Current Nav Ball Is Wrong For Space Navigation

[NeoMorph]

I really didn't get this for a long time but the nav ball is basically an airplane attitude indicator and not a true space attitude indicator.

So what is the difference. Well, an air based attitude indicator works around the ground and the north pole. It uses a heading to let you know which direction you are going and has pitch going from -90 to +90 degrees. A space based attitude indicator won't have a heading because there is no north pole in space. Remember that some planets have their axis tilted compared to the earth so which would you use?

The space based indicator uses yaw instead of heading. It's easy to mix up the two but let's look at what happens when a plane pitches up.

Aeroplane Attitude Indicator.

The pilot is flying straight and level at a heading of 90 degrees. He pulls back on the stick and the pitch reading begins to rise... when he hits 90 degrees the heading flips to 270 and as he passes over the top the pitch reading reduces to 0 again when he is flying upside down and in the opposite direction.

Space Attitude Indicator.

The pilot is flying straight and level at a yaw indication of 90 degrees. He pulls back on the stick and the pitch reading begins to increase... but now when he hits 90 degrees the pitch reading continues to rise and will continue to do so until he does a full flip over at 359 degrees and then back to 0. The yaw indication (not the heading remember) will STAY at 90 degrees as it doesn't change when using a nav ball designed for space use.

As you can see, they look very similar but behave very differently. If you look at the ball used in the Apollo craft it has the three axes labeled 0 to 359 degrees...

It's been niggling me for a while and it was only when I was working with Rich (the guy who wrote Telemachus) that I realised the data I was getting was actually for an aircraft attitude indicator when I really wanted the space attitude data. Pitch should be 0 to 359 no 0 to 90 then back to 0 for the top 180. When you get a pitch of 80 degrees you have to check the heading to find the true position. If

I've been thinking we need a black and white nav ball for a while now (not the least to make it more useful for the colour blind peeps out there... me included) but we do need a nav ball that is labelled like the apollo version. In fact if you use the MechJeb Surf function it actually uses the true Roll, Pitch and Yaw values even though Yaw is labelled heading. Here is an example of what I mean...

Using the Apollo method means that if the pitch is reading as 80 degrees then it can only be one position on the ball. With the current method an 80 degree pitch could be TWO positions on the ball and you have to use the heading to find the actual pitch. Same with heading and yaw. You can set a heading of 180 and get an actual heading of 90 if the pitch is set to 180. But with the Apollo system when you have a yaw of 180 it will always be a yaw of 180 no matter what the pitch and roll is.

Perhaps I have it wrong but I feel that the nav ball currently in KSP is an aircraft attitude indicator/artificial horizon... designed for a fixed reference to north and the plane of the ground underneath the aircraft. What we need is a copy of the Apollo nav ball which is designed for a vehicle in space.

[Morty_jhones]

I sort of get what you is aiming at. but the nav ball we have will do as is for a long time. the thing is the nave ball is a external reference point indecator where as the Apollo nave ball is an Internal ref point indicater. TBH theirs no difference between them except for the fact the way they are marked.

the Apollo one gives all references relative to take off position

the normal one gives all references relative to a fixed point theoretically between the landing gear.

changing the nave ball would be a pain because they would have to implement at least 2 replacements. 1 for space planes and landers, 1 for rockets and probably another for rovers which would be a pain.

also what if the vehicle starts as a space plain, jettisons is wings in orbit an becomes a standard rocket then lands on a planet drooping of a rover? how would the game know to change the navball?

[NeoMorph]

Exactly my point. The problem lies in reference. I was trying to figure out what angle my physical nav ball was at when I came across this problem. Ball is at 80 degrees pitch... but it wasn't... it was actually at 110 degrees. Trying to figure out what the heck was going on brought this problem up...

The thing is we should just do away with heading and use yaw instead... would fix a LOT of problems that way. It allows the aircraft the transition from aircraft to spacecraft easily. I'm beginning to think that a customisable Nav Ball is going to be the way to go tbh.

