Drifting maneuver marker

[A_name]

What causes the blue maneuver marker on the navball to "drift" away from the heading indicator when time warping, so that you usually have to readjust before performing the burn?

[Deddly]

I have been wondering the same thing! I moved the question here to Gameplay Questions so we hopefully get the best answers.

[bewing]

If I understand your question correctly, you have it the wrong way around. It's the heading indicator that changes direction as you orbit in your SOI. The blue maneuver marker always points in the direction that you are going to burn when you finally get to the maneuver node.

 

[A_name]

3 hours ago, bewing said:

If I understand your question correctly, you have it the wrong way around. It's the heading indicator that changes direction as you orbit in your SOI. The blue maneuver marker always points in the direction that you are going to burn when you finally get to the maneuver node.

 

We'll yes but what I mean is, if you move to point exactly at the maneuver indicator then hit time warp, in theory you should stay right on it because the ship will stay stationary during warp, yet the markers shift away from each other. 

[Blaarkies]

22 minutes ago, A_name said:

We'll yes but what I mean is, if you move to point exactly at the maneuver indicator then hit time warp, in theory you should stay right on it because the ship will stay stationary during warp, yet the markers shift away from each other. 

I have only seen this on old nodes(those that are a few orbits in the past already), and sometimes while burning(especially if you are adjusting on the second node while burning on the first). But i have never seen this during normal timewarp, the craft point at the dark-blue node marker, like @bewing pointed out. This because the craft does not rotate relative to the stars, and neither does the node...but the prograde marker doesn't care about stars, it follows the orbital line, which in turn follows the trajectory due to the gravity well.

Maybe you maneuver node is moving forward/backwards in time?
(i get maneuver nodes moving around phantom style when I do Mun slingshots to Minmus, the second maneuver doesn't stay at Minmus Pe but moves forwards and backwards)

[Spricigo]

 

I also notice this behaviour happening sometimes, and more often the marker also shift when I switch to another vessel(or KSC/TS) and back.

My guess its that something is getting recalculated during warp/switching and the marker adjusted accordingly, but no idea why. Its a bit annoying but not really an issue, so I didn’t bother to further investigate.

[sal_vager]

If I remember correctly it drifts because it is showing you where you need to burn relative to your position and the position of the manoeuvre node, which its rarely exactly the same.

[Kryxal]

It seems to happen mostly in lower orbits, I find ... but it feels like in the course of a single orbit, it might move 15 degrees or more.

[A_name]

Well it seems like some people have never encountered this. I'll try to recreate the behavior tonight and post some screenshots.

[DerekL1963]

It's been drifting for me ever since I started playing in .19...  Never worried about it, just drop out of warp a minute or so early and re-align your vehicle.

[bewing]

4 hours ago, A_name said:

Well it seems like some people have never encountered this. I'll try to recreate the behavior tonight and post some screenshots.

Heh. Well, I always need to burn before I get to my maneuver node. So pointing the ship's nose at the blue marker is never useful. And I don't do all that much timewarping. So that's probably why I don't see it.

[A_name]

3 hours ago, DerekL1963 said:

It's been drifting for me ever since I started playing in .19...  Never worried about it, just drop out of warp a minute or so early and re-align your vehicle.

That's what I do as well, but I'm still curious why.

 

[Lybodo]

If this is the issue I'm thinking of, it occurs because of a difference in frames of reference.  I may be mis-remembering the exact details, but I seem to recall hearing that the maneuver node is pinned to the navball relative to a fixed, system wide coordinate system, whereas the prograde, retrograde, and other markers are fixed relative to the current orbit's coordinate system.  As the body you're orbiting moves round the Sun, these two will slowly drift from each other.

[Kryxal]

But it doesn't matter if prograde and retrograde don't follow the node, your CRAFT should be in that same fixed coordinate system under time warp.  The issue here is that the craft which was aligned with the node initially loses that alignment during time warp.

[swjr-swis]

I see the behaviour mentioned here regularly, mostly in low orbits, and if you pay attention you'll also see the dV of the maneuver node change slightly.

I imagine it's for a reason similar to what causes Ap/Pe and AN/DN to keep drifting even when you're supposed to be in a stable orbit under zero acceleration. From the devnotes of previous versions it's clear that there's been several iterations of tinkering with the orbit stability already, but so far it's yet to squash this bug.

[monstah]

On 08/03/2017 at 1:17 AM, A_name said:

We'll yes but what I mean is, if you move to point exactly at the maneuver indicator then hit time warp, in theory you should stay right on it because the ship will stay stationary during warp, yet the markers shift away from each other. 

