Navball changes

[DrD]

Navball as it's modeled is unrealistic. Brown is always "down" and blue "up," ie towards nearest planet or away. This means that when you're in an inertial orbit (one in which you do not rotate) navball moves and it appears pitch is changing. Pitch is in fact changing with regards to ground but not with regards to 3D space. And in fact craft is not rotating, only thing that changes is its aspect with regards to planet.

IRL (at least in Apollo) the navball was fixed to an arbitrary zero point (more on that in a bit) and didn't move unless craft actually rotated. However, there was disagreement between astronauts and engineers. The engineers favored the model I just described, but the astronauts who were all pilots wanted an artificial-horizon type device at least when in orbit. So they got both. When in orbit they activated a device called ORDEAL (Orbital Rate Display Earth and Lunar.) This basically rotated the navball such that is worked just like it does in KSP. Autopilot kept it steady with the result that craft always maintianed same orienttion to earth during orbit, and thus always pointed prograde and made 1 complete 360 degree pitch maneuver in each orbit. (as an aside, they were "heads down" in that earth was above them in reference to the seated position. This was because 3/5 windows faced mostly up, and sextant/telescope was on opposite side and naturally had to face stars.)

Once they left orbit they turned off ORDEAL and went back to the mode in which zero point of navball did not change.

Of course in space the zero point was arbitrary and they actually changed it several times during mission. Zero point had the jaw-breaking name of REFSMMAT (reference to stable member matrix.) They changed it for 2 main reasons:

1. it was convenient during a burn to have the desired attitude be zero in all 3 axes so that if astronauts had to control manually they could just hold the navball centered. (In KSP this would be like having a maneuver prograde be zero point.) So once they calculated the attitude for a burn they would usually change the zero point as described.

2. There were some attitudes in which they could experience "gimbal lock," a situation in which at least 1 gyro was no longer free to move (or actually not move as craft moved around it) such that it would get torqued out of position. This was due to having only 3 gimbals and could easily be solved with a 4th gimbal but for some reason they left that out. The attitude was marked on navball as a redzone and they had to make sure redzone was not on reference point. If a maneuver could result in an attitude that would cause gimbal lock they would simply change the zero point.

So in KSP I think there should be 2 modes of navball: orbital and inertial. Orbital works just like it does now. Intertial would mean that indicated attitude doesn't change unless craft actually rotates. In this mode you might want to have different (or no) colors to remind one that it's no longer in reference to the ground but to 3D space. In this mode there could be buttons and/or hotkeys to change the zero point, one shortcut would be to change zero point to current prograde, another to maneuver prograde, perhaps also to retrograde and target prograde. Maybe a manual option to change zero point in any axis to a desired value, although I'm not sure why one would would do that. Of course it's not critical to be able to change zero point with all the nice markers and such but it would be easy to implement I think.

I doubt we'll ever simulate gimbal lock, but if they did as part of an Apollo pack then you would have to be able to change zero point.

Seems like this could be a simple change, perhaps a mod.

Thoughts?

Edited February 20, 2014 by DrD

[sjwt]

Have you considered looking into Orbiter?

[Kasuha]

I am not against the idea of inertial navball, I just think it would be of only marginal use. Rotating the craft by certain exact angle would be possible with it, and it could perhaps be useful for rovers or boats, but with features available in KSP I never needed to do that with a spacecraft.

[DrD]

1. I've not used Orbiter but I think KSP is better and certainly much bigger ie more content. And Kerbals are cool. I'm not going to change my space sim over a small issue like the navball (I hope I didn't give the impression that it's a big deal.)

2. True it's not important but it does add a bit more realism and I think is an easy implementation. Of course coding being what it is it could be harder than I think.

[sjwt]

2. True it's not important but it does add a bit more realism and I think is an easy implementation. Of course coding being what it is it could be harder than I think.

The realism you are asking to add is only specific to one point in time, if you want to limite KSP to apollo era tech, then Mods would be your best rout.

[majic79]

I'd be happy if up/down keys worked the same way as left/right keys - I want to invert up/down for a flight sim where I'm not using the artificial horizon as my attitude indicator, it's almost (but not quite!) completely nonsensical in KSP to my mind and the number of times I've got up/down crosswise due to a rotation and re-alignment of my axis in relation to the dockng target - grrrrr

[allmappedout]

This argument comes up a lot of times. The problem here is both playability and available data.

From a playability standpoint, it makes it a hell of a lot easier for a new guy to understand that ground is always brown, regardless of what body you're floating around, so pointing away from it is a radial burn, and so on and so forth. It is just much more intuitive than the apollo system.

Secondly, the apollo team had a flight computer to calculate the burns and navball position to perform them, as well as a team back on the ground feeding them data based on their telemetry. Plus the apollo team could use star charts to work out their orientation, etc in case of system failure.

We don't have the same amount of information as they do, and by changing our navball to that type, we are LOSING information from the game (ie: which way to the planetary body), which is not easily gained without adding another dial, or data.

So I'm a huge no.