Remote and cloud production
Cloud production means the control room runs as software instead of hardware. The cameras stay at the venue. Switching, graphics, data, playout and delivery happen in the cloud, operated from a browser by people who never travelled. LIGR is built this way from ingest to output.
Cloud Control Room — live match in a browser tab
1920×1080 MP4 loop, ~10s. Full Control Room UI mid-match: feed preview, graphics rail, score correction.
Cloud production is live video production where the production tools run as software in remote data centres instead of as hardware in a room near the pitch. The camera feeds travel to the software. The software never travels to the camera. Everything else is a question of how far you take it.
Remote production, or remote video production, is the broader term, and the two get used interchangeably. They are not identical. Remote video production means the people are not at the venue. Cloud production means the equipment is not at the venue either. Plenty of broadcasters run remote production into a fixed facility packed with traditional hardware, and that is a legitimate model. Cloud production removes the facility too.
The industry term you will meet most often is REMI.
REMI stands for Remote Integration Model. Some broadcasters call it at-home production. In a REMI workflow the cameras, microphones and an encoder stay at the venue, the signals are backhauled over IP to a gallery somewhere else, and the director, vision mixer, replay operator and graphics operator work the match without leaving that gallery. REMI production spread through tier-two and tier-three sport because it removes travel. REMI sports production is now the default for most competitions below the top tier. Cloud-native production goes one step further and removes the gallery.
Model 1
Model 2
Model 3
Most operations sit between models two and three, and that is fine. Cloud-based live video production has never meant nothing is at the venue. Something always is. It means the only things left at the venue are the things physics insists on. Our cloud production versus production truck comparison puts the two models side by side on cost, crew, latency and coverage.
This is the sentence the whole category rests on. A camera has to be where the sport is. A control room does not. Once you accept that, the only real question is which items in a traditional outside broadcast are attached to the pitch by physics and which are attached to it only by habit.
The cameras
Manned, robotic or AI. A camera has to be where the sport is. Nothing about cloud production changes that.
Audio capture
Effects mics on the pitch, and commentary if your commentators are courtside rather than at home.
An encoder
A hardware appliance, a laptop running OBS or vMix, or an encoder built into an AI camera. Something has to turn light into packets.
Connectivity and power
A connection with reliable sustained upload, not headline download speed. This is the constraint that decides whether a venue is viable.
A pair of hands
Even a fully automated match usually needs someone to mount the camera, switch it on and confirm the feed is live before kick-off.
The truck
With it goes the hire cost, the depreciation, the maintenance, the fuel and the parking negotiation with the ground.
The gallery
Vision mixer, replay servers, multiviewer wall, audio desk and comms. All of it becomes software someone opens in a browser.
The graphics operator
In a truck workflow, a person keys every lower third, every scoreboard update and every sponsor bug by hand. In cloud production the graphics follow the data.
The hardware refresh cycle
No three to five year replacement plan for kit that spends most of its life in a flight case.
Travel
Flights, hotels, per diems, driving time and the crew hours you pay for before anyone touches a camera.
Look closely at what happened to the graphics operator, because it is the part most remote production solutions leave half-finished. Moving the gallery to the cloud without changing how graphics are made just relocates the person keying lower thirds. They now sit in a different building doing the same job at the same cost. The saving is travel, and that is all.
Data-driven graphics remove the job instead of relocating it. If the score changes, the scoreboard updates because the data changed, not because someone noticed. That is the difference between broadcast graphics software that happens to run in a browser and graphics that are genuinely bound to the match. We wrote up the trade-offs in cloud versus hardware broadcast graphics.
Cloud playout is the transmission end of the chain running as software. The playlist, the live event windows, the channel branding, the interstitials, the overlays and the failover are all scheduled and executed in the cloud instead of by video servers and a master control room. For sport it is the thing that turns a set of individual match streams into a channel: a league OTT destination or a continuous channel that keeps running between fixtures.
It matters because most rights holders do not have a playout problem for one match. They have one for four hundred. Scheduling and branding at that volume is software work, and cloud playout is where the per-match marginal cost of a channel finally drops to something a second-tier competition can afford.
Automation replaces the operator wherever the decision is deterministic. If the ball crosses the line and the score changes, the scoreboard should update, and no judgement is required. If a fixture starts at 15:00, the pre-match sequence rolls at 14:55 because a rule says so. If a goal is timestamped at 63:12, the clip boundaries around it are arithmetic.
AI cameras extend the same logic to framing. Veo, Pixellot, Spiideo and Hudl Focus capture a wide field of view and pan a virtual camera to follow play, which removes the camera operator from a large slice of grassroots and second-tier sport entirely.
Soccer scorebug — automated score and clock
1920×540 MP4 loop, ~6s, alpha or chequerboard. Goal event firing the score update.
