Advertised Bit Depth

Facebooktwittergoogle_pluslinkedin

Today’s Question: I’m shopping for a flatbed scanner to scan some old family photos, and am confused about the bit depth specification. How does the 48-bit specification for the scanner relate to the 8-bit or 16-bit options available to me in Photoshop?

Tim’s Quick Answer: The specifications for many film and flatbed scanners list the combined bit depth for all three channels, rather than the per-channel bit depth often presented in software such as Photoshop. As such, 48-bit is actually the same as 16-bit per channel for a three-channel (RGB) image.

More Detail: In most software for processing images, the bit depth is presented as a per-channel value. This is in part because many users will need to work with images with a different number of channels, and thus a combined value for bit depth would be less clear. For example, a grayscale image could have a single channel, a typical color photo could contain three channels (for an RGB image), and a CMYK file prepared for printed output would normally have four channels.

So, if we assume a bit depth of 8-bit per channel, the grayscale image would have a total of 8 bits, the RGB image would have 24 bits (8 bits per channel times 3 channels) and the CMYK image would have 32 bits (8 bits per channel times 4 channels). This could obviously be confusing, since all of the images have the same per-channel bit depth.

Within software such as Photoshop, therefore, the bit depth is described based on the number of bits per channel, with typical images being either 8-bit per channel or 16-bit per channel.

Scanner specifications will often describe this bit depth based on the combined total of all channels, and based on an RGB image with three channels. Thus, a scanner that supports 8-bit per channel scanning would be described as a 24-bit scanner (8 bits per channel times 3 channels) and a scanner supporting 16-bit per channel scanning would be described as a 48-bit scanner (16 bits per channel times 3 channels).

Keep in mind that while image files are generally saved as 8-bit per channel or 16-bit per channel images, devices such as digital cameras and scanners might support other bit depths such as 12-bit per channel or 14-bit per channel. A scanner with 12- or 14-bit per channel scanning could then be described as a 36-bit or 42-bit scanner.

But again, the bottom line here is that scanner specifications often report the combined number of bits for all channels (typically based on a 3-channel RGB image), while imaging software generally presents the per-channel bit depth. As a result, a small amount of math may be required to translate bit depth support between scanners and software.