[Diche Bach]

Once I get up to about 50km and until I'm in orbit, the current Navball confuses the heck out of me. I still dont' feel like I have an intuition for how to read it at those sorts of altitudes. At low altitudes and in space it is okay for me, though I still cannot fathom why the north and south keys seem to cause the thing to tilt in oppositive directions on the nav ball compared to the east and west keys. Meaning. D makes it tilt right, i.e., east, and A makes it tilt left, i.e., west. But the fact that the W Pitch adjustment (up/back) seems to make it point south, whereas the S Pitch adjustment (down/forward) seems to make it point north again, confuses the heck out of me.

I don't know if the Apollo ball would be really any better or not, but yeah . . . some kinda change to the NavBall would be good to me.

This one seems good for flying an aircraft, but not really so good for something that goes into space. Also the numbers on it are so tiny for me on my screen that I basically can't read them.

[Morty_jhones]

lol..... now that more than anything I have a problem with. but idf the nave ball was bigger it would get even more in the way

[Themohawkninja]

While you post a good reason for it to be changed to a proper space navball, what determines when we use what indicator? Would it be based off of the building the craft was made in? The type of CM? Would it be changable in-flight?

Also, the navball still makes a lot of sense as is:

Blue color = Away from parent body.

Orange color = Towards parent body.

0/90/180/270 (horizon) = North/East/South/West relative to the current body.

0/90/180/270 (not horizon) = References your aforesaid place on the horizon.

-90 through 90 = degrees of inclination.

[Ich Dien]

Also, the navball still makes a lot of sense as is

He's not disputing that, It's completely serviceable as it is. What he is disputing is having the attitude indicator for spacecraft be based upon a frame of reference that doesn't exist in space; the horizon.

[Themohawkninja]

He's not disputing that, It's completely serviceable as it is. What he is disputing is having the attitude indicator for spacecraft be based upon a frame of reference that doesn't exist in space; the horizon.

It's the point at which you are neither pointing towards nor away from the parent body.

[Scoundrel]

I completely agree. The navball needs two modes: ADI (ye olde artificial horizon for spaceplanes) and FDAI (ye olde 8-ball for spacecraft), with the ability to choose our reference frames as well. A simple button at either side of the ball would toggle modes and reference frames. I do understand the trepidation of others who are used to the current navball, so I would suggest an option for a "legacy" toggle, so that they can keep their navball. After all, if everyone wins, then everyone wins.

[NeoMorph]

I completely agree. The navball needs two modes: ADI (ye olde artificial horizon for spaceplanes) and FDAI (ye olde 8-ball for spacecraft), with the ability to choose our reference frames as well. A simple button at either side of the ball would toggle modes and reference frames. I do understand the trepidation of others who are used to the current navball, so I would suggest an option for a "legacy" toggle, so that they can keep their navball. After all, if everyone wins, then everyone wins.

Yeah... this is what we need. Did you know that Apollo had two modes as well. ORDEAL (Orbital Rate Display Earth And Lunar) which mimicked the plane type ADI and the normal interplanetary mode that took the reference from the IMU (Inertial Measurement Unit) gyros.

Here is a page from the Apollo training manual. It shows us Pitch, Roll and Yaw, not Pitch Roll and Heading.

Seeing your rates would be excellent as well... I mean your roll, pitch and yaw rates. On my hardware version I'm going to use the error needles to point me towards the next node...

Oh and I will jump for joy when they get around to removing that pink or magenta or whatever the target icon is... I just cannot see it when it goes over the orange part of the the nav ball (darn partial colour blindness). I have to turn the colour up high just to get a glimpse of it.

[Fifi]

I think it should be configurable (both colour and the mode) and the space one should be enabled by default and mentioned alongside the plane one in tutorials. There should also be a keyboard shortcut to switch the modes.

[Fett2oo5]

While you post a good reason for it to be changed to a proper space navball, what determines when we use what indicator? Would it be based off of the building the craft was made in? The type of CM? Would it be changable in-flight?

I don't see how it could be a problem to switch the NavBall midflight. It already switches from surface to orbit when you reach a certain altitude. When in orbit the SAI would be displayed, when in atmo the AAI would be displayed. The game already knows which body you are near. If you reach a certain altitude to any given world then the AAI would be displayed and it's reference would be the world/body you are located on.

[karolus10]

Nav-Ball used now is very good for atmospheric flight, but is inefficient for navigation in space

Now, we use scale grid with poles on "top" and "down", this solution would making sense for aircrafts, because they don't pitch UP/DOWN too often (if yes, You are in trouble).

