Bonded Cellular vs Satellite for Live Contribution

A story breaks at half past two, four hours away by road, and the bulletin is at six. Nobody is booking space segment for that. A crew loads an encoder and a bag of SIMs into a car and goes.

Three weeks later the same broadcaster covers an offshore sailing event: one operator with patchy coverage along that stretch of coast, six hours of continuous live output, and a finish line that moves with the wind. That production is not going out over bonded cellular.

Both calls were correct. Neither was really a verdict on which technology is better. Live contribution runs on bonded cellular, on satellite, on fibre, and most often on a combination, and the decision is made against one specific location, notice period and duration. This article sets out what each path does, where each one breaks, and how to make the call without relearning it at every event.

What each path actually does

Bonded cellular aggregates several mobile connections inside an encoder and presents them to the production as one logical link. The encoder spreads the video across the available modems, monitors each one continuously and reallocates as conditions change, so the loss of any single connection degrades the picture instead of ending the transmission. The detail of how many SIMs and which carriers sits in the batch 1 article on multiple carrier SIMs in one encoder.

Satellite news gathering sends the signal from a truck or flyaway terminal to a satellite and down to a teleport, on capacity booked in advance. Traditional SNG uses geostationary satellites parked at roughly 35,786 km above the equator, which is what gives the service its defining characteristic: the link works anywhere with a clear view of that orbital slot, and it is completely indifferent to how many people are standing around the antenna.

Low earth orbit services changed the second half of that description. LEO constellations operate a few hundred kilometres up, which cuts the round trip dramatically and shrinks the terminal to something a two-person crew can carry. They introduced their own variables, covered below, and they made the old shorthand of satellite meaning slow inaccurate.

Latency is where the two paths genuinely separate

A geostationary hop covers roughly 72,000 km before the signal reaches the teleport. Physics puts the round trip at around half a second regardless of the equipment involved. For a one-way feed into a gallery that is entirely workable, and broadcasters have built around it for decades. For a two-way live interview it produces the pause every viewer recognises, and for talkback it is close to unusable without careful workflow design.

LEO brings the round trip down into the tens of milliseconds, which puts it in the same conversational range as terrestrial connections. The trade is variability: the terminal hands off between satellites as they pass, and the resulting jitter matters to a production in a way that a stable delay does not.

Bonded cellular is where the intuition most often goes wrong. The underlying mobile network latency is low, but the encoder deliberately adds a delay buffer to absorb jitter across the modems and reassemble the stream in order. That buffer is usually configurable from well under a second to several seconds, and it is a direct trade: a longer buffer produces a more robust picture in poor conditions, a shorter one gets closer to live and gives the encoder less room to recover. A bonded link is low latency when it is configured to be, and configuring it that way costs resilience.

The practical consequence is that latency should be specified per service rather than per event. The programme feed can usually absorb a delay. Talkback cannot, and on many productions it is carried separately for exactly that reason.

The uplink question, which is the whole question

Contribution is an upload problem. That single fact explains most of the difference between the two paths in day to day operation.

Satellite capacity is allocated. A booked space segment delivers what it says, and it delivers it whether the venue is empty or holding sixty thousand people. Nothing else contends for it. That predictability is what a broadcaster is actually buying, and it is the reason satellite persists in an era of fast mobile networks.

Cellular uplink is contended and shared with everyone else on the cell, which is why a capacity test during the morning rig tells you very little about conditions at the moment of highest drama. The uplink direction suffers first and worst under crowd load. Consumer connections also throttle upload as a matter of design, which is a separate issue from congestion and is covered in the batch 1 article on symmetric and asymmetric upload.

The mitigation on the cellular side is diversity rather than a larger headline figure. Congestion at a venue is rarely uniform across operators, because coverage, cell density and backhaul all differ, and the network that saturates first varies by location and by hour. Weconnect provides non-steered access across 700+ carrier partnerships in 195+ countries, so an encoder draws on whichever networks are performing at that spot rather than a set fixed months earlier in a procurement round.

Coverage decides more events than latency or cost

The first question is whether usable cellular exists at the location at all. Where it does not, the comparison ends there. Open water, mountain stages, desert rally sections, deep rural coverage holes and large parts of a marathon route through terrain nobody built a mast for are satellite territory, and no amount of bonding fixes an absence of signal.

Where it does exist, the honest question becomes how much of it exists at the specific point the camera stands, at the specific hour of the event, on the specific networks the crew can reach. Multi-network access widens that zone considerably, because a location that is marginal on one operator is frequently solid on another.

Satellite has its own siting constraint, and it is easy to underestimate. A geostationary terminal needs a clear line of sight to a fixed point in the sky. Stadium roofs, tree cover, urban canyons and adjacent buildings block it, and the elevation angle falls the further north the event is, which makes obstructions worse in exactly the markets that run a lot of winter sport. A truck that cannot see the satellite is a truck that cannot transmit, and its parking position becomes a technical decision rather than a logistical one.

Notice period, mobility and the shape of the crew

Space segment is booked. The truck has to drive there, park somewhere with sky visibility and be operated by someone qualified to do it. That chain has a lead time, and the lead time is the reason breaking news rarely goes out over traditional SNG from the first hour.

