Video on the same map as everything else

The vantage point nobody can occupy

Teams almost never ask for video because they want more screens. They ask because there is a specific place that matters and no good way to keep a person in it — the roof overlooking an event, the gate at the far end of a site, the stretch of ground that would take an hour to walk and thirty seconds to fly.

Positions on a map tell you where your people are, which is the harder half of the problem and the half TAK solves first. Video answers the other question, the one that usually comes next: not where is everyone, but what does that place look like right now. Having both in the same client, on the same network, is the difference between an operating picture and a map with a to-do list attached.

Getting a feed in

Create the stream in the console, from the video feeds tab of the network it belongs to. You give it a name, choose RTSP or RTMP, set the quality and set how many people should be able to watch it at once. The console hands back an address — protocol, host, port and path — and that address is the entire configuration.

Put that address into whatever is doing the publishing — and publishing is the load-bearing word. The server accepts a stream pushed to it; it does not go out and fetch one. That is worth checking before you buy anything, because the typical fixed IP camera works the other way round: it hosts a URL for you to pull from, and supporting RTSP is not the same thing as being able to publish over it.

In practice the rule is short. If the device can publish, point it at the address and you are finished. If it can only serve a URL, put an encoder between it and us, which is the ordinary answer for cameras that are already on a wall.

• Cameras whose firmware can publish RTSP or RTMP, which point straight at the address
• An encoder in front of any camera that only serves a pull URL
• A laptop running ffmpeg, which is how most people test a feed before trusting it
• Phone applications that publish RTMP, for a body-worn or vehicle view
• DJI aircraft, which publish over RTMP

Getting a feed out, without sending anyone an address

Operators open the video player in ATAK or WinTAK and choose to download the feed list from the server. The feeds you created are already there. Nobody types an RTSP address into a tablet, nobody reads one down a radio, and a camera added an hour into an operation reaches everyone without a second round of instructions.

Worth being precise about what that convenience is and is not, because the wording invites a wrong reading: nobody has to type an address, but the address is still the thing protecting the feed. Anyone holding a stream path can watch that stream. Treat the path the way you would treat a password, and if one gets out, delete the feed and create it again — the new path is unrelated to the old one and the old one stops working.

One limit is worth planning around rather than discovering: video streaming does not work on iTAK. An iPhone joins the same network and sees the same map, and no setting or hand-typed address changes the video part — it is ATAK on Android and WinTAK on Windows that can watch. If it matters that a particular person can, put them on one of those.

Why every stream has a ceiling

A feed is created with a quality and a maximum number of simultaneous viewers, and the server enforces both — it measures what is actually arriving and drops a stream that sends more than it was set up for.

That is deliberate, and it protects you rather than the server. A camera left on a default profile, pushing several times what anyone intended, is one of the most effective ways to degrade a shared network for every other person on it. Losing one feed is a considerably better outcome than everybody’s map going slow at once. The practical version of this: set the quality to what you will genuinely watch on a tablet over cellular, not to what the camera is capable of.

What it costs and what it needs

Video is a paid-plan capability rather than part of the free tier, and what a feed costs scales with two things you choose when you create it: the quality of the stream and the number of people who can watch it at the same time. A modest feed with a couple of viewers and a high-quality one the whole team watches are genuinely different propositions, and the console shows you which one you are creating before you commit to it. The pricing page covers the plans.

Beyond a plan, three things. A network for the feed to belong to. The permission to create video streams, which is a role setting, so it can sit with the people who run operations rather than with whoever happens to hold the account. And enough spare capacity on the plan to cover the feed — that is the one teams meet unexpectedly, because video draws on the same allowance as your users, your devices and your federation. An organization already using all of it adds capacity before it adds a camera.