"classic" style grid

On orbit this grid is far from perfect, because best grid resolution is near equator/horizon line (and are going worse on both poles) and we mostly using pitch axis.

Best Nav-Ball grid for space would be like this (and was used in Apollo and shuttle spacecrafts):

(Disregard the colors placement, nav-ball is in same position like on first picture)

In this Case orange north line is the horizon and Bold vertical line (horizon ;p) is an pitch axis (it could had 360* pitch scale marked on it).

Poles would be pointing normal/anti-normal axis that isn't used on orbit for most of the time (inclination change).

Also it would very useful to be able to switch nav-ball between relative to the ground (used now) and relative to spacecraft gyro (Arbitrary plane of the universe used by navigation gyro) so It would not move (now nav-ball is constantly rotating, because is showing attitude relative to surface) when orbiting around planet.

This mode can be invaluable when maneuvering with other crafts or estimating maneuver attitude.

Last thing that would be useful is ability to trim/reset attitude gyro, so we could (for example) deflect Yaw 0 line (with precise pitch scale), so we could launch/land easily at proper inclination by following pitch scale .

Of course, both grid scales should be switchable (in-game NavBall settings ?), because both are useful for different applications.

PS: I would like to get Black/White color scheme, it would made scale much more readable (even with denser grid) and made color markers stand out more .

EDIT_1:

Error Needles (yellow ones) are also very useful tool, because You can easily eyeball target attitude to every degree (scale range is switchable, so one line on scale can be 1, 5 or 10 degrees off target) by just keeping 3 needles in the center, making it much more accurate (and always had shown witch way turn, even if target is off the scale) than chasing the marker.

Also 3 White scales around FDAI are showing rotation rate (adjustable from 1,10,50 degree per second at full needle deflection), allowing to control rotation speed for heavier crafts (easier so see when slow down, if craft had large inertia), see if ship is not slowly drifting in some direction, or even gauge rotation to achieve tidal lock (not really useful, because warp is not remembering rotation now).

Edited August 6, 2013 by karolus10

[RexKramer]

Personally, I prefer the aircraft style ADI, with automatic reference to the body being orbited.

I don't think it would be terribly difficult to have the scale or color of the ADI be customized, maybe from display settings. I think most users will prefer the reference continue to be automatically set to the body being orbited.

I would guess that actual spacecraft would need and have the ability to change the reference based on the phase of flight. One benefit to the game automatically referencing the orbited body is you can always quickly determine basic orbital maneuver directions (prograde, retro grade, normal, anti normal etc..) with no additional cues.

[NeoMorph]

How do you determine north when the item in question has no magnetic field. Not every planet or asteroid will have a magnetic field after all.

I'm not saying you are wrong RexKramer but that the reference body is artificial in cases like that and that yaw is more reliable than heading. It also has the benefit of not flipping at some random angle and changing for some weird reason as it does when you pitch higher than 90 degrees. But like others have said, it would make more sense to have more than one system in place to cater to space and planetary navigation.

The big problem with the prograde/retrograde markers is as Karolus says as well... when they move behind the ball you have no idea which way to turn no matter which way up the nav ball is showing. My point isn't the automatic referencing, it's that the ball doesn't really show you what you need to know when navigating in space.

For example... You are going to do a mid course correction with your big assed interplanetary bus. You mistake the amount of power you have and are low on monopropellant. You have set up the node and now need to steer the craft to the right direction but the marker isn't visible as it is somewhere on the other side of the ball... WHICH WAY DO YOU TURN.

If you had the Apollo nav ball it would be a piece of cake as the error needles tell you EVERYTHING you need to know. They even tell you when the direction is coming round as they will start to move before the KSP marker would appear.

Apollo was planned down to the last nut, bolt and washer and they always used Pitch, Roll and Yaw... not Heading... and each value was 0-359... not 0-180, -180 to 0. There must have been a reason for that.

Oh yeah... one thing that I meant to say regarding the aeroplane style ball. Aeroplane nav balls have only TWO AXIS... Yes, even those don't use a heading. They use a compass for that. Now as there isn't a magnetic field in space there had to be a fix and that fix was to make the ball handle three axis of movement and name that extra axis the Yaw axis. It was that third axis that totally blew my mind when I was trying to figure out how they made the ball move in sync with the the spacecraft attitude.

