Some orbit lines have strange behavior

Bug #1490223 reported by Adriano Steffler on 2015-08-30
14
This bug affects 3 people
Affects Status Importance Assigned to Milestone
Stellarium
High
Guillaume Chereau

Bug Description

When I activated the orbit line for Pluto, it was not drawn correctly and was shown as a broken line. To reproduce this bug, try to see Pluto from Caronte, for example. See attached image for visual information about the issue.

Thank you & best regards!

Changed in stellarium:
status: New → Confirmed
importance: Undecided → High
Changed in stellarium:
assignee: nobody → Guillaume Chereau (guillaume-chereau)
Changed in stellarium:
milestone: none → 0.15.0
Alexander Wolf (alexwolf) wrote :

The problem is in the scale of the whole Solar system such as seen Earth's orbit from the Moon or the orbit of Jupiter from Ganymede.

tags: added: solar-system
removed: lines

I am still not entirely sure, but it seems to me that the root of the problem is that the orbit line should be split into segments of different size depending on the distance to the observer. The way it works now, the line is always split into 360 segments of same size. When the orbit passes very near the observer (as in the Pluto seen from Charon case), a single segment projects into a large angle.

A fix might be to render the orbits using the higher level rendering functions, like `drawSmallCircleArc`.

Alexander Wolf (alexwolf) wrote :

Guillaume, can you update code of orbit's drawing to use drawSmallCircleArc?

P.S. Can it fix bug #978576 also?

I will have a look at it. It's a bit tricky, because even if drawSmallCircle works, it might be too slow when rendering a lot of orbits...

I don't think this is related to bug #978576, but it would be nice to fix both at the same time.

Alexander Wolf (alexwolf) wrote :

Yes, current bug is not related to bug #978576. Can we directly use GPU for rendering orbits?

I don't think we can use GPU directly. Maybe we could approximate the orbits with their Kepler solutions. That might be enough for a single revolution, and would make the computations much faster.

I started a new branch to work on this bug:
https://code.launchpad.net/~guillaume-chereau/stellarium/fix-orbits

I use a new way of rendering orbits, using parametric functions instead of tessellated lines. So far it seems to work, but there are a a few bugs I need to fix before we can consider merging.

Changed in stellarium:
milestone: 0.15.0 → 1.0.0
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers