-manipulation often fall below our perceptual threshold—in other words, we can see problems with heavily manipulated images.

The solution some photographers are adopting is to shoot in 16 bits per channel. This is currently possible only on prosumer or high-end professional digital cameras. Photos with 16 bits per channel have 65,536 possibilities per pixel per channel (refer to Table 1.1), and in RGB color mode this equates to 48-bit images (16x3). The benefit to shooting in 16 bit is that you can heavily retouch photos without being able to perceive banding (imperfections) in the result.

WARNING The downside to shooting in 16 bit/channel is the huge volume of data that your system then has to handle. Much larger images require expensive digital cameras with huge storage space available. In addition, your computer must have more than a gigabyte of memory, huge hard drives, and fast processors to reasonably manipulate the data sets.

You can check an image's bit depth by choosing Image > Mode and seeing whether 8 Bits/Channel or 16 Bits/Channel is selected.

To arrive at the total bit depth stored in an image, see how many channels appear in the Channels palette and multiply this number by your bits per channel. For example, if your image has 4 channels (CMYK) and it stores 16 bits/channel, multiply 4 by 16 to arrive at a composite 64-bit image.

NOTE Photoshop CS now offers greater ability to work with images containing 16 bits per channel.

0 0

Post a comment