Upgrading to 16-bit


Today’s Question: As a follow-up to your question about converting from JPEG to raw [in Friday’s Ask Tim Grey eNewsletter], wouldn’t applying that conversion help ensure smoother gradations of tone and color in the photo? I know you’ve mentioned that in the past as a benefit of a higher bit depth.

Tim’s Quick Answer: No. Converting an image from 8 bits per channel to 16 bits per channel will not have any significant impact on the degree of posterization that may occur if you apply strong adjustments to the image.

More Detail: One of the key advantages of working with high-bit data for a photo is the ability to ensure smooth gradations of tone and color even with strong adjustments to an image.

An 8-bit per channel image can have up to 256 shades per channel. A 16-bit per channel image can have 65,536 shades per channel. That’s a significant difference, especially when you consider a typical image is comprised of three channels (red, green, and blue).

As you apply adjustments to an image, such as enhancing contrast, the transition of tonal values among neighboring pixels will change. With strong adjustments, smooth gradations can be made less smooth, creating effects such as banding in what would otherwise be a smooth transition in the sky, for example.

Converting an 8-bit per channel image to the 16-bit per channel mode still means the image contains a maximum of 256 shades per channel. As you apply adjustments to the image, that number could increase as pixel values are possibly distributed differently in terms of the actual value. But that doesn’t provide any real benefit, especially when you consider a 16-bit per channel image would actually start off with so much more information compared to an 8-bit per channel image.

It is worth noting that most digital cameras don’t actually support 16-bit per channel capture. Rather, most camera models support either 12-bit or 14-bit capture. That still translates to 4,096 shades per channel for a 12-bit capture, or 16,384 shades per channel for a 14-bit capture, which again provides a significant potential advantage compared to an 8-bit capture.

The issue of bit depth is one of the reasons I recommend using raw capture rather than JPEG capture whenever possible. And, as noted above, converting a 8-bit capture to the 16-bit per channel mode won’t provide any real benefit compared to capturing at a high bit-depth in the first place.