Most bridges don’t need a video matrix. You need one when a display has to show a source it isn’t physically cabled to, and no amount of front-panel input cycling will fix that. So count destinations, not boxes. For every screen on the bridge, list the sources it must be able to show. If all of them can already reach that screen’s back panel, you have a control problem. If they can’t, that’s the routing problem a matrix exists to solve.
Getting this backwards costs money in a specific way. A matrix bought to fix a control complaint adds a processor, a network and a shared failure path to a bridge that never needed any. A control head bought when the real need was routing leaves you re-pulling cable after the console is machined. Neither mistake shows on a spec sheet. Both show at commissioning.
What Does a Video Matrix Actually Switch?
A matrix routes any input to any output. It takes the video sources on the vessel and makes each one available at each destination on demand. A control head does something different. It reaches out over a command bus to displays you have already cabled and tells them what to do. One moves signal. The other moves instructions.
Three separate layers get folded together in most switching conversations, and they cost wildly different amounts.
- The display’s own input selector. Seatronx marine displays run 10 to 36 inches with options up to eight inputs, including HDMI, DVI, DP, VGA and composite. A screen with eight inputs already holds eight sources, and choosing between them costs nothing extra.
- A remote control head. A small panel that talks to displays over a serial command bus: power, dimming, night mode, and which of a display’s own inputs is live. It changes what a screen shows, not what a screen is connected to.
- A true matrix. The layer that re-routes signal, and on the marine side it usually carries more than video. Seatronx describes its switching command center as modular video, touchscreen, keyboard and mouse switching, so one position can take over another station’s picture and pointing devices in a single action.
The keyboard and mouse half is its own decision and sits outside this article. Stay on the video side and the question stays clean: does the signal need to move, or does the instruction?
How Many Sources and Destinations Justify a Matrix?
Count pairs, not boxes. Walk the bridge and write down every screen that has to show a source it isn’t wired to today. That number is your requirement. Zero means no matrix. One or two short runs mean you pull cable. Past a handful, or when the pairs change by operating mode, switching gets cheaper than copper.
Notice how buyers phrase this when they shop. The most searched configurations are written as grids: 4×4, 8×8, 8×8 HDMI. That’s inputs by outputs, and the habit is a good one. The mistake is picking the grid before writing the pair list, because the grid is a purchase specification and the pair list is the requirement.
| What you have | What it usually is | Why |
|---|---|---|
| One helm, one dedicated screen per source. | Not a switching problem. | Each display’s own inputs already hold everything it shows. |
| Several screens on fixed sources, operators walking the console to dim them. | A control problem. | A command bus reaches displays you already cabled, for a fraction of a matrix. |
| Two operator positions that both have to show any camera, sounder or chart. | A routing problem. | The pairs outnumber the runs anyone is willing to pull through a finished console. |
| Modes that reassign screens for docking, fishing or a night watch. | A routing problem with presets. | The pairs change by mode, and rewiring isn’t a mode. |
| A required navigation station with its own processor and its own feed. | Keep it off the matrix. | Independence is the point, and routing it through a shared box removes it. |
The count is a timing decision as much as a sizing one. Capacity is bought in modules, so the planning question isn’t what you need at handover, it’s what you’ll need after the next two source refreshes. Ask for the current datasheet on module counts, control protocol and power budget rather than an old brochure. Those numbers move between revisions.
What Happens to Redundancy When Every Station Shares One Matrix?
A matrix concentrates. Before it, a dead source took out one screen. After it, a failed processor, network switch or power feed can take out every screen it feeds at once. That isn’t an argument against switching. It’s an argument for deciding early which sources are never allowed onto it.
The regulatory language here is older than the technology and still governs the answer. Setting out the SOLAS chapter V carriage requirement for integrated bridge systems, IMO writes that In case of failure in one part of an integrated navigational system, it shall be possible to operate each other individual item of equipment or part of the system separately. Read that as a constraint on your cable plan and not only on your alarm list. If the single path from a required sensor to its display runs through a shared switch, separate operation isn’t available once the switch is gone.
The split most integrators land on is a two-layer bridge. The situational layer goes on the matrix: cameras, engine and tank pages, secondary chart views, thermal imagery. The required layer keeps its own cable from source to screen, powered and routed independently, which is the same discipline behind what an integrated bridge system has to prove at type approval. That leaves the matrix free to do what it’s good at, which is putting the right picture in front of the right person quickly.