Bonded cellular has almost no lead time. The kit is a bag, the crew is whoever is already closest, and the setup is measured in minutes. For newsgathering, for fixtures confirmed at short notice and for any production where the venue changes weekly, that difference decides the workflow rather than merely influencing it.

Mobility compounds it. A camera on a motorbike, a follow car or a boat is straightforward to serve with cellular and difficult to serve with a fixed terminal. Productions that move through the event, rather than watching it from a fixed position, tend toward bonding for the moving cameras even when a satellite path exists for the main feed.

The two cost models do not compare line by line

Satellite is a booked-capacity model. The segment is paid for whether the event overruns, finishes early or is cancelled by weather, and the truck and operator are day rates on top. For a long, high-value production with a fixed schedule that arithmetic is entirely reasonable, and the predictability is part of what is being bought.

Bonded cellular is consumption based. Data is paid for as it is used, which suits a calendar with heavy matchdays and quiet midweeks. Weconnect operates the broadcast side on that basis, so a fixture list with eight active days a month is not carrying capacity charges for the other twenty two. The SIM selection and commercial detail behind that sits in the best SIM for a LiveU encoder comparison.

The comparison that matters is cost per event including the crew and vehicle each path requires, measured across a season rather than a single fixture. A production that would need a truck and an operator forty times a year is a different calculation from one that needs it twice.

Where each path wins

The table below sets each condition against the path it favours. The same pattern holds across news, corporate live and sportuitzendingen connectiviteit.

Most serious productions carry both

The last row of that table is the one that decides real deployments. Bonded cellular and satellite fail for completely unrelated reasons. Cellular degrades under crowd load, in coverage holes and when a mast has a bad afternoon. Satellite degrades in heavy rain fade, when the sky is obstructed and when a terminal cannot acquire. A production carrying both is protected against each set of failures by the other, which is a stronger position than either path alone at any price.

That is now common in practice. Modern encoders bond a satellite terminal alongside cellular modems and treat it as another path in the same aggregate, so the transition under stress is gradual instead of a manual switch under pressure. Which path is primary depends on the event: satellite leads at the remote endurance event with cellular filling gaps, cellular leads at the city-centre fixture with satellite held as insurance.

The return path deserves separate thought here. In a remote production the link carries talkback, programme return and cues back to site, and those services have different tolerances from the outbound video. How that changes the specification is covered in the article on remote and live sports production.

What to establish before the event, not on the day

Four things settle most of these decisions, and all four can be answered in advance.

Which networks actually serve the camera positions, tested at a comparable time rather than during a quiet morning rig. What the notice period and the duration are, since those two together usually rule one path in or out on their own. Which services the link carries beyond the programme feed, because talkback and return frequently drive the design more than the video does. And what happens when the primary path degrades, decided while there is time to build it rather than during the event.

Weconnect works through those questions per production as part of its broadcast connectivity offer, including coverage mapping at the specific venues on a fixture list.

Veelgestelde vragen

What is the difference between bonded cellular and satellite for live broadcast?

Bonded cellular aggregates several mobile connections inside an encoder into one link, using public networks that are shared with everyone else in the area. Satellite sends the signal to a spacecraft on capacity booked in advance, so it is unaffected by local congestion and works where no mobile coverage exists. Bonding wins on speed of deployment, mobility and cost per event; satellite wins on predictability and on coverage in places mobile networks do not reach.

Is bonded cellular good enough for live television contribution?

Yes, and it is standard practice across news and below the top tier of sport, with growing use above it. A bonded encoder spreads the feed across multiple carriers so no single network failure ends the transmission. The requirement is sustained, consistent upload rather than headline download speed, plus enough network diversity at the location that congestion on one operator does not take the production with it.

What is the latency of satellite for live contribution?

A geostationary satellite sits at roughly 35,786 km, so the round trip is around half a second whatever equipment is used. That is workable for a one-way feed and awkward for two-way interviews or talkback. Low earth orbit services operate a few hundred kilometres up and bring the round trip into the tens of milliseconds, at the cost of more variability as the terminal hands off between satellites.

When do you still need an SNG truck?

When there is no usable mobile coverage at the location, when the crowd is large enough that public network capacity cannot be relied on, or when the production runs long and continuous on a fixed schedule where booked capacity is worth paying for. Line of sight is the practical constraint: a terminal that cannot see the satellite because of a stadium roof, tree cover or surrounding buildings cannot transmit.

Can you use satellite and bonded cellular together?

Yes, and on high-stakes productions it is the normal arrangement. Modern encoders bond a satellite terminal alongside cellular modems and treat every path as part of one aggregate, so degradation on one is absorbed by the others instead of requiring a manual switch mid-event. The two fail for unrelated reasons, which is precisely why carrying both is more robust than buying more of either.

Volgende stappen

Weconnect provides the cellular side of live contribution: non-steered multi-network SIMs across 700+ carrier partnerships in 195+ countries, upload-focused performance for contribution feeds, consumption-based data that follows the fixture list, and central visibility of every SIM in the field. Tell us where the productions are, how much notice you get and which venues have caused problems before, and we will map coverage per location and build the connectivity around it. Challenge us with your connectivity requirements. Direct response within 4 business hours.