Why the app had to be camera-agnostic

I started with a GoPro Hero 7. My wife got tired of me always messing with it and missing the game, so one Father's Day she gave me a Mevo Start.

It was better. I found the Mevo MultiCam app and I was hooked — multiple shots, live switching. Once I saw what was possible, I had to figure out how to do it for our games. But the math never worked for me: $500 for another Mevo Start, $1,300 for a Core and lens in the outfield, and on top of all that, the Mevo MultiCam app needs a paid Pro subscription to enable NDI input — which is the whole reason you'd want the second camera in the first place.

That was one of the things that started me thinking about building something with other options.

I found a 4K POV camera from ZowieTek and bought one because it was about half the price of a Mevo. Great little camera — 4K, native NDI HX3, plus RTMP, SRT, and RTSP straight from the camera. No PC required; it could stand on its own. PoE or USB-C for power.

It was a network camera, and by no means waterproof. So there was a whole process to make it work at a softball field: a Vonets adapter and a power bank at each camera. PoE doesn't help when there's no outlet at the fence. All in, it was still cheaper than the Mevo.

The first one I bought had a manual-focus M-mount sensor — the kind security cameras use — so I could put any lens I wanted on it, from a 35mm to a fisheye. I did a bunch of experimenting with mixed results. Then they came out with an autofocus version, and a C-mount version that opened up a much wider lens selection. I bought a C-mount with a zoom lens to be my outfield camera, and two autofocus units (which work similarly to the Mevo) for the sidelines.

So my setup became: Mevo Start at the backstop, ZowieCam C-mount + zoom in centerfield, ZowieCam autofocus on the first and third base fences. All running through NDI in the Mevo MultiCam app.

When it all worked, it was good. But it didn't always work. NDI sometimes wouldn't get discovered. That number of cameras maxed out the Mevo app. The network was always a challenge.

We were at a tournament in the Tri-Cities — for those of you outside Washington, those are smaller towns about five hours from Seattle. We had a late game, it was cold, and we'd been there all day. In the rush to get out, I left the Mevo on the fence and went back to the hotel. Came back the next morning and it was gone.

That was my most reliable camera. And I still wasn't willing to spend that much again.

I figured the way to make NDI work better was to make sure the network was rock-solid. If I had to invest in something, it might as well be the network — that way I could rely on the ZowieCams as my main view.

The whole time, I kept looking at security cameras and wondering why they wouldn't work. They're waterproof. Bullet-proof. Simple. Cheap. There just wasn't a way to get those streams into a live stream.

That became another key consideration for anything I built: it had to be camera-agnostic.

Why should the app discriminate against lower-budget cameras? I looked all over for a way to do it and couldn't find one. So I wrote it.