I've noticed that too; my guess is that something in the physics fails to keep the ship in the same orientation; in orbit it's hard to judge what's moving since everything is moving in relation to everything.

[Ten Key]

I see this a lot in Kerbin orbits below about 100 km. Above that threshold, the problem seems to mostly go away. But in a low orbit of 70-80 km, I can watch the maneuver indicator drift away from the heading indicator with no time warp at all. I usually have to correct after 1-3 minutes or so. 

[Red Iron Crown]

@Claw explained this one to me a long while back. Below a certain altitude, the frame of reference for the SAS but not the maneuver node is synodic rather than sidereal (I may have which is which reversed, memory is hazy). So as the planet revolves around the sun the direction pointed to by SAS and the direction of the maneuver node change relative to each other. (Test this by timewarping for about 106 Kerbin days, an equatorial node will have drifted by close to 90 degrees). 

The workaround is to make your parking orbit higher than 100km, then the frame of reference changes and the node and SAS stay put. (For Kerbin, other bodies will have different thresholds which I can't quote.)

Edit: The above is not entirely accurate. Below 100km the SAS reference frame rotates with Kerbin's rotation, not its revolution. Props to @jonny for performing an experiment which exposed this error. 

Edited March 14, 2017 by Red Iron Crown

[jonny]

On 10.3.2017 at 11:07 PM, Red Iron Crown said:

@Claw explained this one to me a long while back. Below a certain altitude, the frame of reference for the SAS but not the maneuver node is synodic rather than sidereal (I may have which is which reversed, memory is hazy). So as the planet revolves around the sun the direction pointed to by SAS and the direction of the maneuver node change relative to each other. (Test this by timewarping for about 106 Kerbin days, an equatorial node will have drifted by close to 90 degrees). 

The workaround is to make your parking orbit higher than 100km, then the frame of reference changes and the node and SAS stay put. (For Kerbin, other bodies will have different thresholds which I can't quote.)

I've tested this, and you're right. There are different frames of reference for given altitudes. The maneuver node will always stay in the sidereal frame of reference, but the SAS of a vessel below a specific altitude (i.g. 100 Km for Kerbin) will hold the attitude in a frame of reference which rotates with the planet (for Kerbin 360° every 6 hours)

A small example:

Mission time: 00:00:18; Vessel below 100 Km; Target at KSC

some orbits later mission time: 02:19:20; Vessel below 100 Km; still pointing at the target at KSC
Kerbin rotates 140° in 2:20:00h

The same experiment above 100 Km :

Mission time: 00:00:23; Vessel above 100 Km; Target at KSC

some orbits later mission time: 02:10:52; Vessel above 100 Km; Vessel pointing at +37° while the target is at -90°
Kerbin rotates 130° in 2:10:00h
I would call the error of 3° as a bad timing for my screenshots, but the difference is obvious.
note: a vessel above 100 km behave exactly same as a vessel below 100 Km if you time warp at KSC.

according to https://en.wikipedia.org/wiki/Synodic_day and https://de.wikipedia.org/wiki/Siderische_Periode I do not think the reference frame for vessels below 100 Km is Synodic.

Do not quote me about the following, as it only states how i understand the above mentioned Wikipedia articles.

Some explanation about the synodic and sidereal Refrence frame:
Pictures are better then words:

note: the origin and the orientation of the sidereal reference frame may be completely wrong, but the point is the sidereal reference frame does not rotate in reference to our milky way, while the Synodic reference frame rotate with Kerbin around the Sun.

 

Edited March 14, 2017 by jonny

[Red Iron Crown]

@jonny Thanks for sharing your experiment.

I think I might have misspoken though, from your experiment it looks like it's not synodic vs sidereal, it's absolute vs rotating. If it was synodic it would rotate through 360 degrees once a year; instead it appears to do so once per day. So it's matched to the planet's rotation rather than revolution.

Great job.

[Daniel Prates]

On 08/03/2017 at 1:17 AM, A_name said:

We'll yes but what I mean is, if you move to point exactly at the maneuver indicator then hit time warp, in theory you should stay right on it because the ship will stay stationary during warp, yet the markers shift away from each other. 

Isnt this because the maneuver marker always points to the path of the maneuver as it was when you planned it, whereas the other markers move inside the actual orbit as it is deformed by the burn? Specially in long burns, there is always gonna be a difference between, for instance, actually burning prograde and burning inside the path of a prograde maneuver pre-stablished. I imagine so, since as the burn goes, the conic path of the maneuver stays the same but the actual orbit will change shape while you burn. In my perception the maneuver node remains static.

At least that is the impression had!