Video Frame Size Calculator

An uncompressed frame costs exactly what its samples cost: pixel count multiplied by the average number of component samples per pixel multiplied by the bit depth, divided by eight for bytes. Chroma subsampling sets that middle term. Full 4:4:4 carries three samples per pixel; 4:2:2 halves the chroma horizontally and averages two; 4:2:0 halves it both horizontally and vertically and averages 1.5. An alpha channel adds one more full-resolution sample. Bit depth sets the tonal resolution - 8 bits gives 256 levels per channel, 10 bits gives 1024, 12 bits 4096 - and moving from 8 to 10 bits inflates the frame by 25 percent. A 1920x1080 frame at 4:2:0 8-bit is therefore 3,110,400 bytes, about 2.97 MB, while the same frame at 4:4:4 12-bit is 8.90 MB. Multiplying by frame rate turns frame size into a data rate, and that is what determines whether an interface or drive can carry the signal. HD-SDI runs at 1.485 Gbps, 3G-SDI at 2.97 Gbps, 12G-SDI at about 12 Gbps for single-link 4K, HDMI 2.1 up to 48 Gbps. Uncompressed 4K at 4:2:2 10-bit and 24 fps is roughly 3.2 Gbps, or about 400 MB/s sustained, which is why NVMe drives and RAID arrays appear in capture workflows and why nearly everything is compressed somewhere in the chain. Two caveats on the arithmetic: subsampling factors like 1.5 are averages across a pixel block rather than a per-pixel truth, and real container formats add padding and packing overhead, so v210 10-bit 4:2:2 stores each group of six pixels in sixteen bytes and lands a few percent above the theoretical figure.

Chroma subsampling predates digital video. Analog color television, beginning with the NTSC system approved in 1953 and the PAL and SECAM systems of the 1960s, already transmitted a full-bandwidth luminance signal with narrower-bandwidth color-difference signals, exploiting the eye's lower acuity for color detail and preserving compatibility with monochrome receivers. The 4:1:1 style notation came later, describing sampling frequencies as multiples of 3.375 MHz, which is why the digital component standard CCIR Recommendation 601, published in 1982 by the CCIR and now maintained as ITU-R BT.601, sampled luminance at 13.5 MHz and each chroma channel at 6.75 MHz - the 4:2:2 scheme still used in professional acquisition. Higher resolutions followed the same pattern of standardization. ITU-R BT.709 defined the 1920x1080 HD parameters in 1990, and BT.2020 specified the 3840x2160 and 7680x4320 UHD grids with wide gamut and 10 or 12-bit depth in 2012. Serial Digital Interface, standardized by SMPTE as ST 259 and later ST 292 and ST 2082, fixed the data rates that frame size calculations must respect. On the file side, DPX grew out of Kodak's Cineon format from the early 1990s for 10-bit log film scans, and Industrial Light and Magic developed OpenEXR internally in 1999 and released it as open source in 2003, providing the 16-bit half-float storage now standard in visual effects compositing.