Recent content by Mick West

Minor point, but there are no clouds in the video. The camera is looking down at the ground. The light area at the bottom is a ridge, but the upper area is ground beyond it

Here's the official release combined with the earlier leak's OSD, this actually give slightly more motion info, as you can see more of the background behind the redaction boxes - although the leaked version has some duplicated frames

This video demonstrates an objection to the camera-motion hypothesis posited by Marik von Rennenkampff, on X. He notes that the camera-relative trajectory of the object changes and suggests that this implies it is independent of the camera motion. However, he seems to be missing that the...

Yeah, my bad.

Some others for reference: PR034 - less clear cut, but still blinking, then change direction PR038 - (Chandelier) single blink before loss. [EDIT: 38, not 36] I don't think it's a hard timeout. I think if the track quality is worse, it will end sooner.

PR098, It gets a false lock, start blinking pretty quick, then when it ends, the tracking stops.

PR078, a pattern is emerging.... Loss of lock, blinking, then zip-off as the camera changes direction. I think this is pretty definitive at this point.

And from PR077 Less dramatic here, but still visible. The white dot on left changes apparent direction at the end of blinking as the camera speeds up.

The blinking mentioned also happens here. From PR59 So it's blinking because it's losing confidence in the lock, and then it just transitions to the other mode and we get a zip-off.

Looking over them again, I think the 25% one is the best one to use. It's longer than the 5% and it has the smallest redaction boxes, and is slightly better quality than the 50% one

Some thermal cameras (like mine) will shut down the instant the sun is in frame, dropping a shield in front of the sensor.

Okay, here's the unique frames from the 5% segment (I added a "unique frames only" option to Sitrec specifically for cases like this - well, specifically for THIS case, but I imagine there will be other usages). Then there's longer segments derived from the 50% and 25% clips. This is...

Speaking of missing frames, there's a segment of the video at 1:22, titled "5% Speed". It actually seems more like 10% speed, but it also seems to have all the frames (obviously duplicated). I think THIS is the real best footage of the zip-off. I'll extract the unique frames...

Definitely, you can see it move so it stays lined up with the heat source (probably a gas flaring burn), and the center of the screen.

The lock was already lost (the object is not boxed). We are seeing a transition from RPOINT (attempting to hold a lost lock) to RATE-G (manual tracking). This appears to be something the operator did as the object was about to drift off-screen. They likely did not anticipate the parallax...