Mick West's latest activity

I updated from 4o-mini to 5.4-mini, and it seems to work.

So when I use GPT-5 Mini, it gives UTC-3 (wrong). But switching to Claude Opus 4.8 it gives the correct UTC-4 So I guess the simple solution is to use the better AI model wherever possible. I'll see if I can update the ChatBase one - but...

Oh, wait, that's the "Sitrec Chatbot", not the Sitrec Assistant. I've not updated the underlying models there (or used it) for a while. Let me try the AI assistant...

Chile is in Chile Standard time after Apr 4, that it UTC-4, so add another hour. The chatbot made a mistake. What model do you have it set to in Sitrec/Settings?

ProPixel is just Billy's photography business; it's just him doing standard photography (weddings, portraits, real estate, etc). It's nothing to do with video or photo forensics. Based on his X timeline, he seems to have got caught up in the NJ...

How would you know?

If it happens again, then click on Help/Export Debug Log, and then attach that file here (see example attached) and note the file name of the KML that you are trying to load

Can you make this happen? Is there a particular sequence that demonstrates it. When dropping a KML, it only shows a dialog if there's more than two tracks in it, otherwise it just loads it. If the track is far away, you might not see anything...

@jarlrmai I agree that the airliner candidates dont seem to fit. I added some buildings to the stitch to try and get a better line of sight to the light, and after scrubbing the video back and forth I'm not sure that it even is a plane. It could...

Yeah, I've talked a bit with Philip about this. I suggested he start a thread, but he never joined up. It's a nicer-looking UI. Sitrec does suffer from being over-complicated. I actually just started a push towards modernizing the UI today.

Since we don't know the geometry of the camera, we can't really rule out an internal reflection that's different from other examples. And, no, I don't know what the geometry might be. You don't seem to have addressed the correlation between the...

Shut down the camera, say a Rolleicord, to a very small aperture (f/16). Place a cloaked figure in position and expose the film for five seconds—then cover the lens. Remove the figure and carry on with exposure for another 15 seconds. The result...

No, I think I was wrong there. The two bright stars have identical-looking little curves, so they can't have the same center of rotation, so it's really just all from camera motion without rotation. I think the "or quicker" is actually the...

If it were a simple motion line, or quicker, then yes, but I think this includes a rotation component - something that's not in the "Nudge" in Sitrec.

You can see the actual camera motion in the small arcs near the two brightest stars. Looks like a small bump and a rotation, maybe over half a second