A Bit on Depth and Gamut
There are two aspects of digital photography that all serious photogs should know about; bit depth and gamut. Understanding them can help you make better decisions on whether or not to capture your images in RAW, and what happens when you process them.
Our eyes see in two ways; luminosity, and color. They have two types of receptors within the retina (which equates to the sensor on a digital camera); “Rods” handle the luminosity (shades of gray) while “Cones” handle the chroma (hues of color). These names come from the general shape of the photo receptors on a microscopic level.
This is so, because low light levels present problems for vision just as they do in photography. Low light vision is called “scotopic” vision, and it’s the region in which the rods operate. In this way, humans can see general shapes in extremely low light, although with somewhat compromised spatial acuity.
When light levels are more normal, the cones kick in, and we can see color and much greater detail. This is called “photopic” vision. At these levels, both rods and cones can be active, and this is known as “mesopic” vision.
Sadly, digital cameras (and film) only have the equivalent of cones, which might account for why human vision has such a wide dynamic range, that is, the ability to perceive an extremely wide range of tonal information. “Tonal” is a key word here, because it refers to shades of gray, also known as “Luminosity” or simply “L”.
The good news is that while cameras capture light with only one type of receptor, they process this information much like our eyes do—by separating luminosity from chroma (another word for color).
Within the retina, there are three types of cones which correspond to long, medium or short wavelengths—red, green and blue. But these three nodes of information are then expanded to four in two groups; green-magenta, and blue-yellow. This is known as an “opposing color” space. If you look closely at your camera’s white balance settings, you will see this in action. A square with two axes, blue and yellow on the horizontal axis representing “temperature”, and green and magenta on the y axis representing “tint”. Temperature and tint are white balance terms.
Just as this information is sent to the brain, so the color information that you capture in a RAW image is stored “internally” in a color space unfamiliar to most—L*a*b. In this space, “Luminosity” information is stored in the “L” channel separately from the “Chroma” (the “a” and the “b” channels). But far more information is stored in these spaces than the typical device can handle, and so they are represented in another space—RGB. And this is what takes place when you open a RAW image, and then save it as a JPEG or TIFF image.
So where do bit depth and gamut factor in? Luminosity is represented by “bit depth”, and color is represented by gamut. A RAW image has 12 to 14 bits of luminosity information, which equates to a minumim of 4,096 shades of gray. This is sixteen times the amount of information that’s in an 8-bit RGB image. So, when you adjust exposure in lightroom you can choose the range of tones you actually need, and store them in the RGB file upon conversion.
Gamut = Amount of color information or “color profile”
Depth = Amount of tonal information or “bit depth”
But the color is a little different. Color is represented by “Gamut”. The RAW image has an enormous “device independent” color gamut of 68,000 colors, far more than can be represented by any device. So, we have to convert this to a color space that can be represented (16 million colors), and this could be sRGB (smaller gamut, can be represented by any device), Adobe RGB (larger gamut, can be represented by some devices) or ProPhoto RGB or ColorMatch RGB (enormous gamut, cannot be represented by any device).
So, why would we ever need more than 8-bits of luminosity or a greater gamut of colors than the sRGB color space can provide? Simple. When we work in Lightroom or Camera RAW, we make non-destructive adjustments to a huge amount of information which can eventually be stored within a smaller space that can be represented on a different device such as a monitor or color printer. The moment we work outside of the RAW space, we limit ourselves to what we can do, because we’ve already “thrown away” a good portion of information in the process of opening the image in photoshop. With RAW, we can choose which tones and colors make up the final image.
All cameras shoot RAW. And once the RAW capture is saved by your camera in JPEG format, it’s 4,096 shades of gray have been reduced to 256, and its color gamut has been reduced from 68 million to 16 million. Now, if your settings cause the camera to choose the right tones and colors, you’re good. And you can tweak them a little bit to get them better. Your skill in setting the camera’s “processing” determines how successful your JPEG captures can be. And, they can be marvelous.
But, if your camera also allows you to save the RAW image that it captures, you have access to all 4,096 shades of gray and 68 million colors to work with. This means you can make much larger adjustments to get closer to achieving on paper (or on a display monitor) what the human eye can see.
So if you’ve read this far, you’ve earned the prize; here’s the secret of how to work with RAW data outside of Lightroom and Camera RAW.
Set your workflow preferences to open the RAW image into a high-gamut color space (ProPhoto RGB or ColorMatch RGB) with a bit depth of 16 Bits/Channel. This will make all the information in the RAW image available to you in Photoshop. Yes, the file size will be enormous. But, this is only a “working” space. Once you have finished your “retouching” (color and tonal adjustments and pixel editing) you will then convert back to an 8-bit/small gamut color space (sRGB should only ever be used for the final destination space that will be printed or displayed on a standard monitor).
RAW > ProPhoto RGB, 16 Bits/Channel > Retouching > sRGB
Now, if you want to achieve some amazing color and exposure work, after you open in the high-gamut color space, immediately convert to the L*a*b color space. In this space (which does not produce a a larger file size) you can make adjustments to exposure without having any affect on color, and vice-versa. That means you can create wonderfully saturated colors without losing shape (contrast) or add shape and contrast without taking away from colors. You can overcome the curse of the saturation tool in RGB, where in the process of increasing the saturation (purity) of color, it sucks away the contrast.
RAW > ProPhoto RGB, 16 Bits/Channel > L*a*b Color > Retouching > sRGB
Yes, RAW images are larger in file size, but you’ll begin to appreciated just how small they are compared to the amount of information they contain.
You may be thinking, “why can’t I just open my 8-Bit JPEG images into a high-gamut, hi-bit space?” Well, you could, but there’s really no advantage in doing so. It would be like pouring 5 gallons of water in a ten-gallon bucket. You still have to make up the space somehow, and Photoshop does this by spreading out the tones and color information over a wider area, which creates “gaps” in between. And this can lead to some pretty nasty looking images.
Have digital photography related questions? Always open at dreuxsawyer@gmail.com
