I’ve often marvelled at the computer generated lines they draw across the fields in the football game broadcasts. Though not really expecting an answer, I’ve asked aloud during several broadcasts, “How do they do that? How do they draw those lines on the field.”
Inevitably, the answer comes back, simply, “With computers.”
“I know that, but how do they do it. Look! The lines don’t even cross over the players, it actually appears underneath them as if it were actually on the field.”
“I don’t know.”
That’s been the typical conversation all these years, and I’ve come no closer to getting an answer. (Strangely, I’ve never Googled to find an answer.) Until today.
I learned that the first time this was done was during a game between the Baltimore Ravens and the Cincinatti Bengals back in Sept. 27, 1998. (Bengals lost.) That is a long time to be wondering about this without ever investigating. I should probably be embarrassed.
Turns out Matt Lake of the New York Times published an article about this 10 years ago. The article is titled, “When the Game’s on the Line, the Line’s on the Screen”. In summary, this is how it’s done according to Matt Lake’s report:
- Surveying equipment is used to help create a 3-dimensional computer model of the football field. Each field being broadcast from needs it own model. A standard, generic model isn’t used.
- The video cameras used to broadcast the game have sensors that detect the cameras position and angle of view of the field including their zoom and focus settings.
- The raw video feed is received by a one of several computer graphics companies that specialize in handling the computer work for the broadcasts. They receive the feed just a few seconds before they broadcast goes over the air.
- The camera view information is matched against the computer generated model of the stadium. Then the field marker information is entered into the computer.
- The computer image is then overlaid over the video feed, but the computer work doesn’t end there. The grass color and lighting, adjusted for weather, needs to be matched so the computer image looks natural, but the computer needs to be informed of the teams colors so the image doesn’t cover the players. (That’s the part I always wondered about.)
So that, in a nutshell, is how it’s done. Amazingly, all this is done in about 3 seconds. The same technology is used to enter those little ads, which seem to be getting bigger, in the corners of the TV screens.
You can read Matt Lake’s New York Times article for the full story. There’s another article at HowStuffWorks.com, too.