Technical Topics
4K Video Quality
12min
this page describes how color space, chroma subsampling, bit depth, and video compression relate to 4k video output it also outlines best practices for encoding 4k video color space each frame in a digital video, like a digital image, is made up of numerous individual pixels to encode, store, and transmit these pixels, we represent them as numerical values for digital video, there are two popular systems for representing a pixel as a numerical value rgb (red, green, blue) the pixel is made up of three distinct color values, which are then combined on the display to create the visual quality of the pixel the amount of data used to represent each color value is the same ycbcr (luma, blue difference, red difference) the pixel is made up of a luma value, which represents the brightness of the pixel, and two color difference values, which when combined represent the color of the pixel unlike the rgb system, the amount of data used to represent the y (luma) value can be different from the amount used to represent the cb and cr (color) values the ycbcr system takes advantage of a crucial factor in human vision we can discern differences in light (luma) much more finely than differences in color (chroma) this means that ycbcr allows us to reduce the amount of data dedicated to representing color, while keeping the amount of data used to represent light the same, thereby reducing the total amount of data needed to display a video this advantage is key to understanding the purpose of chroma subsampling chroma subsampling ycbcr chroma subsampling ratios such as 4 4 4, 4 2 2, and 4 2 0 represent how much data of a pixel is dedicated to color 4 4 4 – the same amount of data is dedicated to color values as for the light value in fact, this ratio can refer to either an rgb or ycbcr value because, in this case, there is little difference between the two 4 2 2 – half of all color information is being preserved compared to the 4 4 4 ratio as a result, the amount of data required for storing, decoding, and displaying the video is reduced by ⅓ 4 2 0 – one quarter of all color information is being preserved compared to the 4 4 4 ratio as a result, the amount of data required for the video is reduced by ½ bit depth bit depth is the other major factor in determining the size and color quality of a video the bit depth of a video determines how many bits can be used to describe the rgb or ycbcr quality of each pixel for example, an 8 bit video depth means that each y, cb, and cr value for a pixel can have one of 256 values, while a 10 bit video depth expands the possible values to 1024 (the same rules apply for an rgb video) it is important to note that reducing the bit depth and adding chroma subsampling both reduce the data rate of the video, negatively affecting video quality these losses in quality occur on top of any losses that result from the compression of the video codecs themselves encoding video codecs support different combinations of bit depth and chroma subsampling furthermore, hdmi ® standards support their own sets of bit depth and chroma subsampling to ensure optimum video quality, you will need to consider the capabilities of both the video codec and the hdmi output the hd (h 264), 4k (h 265), and jpeg standards define what chroma subsampling and bit depth can be used when encoding media brightsign xtx44, xtx43, xdx34, xdx33, and 4kx42 players support the following values h 264 4 2 0 subsampling with 8 bits of depth (main profile and high profile)–at up to 1920x1080x60p h 265 4 2 0 subsampling with 8 bits (main profile) or 10 bits (main 10 profile) of depth–at up to 3840x2160x60p main 10 is the maximum video quality offered by the h 265 version 1 codec jpeg 4 4 4 subsampling with 8 bits of depth–at up to 1920x1080 hdmi output once the video is decoded by the player, it is then output to the display over hdmi at this point, a second set of subsampling and bit depth values are chosen to best match the capabilities of the screen that is connected to the brightsign player brightsign xtx43, xdx33, and 4kx42 players can output the following values over hdmi resolution 8 bit 10 bit 12 bit 4kp24 4kp25 4kp30 4 4 4 (rgb) 4 4 4(rgb) 4 4 4 (rgb) 4kp50 4kp60 4 4 4 (rgb) 4 2 0 4 2 0 4 2 2 4 2 0 this is the only 4k video profile supported over hdmi 1 4 all other profiles require a display with hdmi 2 0 support encoding best practices the following video encoding guidelines should help you achieve the best 4k video quality allowed by your display note that hdmi 2 0 is backwards compatible with hdmi 1 4, so you can use hdmi 2 0 compatible players with an hdmi 1 4 display hdmi 1 4 frame rate 30p subsampling ratio 4 2 0 color depth 8 bit (main profile at level 5 0) hdmi 2 0, 8 bit color frame rate 60p subsampling ratio 4 2 0 color depth 8 bit (main profile) hdmi 2 0, 10 bit color frame rate 60p subsampling ratio 4 2 0 color depth 10 bit (main 10 profile) hdmi 2 0, 12 bit color frame rate 60p subsampling ratio 4 2 0 color depth 10 bit (main 10 profile)