Adobe Photoshop CS5 for Photographers - Martin Evening [140]
JPEG images are always limited to 8-bits, but TIFF and PSD files can be in 8-bits or 16-bits per channel. Note though that Photoshop only offers 8-bits or 16-bits per channel modes for standard integer channel images, while 32-bit support in Photoshop uses floating point math to calculate the levels values. Therefore, any source image with more than 8-bits per channel has to be processed as a 16-bits per channel mode image. Since most scanners are capable of capturing at least 12-bits per channel data, this means that scanned images should ideally be saved as 16-bits per channel images in order to preserve all of the 12-bits per channel data.
Figure 5.11.
The bit depth of an image is a mathematical description of the maximum levels of tone that are possible, expressed as a power of 2. A bitmap image contains 2 to the power of 1 (two levels of tone), in other words, black or white tones only. A normal Photoshop 8-bit grayscale image or an individual color channel in a composite color image contains 2 to the power of 8 (2 8) and up to 256 levels of tonal information. When three RGB 8-bit color channels are combined together to form a composite color image, the result is a 24-bit color image that can contain up to 16.7 million shades of color (2 8 × 3).
Bit depth status
You can check the bit depth of an image quite easily by looking at the document window title bar, where it will indicate the bit depth as being 8-, 16- or 32-bit.
In the case of raw files, a raw image contains all the original levels capture image data, which will usually have been captured at a bit depth of 12-bits, or even 14-bits per channel. Camera Raw image adjustments are mostly calculated using 16-bits per channel, so once again, all the levels information that is in the original can only be preserved when you save a Camera Raw processed raw image using 16-bits per channel mode.
Why is 16-bits really 15-bits?
If you have a keen knowledge of math, you will notice that Photoshop's 16-bits per channel mode is actually 15-bit as it uses only 32,768 levels out of a possible 65,536 levels when describing a 16-bit mode image. This is because having a tonal range that goes from 0 to 32,767 is more than adequate to describe the data coming off any digital device. Also, because from an engineering point of view, 15-bit math calculations give you an exact midpoint value, which can be important for precise layer blending operations.
8-bit versus 16-bit image editing
A higher bit depth doesn't add more pixels to an image. Instead, it offers a greater level of precision to the way tone information is recorded by the camera or scanner sensor. One way to think about bit depth is to consider the difference between having the ability to make measurements with a ruler that is accurate to the nearest millimeter, compared with one that can only measure to the nearest centimeter.
There are those who have argued that 16-bit editing is a futile exercise because no one can tell the difference between an image that has been edited in 16-bit and one that has been edited in 8-bit. Personally I believe this to be a foolish argument. If a scanner or camera is capable of capturing more than 8-bits per channel, then why not make full use of the extra tonal information? In the case of film scans, you might as well save the freshly scanned images using the 16-bits per channel mode and apply the initial Photoshop edits using Levels or Curves in 16-bits mode. If you preserve all the levels in the original through these early stages of the edit process, you'll have more headroom to work with and avoid dropping useful image data. It may only take a second or two longer to edit an image in 16-bits per channel compared with when it is in 8-bit, but even if you only carry out the initial edits in 16-bit