Camera positions for a full court, the encoding pathway that survives a venue's wifi, a federation graphics package, scoreboard data and simultaneous multi-platform distribution. The complete operating picture for a season — not a single game.
Four cameras is the sweet spot where a basketball stream stops looking like a webcam and starts looking like a broadcast — and you can run a whole season of it with a crew of two to three and no OB truck.
Most federations do not need a bigger production. They need a consistent one — the same watchable, professional broadcast every fixture, for a budget a domestic league can actually carry. This is how that is built, position by position.
A single locked-off camera gets the game online. It does not make it watchable. The eye tires of one angle inside a quarter, replays are impossible, and the result reads as a recording of a match rather than a broadcast of one.
Two cameras — a follow and a wide — is the realistic floor. But four is the point where the stream crosses into broadcast territory: you can cut to the action under either basket, hold a clean tactical wide whenever you need it, and serve a replay that actually shows what happened. Beyond four, for a domestic federation season, you start adding cost and crew faster than you add value the audience will notice.
The whole production runs from a single broadcast OS on a fold-out table — switching, graphics, replay and encoding in one place. Here is the build, in the order you set it up on game day.
One elevated centre-court camera for the live follow, two opposite-baseline cameras for the under-basket action, and one fixed tactical wide that holds the entire court for replays and graphics-safe cutaways.
Encode the program feed through a bonded or dual-path uplink so a single venue wifi drop never takes the stream off air. Test the real, sustained bitrate at the venue — not the number on the router.
A consistent scoreboard, team lower-thirds and a live clock — built once and reused identically every fixture, so every game looks like the same broadcast.
Feed the venue's scoring and clock data into the graphics so the on-screen score and time update automatically, rather than being typed by hand all night.
Send a single clean program feed to a CDN that fans it out to the federation platform and the clubs' channels at once — then record the full match for the archive.
Position is where most self-produced streams quietly go wrong. The cameras are fine; the angles are not. For basketball, four placements do almost all the work:
Here is the uncomfortable truth of live sport: the thing that takes you off air is almost never a camera. It is the internet. A sports hall's wifi is shared with several hundred phones, and the upload you tested at 9am is not the upload you have at tip-off.
So the encoding path matters more than the resolution. Build redundancy in: bond the venue line with a cellular connection, or carry a second independent uplink, and let the encoder fail over without dropping the broadcast. Encode at a bitrate you can actually sustain, not the highest number the venue will momentarily allow.
Always measure the sustained upload at the camera position during a busy moment — a half-full hall behaves nothing like an empty one. Plan your bitrate around that figure with headroom to spare.
A clean feed with no graphics reads as amateur however good the cameras are. A federation graphics package — a proper scoreboard, team and player lower-thirds, a live game clock — is the single cheapest upgrade from "someone filmed it" to "this is a broadcast."
The part that earns its keep across a season is live data: wiring the venue's scoring and clock system into the graphics so the on-screen score and time move on their own. Typing the score by hand is survivable for one game; over a season it is exactly where errors and lag creep in.
Broadcast & Live Streaming — federation-grade multi-camera production on a lean crew.
A single great game is a production. A great season is an operation. The difference is repeatability: the same plot, the same graphics, the same encoding profile and the same hand-off, run identically every fixture, so quality never depends on who happens to be in the building that night.
Distribute one clean feed to a CDN and let it fan out to the federation platform and the clubs' own channels simultaneously — not platform by platform. Record every match in full and hand the archive and pull-ready clips back after the final whistle. Do that for a season and the audience learns the games are always there, looking the same every week. That reliability, more than any single highlight, is what builds a federation's broadcast.
The audience never notices a great season. They only notice an inconsistent one. Consistency is the product.
Two is the floor for a watchable stream, but four is where it reads as a broadcast: a centre-court follow, two baselines for the under-basket action, and a tactical wide for replays. Beyond four, you add cost and crew faster than value for a domestic federation season.
Yes. A modern broadcast OS runs all four cameras, graphics, replay and encoding from one touchscreen on a fold-out table — no outside-broadcast vehicle, no gallery crew. A team of two to three operates the whole production.
Enough sustained upload for your bitrate plus headroom — but above all, a redundant path. Bond two connections (venue line plus cellular) or carry a second uplink, because the failure that takes a stream off air is almost always a single internet drop, not the cameras.
By feeding the venue's scoring and clock data into the graphics package so the scoreboard and game clock update in real time. Typing the score by hand works for one game; across a season it is where mistakes and delays creep in.
Two to three people per fixture: one on the broadcast OS (switching, graphics, replay) and one or two on the manned cameras. The baselines and tactical wide can run locked-off, which is what keeps a full season affordable.
The guide is how we think. The work is what we deliver. Tell us about your season and we will tell you exactly how we would produce it — cameras, crew and cost.