Bit Depth for RAW

Facebooktwittergoogle_pluslinkedin

Today’s Question: As a follow-up to your answer about compression for RAW captures, can you address the impact of the bit depth option for cameras the offer several different bit depth settings for RAW captures?

Tim’s Quick Answer: The bit depth of your photos primarily relates to the total range of tonal and color values available. Therefore, a lower bit depth can reduce color fidelity and dynamic range, and increase the risk of posterization in your photos.

More Detail: When information in an analog form (such as light) is converted to a digital value (such as the numeric color values for a single pixel), bit depth plays a role. Put simply, bit depth determines how many individual values are available. In the context of a digital photo, bit depth determines the total number of tonal and color values available.

For example, an 8-bit per channel color image can make use of almost 16.8 million individual color values. A 16-bit per channel color image can theoretically make use of more than 281 trillion individual color values.

When you have more color values available, the differences between neighboring color values are very small, and so gradations in an image can be incredibly smooth. When you have a relatively small number of color values available, the difference between neighboring color values is relatively large, and so it can be difficult (or impossible) to retain smooth gradations in an image.

Many (perhaps most) digital cameras today use 14-bit per channel analog to digital conversion for RAW captures. Some higher-end cameras use 16-bit per channel, and some older or lower-end cameras employ 12-bit per channel conversion. JPEG captures, by the way, always represent 8-bit per channel conversion.

If your camera is capable of 14-bit conversion, it might also offer a 12-bit option for RAW captures. The result would be a smaller file size, but also a smaller number of possible tonal and color values in the image. Specifically, an image with a 14-bit conversion could employ up to about 4.4 trillion possible color values, while a 12-bit conversion would translate to a maximum of about 68.7 billion possible color values.

Whether or not the “extra” potential color values available at a higher bit depth are going to be of much help depends in large part on the amount of processing you’ll apply to the photo. If you apply many strong adjustments to a photo, you’re going to lose a certain amount of detail, and a higher bit depth will be helpful. If a photo requires only modest adjustments, a lower bit depth will work out perfectly fine.

Naturally it is difficult to quantify the precise potential benefit for a given photo. Some photos will benefit more from a higher bit depth than others, and your workflow plays a significant role here. That said, my personal preference is to opt for the higher bit depth, even though that translates into larger file sizes. I may be getting a benefit I don’t really fully leverage, but I still prefer the peace of mind of having as much information as possible in my image files.