In fact the ball in an aircraft is mounted in a C holder with the C normally in the horizontal position. When the ball is rotated in the mount it denoted the pitch. When the C clamp is rotated that is the roll.

In the Apollo FDAI they did is slightly different. They took a the ball, made it hollow and cut it into two. Then they took a disc the same diameter of the ball and put it vertically between the two halves and put an axis through the disc which would now become the pitch. The disc is then mounted in a C mount like the aeroplane version only instead of it being horizontal it is now vertical. when the disc rotates it is now denoting the yaw. When the C mount rotates that is the roll as normal.

Edited August 7, 2013 by NeoMorph

[karolus10]

About colors... even if 8 ball style indicator could be black and white, the classic horizon could use also common, Blue-black style... take a look :

[Skorpion]

I've always liked the idea of having a screen to customize your own HUD. It seems only natural that if you can design your own rockets, you should be able to design your own cockpit controls. That way you could start out with the basic navball, but advanced users could swap it out for an Apollo-style navball, or maybe a docking indicator. Rovers, on the other hand, could have an entirely different HUD that's better for driving.

I'm not convinced the manpower required to implement this this kind of customization is really worth it, but it world certainly be a welcome addition.

[DerekL1963]

How do you determine north when the item in question has no magnetic field. Not every planet or asteroid will have a magnetic field after all.

By defining one direction as North and entering that data into your GNC.

[TMS]

Low priority I feel.

One for the modders.

[innovine]

I would LOVE LOVE LOVE a better navball!!! It's my favorite feature!

[NeoMorph]

If you guys are wondering why it is taking so long to make my own navball (actual physical navball) this is the reason (just check the price and remember this is for a TWO axis aeroplane nav ball... not a three axis one like KSP and Apollo used)

http://www.ebay.co.uk/itm/STAND-BY-ATTITUDE-INDICATOR-GOODRICH-MODEL-AI-330K-PART-NO-501-1568-10-/190876189612?pt=UK_CPV_Aviation_SM&hash=item2c711b83ac

I'm pretty much making mine out of sticks and stones and ducktape in relation to that. If it works I think I'll probably pass out... and drop the thing lol.

[Harguinxx]

I vote for it as an option!!!!!

[NeoMorph]

I do wish the values for Pitch, Roll and Yaw were available natively.... Rich (who writes the Telemachus telemetry plugin) has been great to output the pitch and yaw in a 0-359 degree scale.

If it wasn't for Rich I wouldn't be able to reach step one of my hardware version. Unfortunately the thing is really kicking my butt at the moment because although I can get it to work at medium speeds I tend to start to get lag when I go into a fast spin. I just need to get a system that can handle ANY speed without generating lag.

Another thing that would be cool would be to replace the icons floating on the ball (which are hard to find when they are on the opposite side of the ball) with the three error needles. For example the say the pitch, roll and yaw are all at 0 degrees. The node angles are Pitch: 110 degrees, Roll: 0 degrees and Yaw: 0 degrees. The three error needles would show that you need pitch up but keep the yaw and roll the same If you had the current method of angle display, once you pass 90 degrees pitch your heading would flip to 180 degrees. Yaw however would stay the same.

But the beauty of using the three needles error needls is that they show you which direction to go where the current icons on the ball don't.

[karolus10]

I won't think that replacing markers with error needles is a good thing, because You can see multiple markers floating on the ball when error needles only show Your deviation (AKA: where you should point your ship) from set target.

Best solution would be keeping markers intact and letting player to select target for error needles (Automatic, one of orbital vectors, maneuver node, target attitude for docking ,etc).

Selected mode could be represented by icon (showing mode or vector symbol) placed near the NavBall.

Error needles would be extremely useful if flight planning would include planing trajectory, so You could design ascent path to get at properly inclined orbit and then player could follow planned attitude by keeping needles in the center.

Edited August 14, 2013 by karolus10

[maltesh]

How do you determine north when the item in question has no magnetic field. Not every planet or asteroid will have a magnetic field after all.

North isn't defined by the magnetic field. It's defined by the axis of rotation of the body you're on

And it's also currently defined differently based on whether the object is a Planet or Not., but I digress.

At any rate, you find the appropriate direction along the rotational axis, and then can use gyroscopic instruments to keep a sense of which direction that is.