Basketball scorebug — quarter, shot clock, team fouls
1920×540 MP4 loop, ~6s, alpha or chequerboard. Shot-clock reset and a scoring run.
Rugby league scorebug — sponsor rotation in the bug
1920×540 MP4 loop, ~6s, alpha or chequerboard. Two sponsor slots cycling mid-half.
Editorial judgement. The reaction shot of a manager who has just lost a two-goal lead. Knowing that the story of the match is the substitute who came on at 70. Handling something nobody scripted, a crowd incident or a serious injury where the correct decision is to cut away immediately.
Automated framing also degrades in poor light, heavy weather and grounds where the camera cannot be mounted high or central. And automation is only ever as good as the data behind it. A scoring app nobody is operating produces graphics that are wrong, confidently and on air. Automation reduces headcount. It does not remove accountability for the feed.
We set out how far the zero-operator model actually stretches in zero-operator sports broadcasting, and the rule engine itself lives in automation.
Cloud production does not delete cost. It moves it, and it changes its shape. That is worth understanding before anyone builds a business case on it, because the savings are real but they do not arrive where people expect.
Truck and hardware capital
Ownership or hire, plus depreciation, maintenance and the refresh cycle on mixers, replay servers and graphics machines.
Travel and crew time
Not just fares and hotels. The billable hours spent getting to a ground and rigging once you are there.
Fixed crew per fixture
You no longer send the same minimum team to a reserve fixture that you send to a final.
Connectivity at every venue
A convenience becomes a hard requirement. Fixed line where you can get it, bonded cellular where you cannot, and a real budget line either way.
An encoder per venue
One-off, and small next to a truck day rate, but it multiplies by the number of grounds you cover rather than the number of matches.
Platform and delivery
A recurring subscription plus the cost of getting the finished streams to the destinations that carry them.
Central technical cover
Someone who can see every venue's feed health during a match window and act when a link degrades.
The important change is not the total. It is the curve. Truck-based production carries a high fixed cost per match that barely varies with how important the match is. Cloud production carries a low fixed cost per match and a modest recurring cost per venue. The more fixtures you cover, the harder that maths favours cloud, which is exactly why remote production reached lower leagues, women's competitions, youth and community sport first. Those fixtures were never going to pay for a truck at any price.
The variable that actually bites is connectivity. Bandwidth stops being an afterthought and becomes the thing that decides whether a ground is coverable at all. Budget for it per venue, test it before the season rather than on the day, and treat a venue survey as part of the production plan. The full truck comparison walks through where each model wins.
A page about cloud production with no limitations section is a brochure. Here is where the model genuinely loses, so you can decide whether any of it applies to you.
A local hardware chain, camera to mixer to screen, holds end-to-end latency comfortably under half a second and often within a handful of frames. A cloud path cannot match that. You add encode, a transport buffer sized for retransmission, cloud decode, processing, re-encode and delivery. A well tuned contribution leg over SRT can sit in the high hundreds of milliseconds. The full glass-to-glass figure on a delivered stream is normally measured in seconds, and OTT packaging adds more. If your output feeds in-stadium video boards or anything a crowd will compare against what they can see with their own eyes, that gap is real.
Cloud production turns a ground's internet connection into the single point of failure the truck used to carry with it. Bonded cellular, dual carriers, SRT retransmission and a local recording as insurance all help. None of them turn a rural pitch on one contended line into a good venue. Survey connectivity before you commit a competition to remote production.
Cloud replay works. It does not feel the same. An operator scrubbing a jog wheel expects the picture to respond instantly, and any round trip is felt. For high-value officiating review, you still want the frames on local storage next to the person making the call.
Cloud production wins decisively at one to four feeds and high fixture volume. At twelve or sixteen cameras with a talkback-heavy crew and a live director cutting on instinct, the operational overhead climbs and the argument for a truck returns. Most competitions are not that. Some matches in them are.
None of that is an argument against cloud production. It is an argument for matching the model to the fixture. Most rights holders have a small number of matches that deserve a truck and a very large number that currently get nothing at all. Cloud production is how the second group gets covered.
LIGR is remote production software built for sport specifically, rather than a general video tool pointed at a pitch. The chain runs end to end without a gallery.
01
SRT, RTMP and NDI in. That covers hardware encoders, OBS, vMix and the AI cameras clubs already own: Veo, Pixellot, Spiideo and Hudl Focus. You do not replace the venue kit to move production to the cloud.
How ingest works02
Fuse binds Rive designs to live match data with no code. Scoreboards, lower thirds, team sheets, statistics and sponsor overlays read the match rather than waiting for a keystroke. Change a design once and the next fixture uses it.
Fuse03
The LiveScore app gives a scorer real match-event buttons. Every press is timestamped. That single fact is what makes automated statistics and automated highlights possible later without anyone logging a match by hand.