There’s a quieter dependency worth naming. The Seatronx switching processor is powered over Ethernet, which removes a power cable and makes the network switch part of the availability story. Whatever supply protects the equipment being controlled should protect that switch too.
Shared control also needs a rule about who is driving. The revised IMO performance standards for integrated navigation systems say only one task station should at any time be assigned to accept control commands, and that the station holding control has to be clearly indicated. A take-control feature that lets any helm grab a display is useful in heavy weather and dangerous the moment two positions both think they have it.
Where Does Remote Display Control Fit?
In more bridges than a matrix does. Most complaints about a multi-screen helm turn out to be control complaints. Nobody can dim eight displays at 0200 without walking the console, and nobody can restore a working layout after someone changed one. That’s fixable without touching a video path.
The PHT/SRT-CTL is a compact control head that daisy-chains up to twelve displays over RS485, using the NCOM or SCOM protocol, on a 9 to 36 VDC input. The panel is a 4.3-inch projected-capacitive touchscreen at 480 by 272 pixels, rated sunlight readable at 800 cd/m2, in an 88.2 by 139.9 by 40.1 millimetre body that flush-mounts into a console or sits on a desk. From it you power displays on and off, dim them, and select an input on any of the twelve.
Notice the boundary. It selects an input the display already has. It doesn’t deliver a source the display isn’t connected to. That’s the line between the two products, and your pair list decides which side of it you’re on.
When a switching question comes to us, the first thing we ask for isn’t a source count. It’s the console drawing with every existing run marked on it, because a destination that is already cabled to the source it needs doesn’t belong in the matrix math. Often enough the drawing shrinks the requirement to something a control head handles on its own.
Decide This Before the Console Is Machined
Architecture decisions have a deadline, and it lands earlier than most refit schedules admit. Once the cutouts are cut and the trays are closed, adding a run costs more than the hardware it carries.
Three things belong on the drawing before anything is ordered. First, the pair list: every destination, every source it has to show, and which of those pairs don’t exist yet. Second, the exclusion list: the sources that stay off the matrix because a required function depends on them. Third, the growth allowance: how many spare inputs and outputs you’re buying against sources that will land in the next refresh.
Everything after that is downstream. Interface choice, cable length and extender pairs, and sizing and mounting the screens themselves get easier once you know whether a signal reaches a screen directly or through a switch.
Frequently Asked Questions
What is a video matrix switcher?
It’s a device that connects a set of video sources to a set of displays and lets any source appear on any display. Products are sized as a grid of inputs by outputs, which is where names like 4×4 and 8×8 come from. On a vessel the inputs are systems such as chart processors, radar and cameras; the outputs are the screens.
Does a matrix replace a multifunction display?
No, and the two solve different problems. An MFD combines several functions inside one unit and one interface. A matrix leaves every system as it is and changes only where each picture appears. Adding an MFD usually reduces the number of external sources a matrix would have to carry.
Can one control head run displays from different manufacturers?
Only where they implement the same control protocol. The PHT/SRT-CTL uses NCOM or SCOM over an RS485 daisy chain, so a display that doesn’t answer that protocol won’t respond, whatever the cable looks like. On a mixed console, confirm protocol support model by model against current documentation first.
Does switching add delay to a camera feed?
Any device added to a signal path can add delay, and how much depends on the hardware and how the signal is carried. For docking, crane and machinery views, treat this as a number to obtain and measure on the actual configuration, not one to assume. Ask for the figure for the exact modules quoted.
Can a matrix be added to a bridge after it’s built?
Usually yes, but the cost lives in the cable rather than the box. Every source has to reach the switching point and every destination has to be fed from it, so the work is routing and terminating. On a certified vessel, ask first whether the change touches any function covered by the existing approval.
Bring the Pair List Before You Buy a Matrix
The useful version of this conversation starts with a drawing, not a catalogue. Once the destinations, the sources and the missing pairs are written down, the answer is usually obvious, and often smaller than the first quote. When it isn’t, the modular switching command center Seatronx builds for multi-display bridges is the layer that carries video, touchscreen and pointing devices together.
If a new build or a refit is at the console stage, talk it through with the Seatronx display team while the pair list can still change the drawing.