Which Gimbal scenario is more likely — 30 NM or 10 NM?

[Mendel]

But if we consider that -2 means -1.5 to -2.5...then -2.593 is a trifling 0.93 degrees off that upper range and is well within that .35 degrees FOV. If the -2 is actually -2.5, then the true horizon at -2.593 is 1/3 of the field of view below the cross hairs in the video.

Yes. How good are you with fractions?
The crosshairs are at 1/2 FOV, minus 1/3 FOV means the horizon sits at 1/6 FOV height, closer to the bottom edge than to the crosshairs by a factor of 2. Does your screenshot look like that? No.

 

[Mendel]

True, why I think a readjustement on target of 2-3° at each step rotation (Sitrec Gimbal sim) would cause more than a little bump in the image.
This implies 6-8 FOV are crossed one way by the pod roll to catch up, t

A pod roll of 2-3⁰ doesn't affect the FOV by 2-3⁰. If the ATFLIR was an airplane, banking it 3⁰ wouldn't make pitch or yaw change by 3⁰, they'd change hardly at all. @Mick West did a lot of graphs on this before sitrec where you can see this exactly.

I don't see evidence for the camera readjusting by 6-8 FOV, when looking at the bumps frame by frame. It looks like a minor readjustment. I know of a quote from Lacy Cook (patent holder for the ATFLIR gimbal system) saying that the range of the internal pointing/tracking/stabilization mirror is "something like XXX micro-radians". 999 micro-radians=0.05°.

FYI we also have another Raytheon engineer on record saying that the "pitch/roll/yaw motors are constantly working" (Twitter).

Could you find us the sources for this, please?

 

[TheCholla]

It's not 2-3⁰ of pod roll, it's 2-3⁰ of deviation from target (where the pod is looking, versus where the target is).

The first quote is from a private exchange L. Cook had with Marik, who is the co-author of our Gimbal paper so I know about it.

The second one is public on Twitter:

Source: https://twitter.com/OldManTenno/status/1635132382759223296?s=20

In further private discussions with the Raytheon engineer, he added that from his experience, he had only seen a step motion for the pod in case of malfunction, when motor resolvers are dirty and "stuck", and need more voltage to kick in.