Scoring and data04
Automation rules trigger graphics from match events or from the clock. Pre-match sequence at kick-off minus five. Scoreboard on every goal. Sponsor bug on a rotation. This is where the operator stops being required.
Automation05
Control Room is the manual layer over the automation. Open it on a laptop anywhere, watch the feed, fire a graphic, correct a score, take over a sequence. No gallery, no install, no dongle.
Control Room06
Multi-stream orchestration sends the same match to several destinations at once with different graphics, languages or sponsors on each. Broadcast, OTT and social do not need three productions.
Multi-stream orchestration07
Because every event carries a timestamp, clip boundaries are arithmetic. Highlights are ready close to full time rather than the following morning, which is when they are still worth posting.
Automated highlightsGame Plans — automation firing graphics off match events
1920×1080 MP4 loop, ~8s. Split screen: rule editor on the left, the resulting overlay on the programme feed on the right.
The delivery side is documented in output and streaming, and the reasoning behind running several graphicised feeds from one match is in multi-feed orchestration.
Remote production is easy to demonstrate once. Live sports streaming production is judged over a full season, across venues you do not control, in weather you did not choose, every weekend.
More than 640 matches a season, produced remotely across dozens of grounds with no truck at any of them.
Australian rules at national and state level, with automated data-driven graphics running on live feeds.
Rugby league across a state-wide competition footprint, produced from the cloud rather than from a fleet.
Professional tennis, where fixture volume is high, venues are spread across continents and sending a crew to each one is not a plan.
100K+
Hours Broadcast
20+
Sports Supported
50+
Countries
$20M+
Sponsorship Revenue Driven
Football Australia runs LIGR across 4,000+ live matches annually, spanning nine State Member Federations. Alongside that sit Basketball NSW, Cricket Tasmania, the Corsican football league, World Sevens Football in Miami, the Horizon League and the EuroMillions Basketball League in Belgium. If you are a rights holder, start at broadcasters. If you produce for other people, start at production companies.
Cloud production is live video production where the switching, graphics, replay, playout and delivery run as software in remote data centres rather than as hardware at the venue. Cameras, audio and an encoder stay at the ground. Everything downstream is operated from a browser by people who never travelled. It is also called cloud video production or live production in the cloud.
REMI stands for Remote Integration Model, sometimes called at-home production. Cameras, microphones and an encoder stay at the venue, the feeds are backhauled over IP, and the director, vision mixer, replay and graphics operators work the match from a gallery somewhere else. REMI production removes travel rather than hardware. A fully cloud-native workflow goes further and removes the physical gallery as well.
Cloud playout is channel transmission running as software. The playlist, the live event windows, channel branding, interstitials, overlays and failover are scheduled and executed in cloud infrastructure instead of by video servers in a master control room. For sport it is what turns a season of individual match streams into a continuous OTT or FAST channel that keeps running between fixtures.
There is no single number, only a shape. Your requirement is the encoder bitrate of one contribution feed, multiplied by the number of feeds leaving the venue, plus real headroom for retransmission rather than a tight fit. It scales with resolution, frame rate and codec. What matters is sustained upload under contention, not advertised download speed, so test the ground's actual uplink at the time of day you will be producing.
The camera, because it has to be where the sport is. Audio capture. An encoder, which can be a hardware appliance, a laptop running OBS or vMix, or built into an AI camera. Power, and a connection with reliable sustained upload. Usually one person to rig the camera and confirm the feed is live before kick-off. Nothing else needs to be there.
Usually at volume, but the saving is structural rather than a discount. Truck hire, hardware refresh, travel, accommodation and a fixed crew per fixture come off the budget. Connectivity at every venue, an encoder per venue and a platform subscription go on it. Truck production carries a high fixed cost per match, cloud production a low one, so the advantage grows with the number of fixtures.
More than local hardware, and you should plan for it. An on-site chain of camera, mixer and screen holds end-to-end latency well under half a second. A cloud path adds encode, a transport buffer sized for retransmission, decode, processing and re-encode. A tuned SRT contribution leg can sit in the high hundreds of milliseconds, while a delivered OTT stream is normally measured in seconds.
It replaces the deterministic parts. Scoreboards, clock-driven sequences, sponsor rotations, statistics and highlight clipping are rules and timestamps, and they need no operator. Editorial judgement is different. Reaction shots, knowing the story of a match and cutting away from an incident nobody scripted still need a person. Most competitions automate the routine fixtures and staff the handful that deserve a director.
Multi-stream orchestration — one match, four branded outputs
1920×1080 MP4 loop, ~8s. Quad-split of the same fixture with different sponsors, languages and scorebugs.
Tell us how many fixtures you have, what is already at your grounds and what your connectivity looks like. We will tell you honestly which of them suit cloud production and which do not.