Weird behavior directly over North Pole?

[sndrtj]

Hi there,

For a survey contract on the other side of Kerbin I flew a plane over the north pole. Everything went perfectly fine, up until I flew directly over the pole. I suddenly spiralled left and fell from approx 18km to 14km in half a minute or so. Normality restarted a minute later or so. Is there anything special about flying directly over a pole? Or was this just bad luck?

[Anquietas314]

There is an issue where the camera and navball freak out over the north (and south) pole, but that usually doesn't affect anything except perhaps making you a little nauseous. I've certainly never experienced an issue where the plane/rocket I was flying suddenly didn't work properly. Maybe post a screenshot of the craft in question (with CoM/CoL/CoT visible)? It's possible the timing was just a coincidence.

The camera issue's caused by using a coordinate system that doesn't handle spheres very well; physics uses a different coordinate system.

[sndrtj]

Well, it did look like the view was turning around, rather than the plane doing anything. I had SAS on at the moment; could that be mixing up the two coordinate systems?

[Anquietas314]

I had SAS on at the moment; could that be mixing up the two coordinate systems?

I don't think so, but I suppose anything's possible. I don't know for sure which coordinate system SAS uses, but I would imagine it'd be relative to the ship/plane itself, which means it shouldn't matter where the plane is, what the camera's doing, etc.

[sndrtj]

I recreated the flight, and you're right, it's only a camera thing and the navball. Guess I got confused with that the first time it happened. Of course, the navball flipping around makes sense, but the camera changing view is a bit misleading. Proof: the following pics were made with me not pressing any buttons (except printscreen ), camera shifted automagically.

Moments before reaching the pole

Weirdness starting

More weirdness

Hi there, sun!

Nearly completely shifted.

[Anquietas314]

Of course, the navball flipping around makes sense, but the camera changing view is a bit misleading.

Good to know it's not a nasty physics bug (or, let's hope...). The navball and camera probably use similar coordinate systems.

If you want to get into the maths of it, both probably use the Euler angles system. It's nice for cameras because it makes using the mouse to control it much easier, but it suffers from "gimbal lock" (nothing to do with the ingame one) directly above and directly below the target coordinates ("where the camera is looking at"), which in the case of a planet in KSP is the north and south poles.

The camera's a little more complicated mind since you're tracking the plane, but the camera has a "position" in world coordinates, which in the case of surface/orbital camera is relative to the surface of the nearest planet, meaning you need a coordinate system based on angles.

[hyperbolicj]

It's similar effect to when you get into an orbit. The camera flips to the front of the vehicle for a moment.

[Anquietas314]

It's similar effect to when you get into an orbit. The camera flips to the front of the vehicle for a moment.

That's actually a completely separate thing; the game's just switching between two camera modes, which happen to be in two completely different orientations.

[PB666]

It's similar effect to when you get into an orbit. The camera flips to the front of the vehicle for a moment.

Just so as not to confuse folks, the camera switch occurs in my experience when your periapsis crosses between 23,000 and 24,000 meters. Technically that is not an orbit, because on just about any speed craft if your periapsis falls below 32,000 km drag exerts such a retrograde force that the orbit will never again exceed the atmosphere. You could approach at much higher speeds (say 6000m/s) but from the few times I have done this the vehicle goes white hot at high altitude reentry and disentegrates (or undergoes compression g-forces). For the missions orbit is any celestial bound eclipse in which the apoapsis is a positive number greater than atmosphere ceiling (Kerbin = 70k) and the periapsis is also greater than 70k. On the mun it is somewhat different because you could orbit at a low enough altitude that you crash into terrain, there is no atmosphere.

- - - Updated - - -

Just so as not to confuse folks, the camera switch occurs in my experience when your periapsis crosses between 23,000 and 24,000 meters. Technically that is not an orbit, because on just about any speed craft if your periapsis falls below 32,000 km drag exerts such a retrograde force that the orbit will never again exceed the atmosphere. You could approach at much higher speeds (say 6000m/s) but from the few times I have done this the vehicle goes white hot at high altitude reentry and disentegrates (or undergoes compression g-forces). For the missions orbit is any celestial bound eclipse in which the apoapsis is a positive number greater than atmosphere ceiling (Kerbin = 70k) and the periapsis is also greater than 70k. On the mun it is somewhat different because you could orbit at a low enough altitude that you crash into terrain, there is no atmosphere.

32000 M , not km

[Anquietas314]

PB666: There's an edit button for fixing those sorts of mistakes