Mick West's latest activity

I forgot all about fixing this. So now here it is in the new and improved Sitrec with 3D buildings. https://www.metabunk.org/sitrec/?custom=1/Object%20Falling%20Pembrook%20MA/20260727_195252.js It does not work perfectly, as the camera moves...

It's likely a helicopter, someone on Reddit has already done all the work so I'll share their comment This looks like a helicopter, probably something like an EC-135 or similar based on the tail, traveling away from the camera... not an object...

Without de-rotation mechanisms the camera's roll depends on the drone's bank because of the geometric constraints of the gimbal I created a gimbal simulation in blender and get those results : No drone bank, the camera pans from front left to...

Unfortunately, there is a "ground level" when you have 3D buildings enabled. The normal elevation data is too low resolition, so the "ground", (e.g. the street) is hard to determine. What's a building? What's a tree? What's a hill? I try to...

If you increase the elevation and the object does not move, that might be because the elevation (relative to sea level) is underground. If the position is underground, then the object is displayed snapped to the surface, so try a bigger number...

Please post a sitch where this is a problem (both the jumping and the "not adjustable" camera.

Update on my "Vasco60" work. I have been working on a slightly different version of the pipeline for some time. Code is not yet public. Also testing whether a looser, more permissive version can approach the third-party ~107K-scale candidate...

They look like Chinese Lanterns. Like these: Source: https://www.youtube.com/watch?v=kUegTtkUZz8

It's just too hard to see what motion comes from camera motion parallax, operator slewing, or actual object maneuvering. There's a background pattern, but we can't really tell what is fixed sensor noise, and what is possibly the ground. The...

I could create a Nimitz scenario generator from the verbal accounts. This might be a good test for the AI integration I've been working on.

I suspect it's using a different optical pathway to detect the laser returns and then overlaying them. The LTD seems likely as there's a flashing red dot that moves when the camera moves. That would be the laser on the ground.

Yeah, I'm leaning towards it NOT hitting the wire now. I'm playing around with trajectories, and it does not seem like the ball is high enough to hit the wire.

We'd need to see the video with timestamps, which isnt going to happen. But then we have Nyland, the goalkeep who protested immediately after the goal. As England celebrate, Norway are up in arms, no one more so than Nyland, who runs towards...

Highly motivated reasoning.

It's an interesting event given that we have multiple high-quality videos from different directions and yet two distinct camps of "I know what I saw" (see) have emerged.