This is one of those questions that seem to arise when still photographers start working in video. Still imagery has always been referred to as having “True Color” or “24bit” with “millions of colors”. Those were technically accurate terms when they were first used for marketing still digital cameras.
“True Color” vs actual numbers
“True Color” to the still photographer refers to the fact that human vision perceives roughly 10 million possible colors. So the “True Color” colorspace, with its 16,777,216 displayable colors, encompasses all of what is considered viewable in the spectrum of human vision. That is the ability to display at least 265 shades in each of the Red, Blue and Green color channels. Computer graphics cards also support what is referred to as “Deep Color” referring to the 30 / 36 / 48 / 64bit color as defined in the graphics world but as 10, 12, 14 and 16bit color in the film and video space.
In the digital age, “Bit Depth” alluded to the actual number of physical 1 or O bits that were allocated to each pixel to allow it to reproduce a specific color. On a computer screen, an 8bit image could reproduce one of only 256 distinct colors per pixel onscreen at anyone time (8 x 8 x 8 = 256). The thinking person reading this would identify that as a first generation web graphics palette, and called “Indexed Color”.
Moving pictures use a different math (not really)
As covered previously in the Color Conundrum article, video uses a different attitude, allowing for color compression and the ability to separate the Chroma (color info) from the Luma (a monochrome or grayscale image). So why wouldn’t they choose a different methodology for identifying color info?
Bit Depth definitions for video use the same info. SMPTE defines color spaces based on the bpp (bits per pixel) nomenclature. So that True Color image we started with had the ability to show 256 shades or levels of information per color channel, with more than 16.7 million distinct colors. It’s equivalent to the 8bit color that your HD camera records at, since it allows for 256 levels of density in each color channel.
Wait? So True Color is actually only 8-bit?
Technically yes, when using the bpp (bits per pixel) nomenclature used for video signals since it defines the same number of color data points when using 256 levels of gray per channel. The majority of the video content we look at is viewed this way except for those that are working with higher end cameras or those working in post production.
Few realize that 8-bpp (lets just call it 8-bit Ok?) for video is by far the most prevalent, since both OTA (over-the-air) and cable transmissions all rely on 8bit delivery. Until we started working and watching television in HD, no one really noticed the banding or color compression artifacts.
As cameras moved into HD or Digital Cinema and are essentially replacing film in acquisition, it has become necessary to also increase the amount of information recorded. Film’s ability to record as much as 14 stops of latitude was not possible within the confines of an 8bit recording scheme, mainly because there was just not enough information being recorded.
Cameras today record information now at 10 or 12-bits (like Apple’s ProRes) or in Raw Data formats that allow recording up of 14bit and even 16-bit color depth so they match the Deep Color profiles (add bpp data per color channel to get Deep Color numbers of 30 / 36 / 48 / 64bit).