Constellations don’t work

I have received a lot of positive feedback for my recent map of the interstellar neighborhood. A lot of people gave me some great suggestions for how to improve the map and I’m certainly motivated to work some more on it.

One thing a few people suggested was to implement constellations. Well, I knew it doesn’t work as good as people think because I have seen this feature already in a similar tool called HippLiner. To make my point I have recoded my map to show only the stars of the Leo constellation. Note that those stars are much further out and the scale has been increased so each grid is now 20 light years. Check it out here.

So what do we learn from it? (More after the Jump):

Well, as you can see, the constellation really works only when you look from the perspective Solar System. As soon as you move just a few light years from the Solar System, the constellation falls apart into a jagged mess. The stars that make up the constellation are hundreds of light years apart. In fact, even I fell for it. I thought it would be sufficient just put in the Leo stars in my old map. It turned out that because my old map was using orthographic projection, the constellation was messed up even if you centered on the solar system and pointed in the direction of Leo. So I had to go an extra mile and implement true perspective projection to make it show up properly. It’s THAT hard to reproduce the effect.

That’s why I really don’t like constellations. They are a very poor way to organize stars. They come from a time when we did not understand stars as objects in 3D space. They have survived so far because most of astronomy is still done as it was done thousands of years ago: through telescopic observation. In fact, you will find lots of strange artifacts of this method in astronomy today.

• Star positions are encoded by where they appear on the sky, not by where they actually are in 3D space.
• Because of that, the coordinate system is aligned to the ecliptic, which is the plane the planets of the Solar System orbit in. Well, that wouldn’t be as bad but it was actually used because it is also the apparent path that the Sun traces out in the sky during the year.
• The movement of stars (proper motion) refers to the movement of stars relative to the sky. It is measured in mas/yr, so it is highly dependent on the distance of the star to the Solar System. I have yet to find a source which also gives the actual velocity of stars.

Of course, this is also because measuring the position of a star on the sky is easier than measuring the distance of that star from earth (or even its velocity).

So there you have it. Implementing the true perspective view was also a pain in the ass, by the way. It adds another variable (FOV) so controlling the view is more difficult. The math is more complex and it requires you to implement a clipping plane. Also, distances are more difficult to compare. On the other hand, it is much easier to understand the perspective.