Chroma Subsampling

What this is about

Human vision is far sharper for brightness than for colour, and compression exploits exactly that. In 4:2:0 the colour information is halved horizontally and vertically, which is invisible in normal viewing and standard for streaming. Trouble starts where software has to read colour edges: green screen keys tear, fine coloured text smears and heavy secondary grading reveals blocky patches. 4:2:2 halves colour only horizontally and is the practical floor for anything that will be keyed. 4:4:4 keeps full colour and is reserved for compositing and finishing.

A term only helps when everyone on set means the same thing by it — so what follows is how it is used, not a dictionary definition.

What runs differently here

Where this differs from the general case:

• 4:2:0 is fine for viewing, 4:2:2 for grading and keying, 4:4:4 for compositing and finishing
• Green screen work in 4:2:0 produces ragged edges that cost hours to clean up afterwards
• Coloured graphics and fine text suffer most from subsampling, plain camera images least

With TillyGen

When the breakdown marks scenes as green screen work, that flag reaches the camera department early enough to set a suitable recording format.

Change one constraint and the consequences travel through the whole plan: affected shots are flagged, the call sheet is regenerated, and nobody keeps working from yesterday's version.