Bit-Depth for Black and White


Today’s Question: If a black and white JPEG image has only 256 different tonal values, which risks banding, is there any recommendation when entering exhibitions of monochrome images when the exhibition requires JPEG images only? Is it OK if the image is prepared as a TIFF and exported as a JPEG only as a last step?

Tim’s Quick Answer: As long as your adjustments were all applied to the original 16-bit per channel version of the image, and that result demonstrates smooth gradations without extreme transitions of contrast, saving the final result as a JPEG won’t cause a serious degradation in image quality.

More Detail: The primary reason it is important to work in the 16-bit per channel mode is to minimize the loss of information in the photo when applying adjustments. The various adjustments you apply to a photo can cause a loss of overall information. For example, increasing contrast can cause a reduction in the total number of tonal values represented in the image.

If you start with 16-bit per channel data, even very strong adjustments won’t likely result in posterization (the loss of smooth gradations of tone and color in an image). If, on the other hand, you start with only 8-bit per channel data, there is a very real risk of posterization, especially if you apply strong adjustments to an image.

This issue is magnified for black and white images, since there is so much less information in a black and white image compared to a color photo. With a color image you start with a theoretical maximum of more than 281 trillion possible color values available. Converting to 8-bit per channel reduces that number to a little more than 16.7 million colors.

For a black and white image you start with a maximum of 65,536 possible shades of gray when working in the 16-bit per channel mode, but only 256 shades of gray when you have converted to the 8-bit per channel mode.

As long as your adjustments are applied while working in the 16-bit per channel mode, you will minimize the impact of saving a copy of the final result as a JPEG image, which in turn means that you’ll be creating an 8-bit per channel version of the photo. Of course, it is also important to keep in mind that if your adjustments produce extreme transitions of tonal values, there may still be some posterization evident in a copy saved in the 8-bit per channel mode, even if you worked on the original in the 16-bit per channel mode.

However, if the image looks very nice in 16-bit per channel mode and you don’t apply any adjustments to the 8-bit per channel copy you create to submit to the competition, you can expect that 8-bit version to be of very good quality as well, with little or no posterization evident.

To be sure, there is still some degree of risk of visible posterization with any monochromatic (black and white) 8-bit per channel image. In the context of a photo contest I suppose you can be reassured by the fact that all images that are submitted will face the same limitations. But posterization should be minimal if you have performed all of your adjustments in the 16-bit per channel mode before creating the 8-bit JPEG for purposes of submitting a photo to the contest.