Color depth and chroma subsampling are probably two of the most misunderstood aspects of digital video. Trying to educate someone about them usually devolves into a simple “more is better” conclusion, without actually explaining what’s really going on. Unless you’re Griffin Hammond, documentary filmmaker and Panasonic’s brand ambassador for the upcoming Lumix GH5 mirrorless camera, who explains the concepts effortlessly in a way that anyone can understand.
Color in consumer digital video is typically recorded at a bit depth of eight bits, with chroma subsampling measured at what’s called 4:2:0. This is already more information than the average consumer needs to worry about, but if you’re considering upping your video skills, understanding what this means is important.
Bit depth refers to the overall number of levels of red, green, or blue that the camera records. Eight bits means there are 256 levels (zero to 255) of each color, or roughly 16.8 million combinations in total. Now, 16.8 million sounds like a lot, but one common situation where you can notice the drawback of 8-bit color is in areas of light falloff from bright to dark, which may show up as disparate bands of shading rather than a smooth gradient.
Bumping up to 10 bits multiplies the levels of color by four. That’s 1,024 available values for green, red, and blue, or a whopping one billion total combinations. As Hammond explains, however, you may not notice the difference: Your monitor likely only supports eight-bit color, as does your TV, as does YouTube. (Higher-end, professional monitors can show more colors).
Chroma subsampling is a separate beast altogether. This is often called the color resolution, as compared to the spatial resolution, of an image. As an example, 4K Ultra HD video has a spatial resolution 3840 x 2160 pixels — but the color of each pixel is derived from a much smaller sampling than that.
With 4:2:0, for every two rows of four pixels, color is sampled from two pixels in the top row and zero pixels in the bottom row. Surprisingly, this seemingly dramatic approximation has little effect on how our eyes perceive color. If your camera supports 4:2:2 subsampling, this doubles the color resolution by including color from an additional two pixels on the second row.
So if moving to 10-bit 4:2:2 has little effect on what we can actually see right out of the camera, why is it important? It all comes down to postproduction.
Ten-bit color is especially important to colorists, as it offers more room to push the color and exposure of video. Even if your final output is still an eight-bit monitor, working in a 10-bit space will give you more control and yield a better result.
Also, 4:2:2 chroma subsampling is helpful in the coloring process, but is particularly useful for chroma key, or “green screen,” compositing. Here, the extra color resolution can be the difference between a smooth mask or a jagged outline.
Many video professionals working with mirrorless or DSLR cameras will use external recorders in order to capture more color information than what the camera can process internally. This is usually in the form of either 8- or 10-bit color depth with 4:2:2 chroma subsampling.
The Lumix GH5 is unique in that it will offer internal recording at 10-bit 4:2:2, which could save pro videographers time and money by not requiring an external recorder.
If this explanation still left you scratching your head, rest assured that these concepts are explained much more succinctly in Hammond’s video. So if you haven’t yet, go ahead and give it a watch.