Visualization Enhancements

[nthallen]

• [X] Get a light blue background. I accomplished this interactively using the mayavi interactive tools. The thing I changed was the 'Scene' background. If we can figure out how to reference the Scene, it should be easy to update. Resolved in 7228794d0ccf5a18e1d0bcd375cafb87d50cc10c
• [x] Try adding a horizon reference: Add a brown sphere 30km below the gondola with 6,371 km radius. Partially Resolved. see following comment.
• [X] Possibly add some sort of directional vectors (N, W) Resolved

[nthallen]

• Direction vectors, when present, should probably be anchored to the current gondola position.

• Simulated horizon could be implemented as a disk of radius 504 km located 40 km below the gondola. It could also be a proportionally smaller disk proportionally closer (e.g. 504 m 40 m below the gondola), but then the disk position would need to be recentered whenever the gondola moves.

  • This works, but objects tend to disappear, particularly when zooming in. I suspect this may happen when the view distance gets too small. Specifically, it makes no sense to zoom past the focus point, but sometimes that is exactly what we want to do. It might make sense to update the focus point when the distance gets too small (maybe < 1000, maybe even further) and instead of setting the focus point to the gondola, simply update the focus by the same distance and direction as the gondola moved.
  • To determine whether this is worth working, we need a test function to move the focal point further away while increasing the distance.

Given Zenith Angle Ze and Azimuth Angle Az, the direction vector from the focal point toward the camera is given by:

• x = sin(Ze)cos(Az)
• y = sin(Ze)sin(Az)
• z = cos(Za) To double the focal distance:
• focus = focus - distance * (x,y,z)
• distance = distance*2

If we code this into an interaction, we could:

• run the simulation to a position where elements are obviously disappearing
• pause the simulation
• double the focal length and observe the effects

If this works, it means we need to allow the focal point to be somewhere other than the gondola position. We can figure out the details after running the test, but it might be possible to simply set a reasonable distance range, and update the focal point if the zoom is too close or too far.

[nthallen]

Clipping is still an issue. We have documented the problem and submitted an issue on the mayavi GitHub. The current workaround is to make the planetary disc really small. The disadvantage of this is that you can fall off the edge!