Hi, HarbingerDawn here. I’m a SpaceEngine team volunteer who sometimes makes contributions to SE’s code. Today I’ll be talking about some of the work I’ve done for the upcoming version,


I have made many changes to how SE chooses atmosphere models and atmosphere hue and saturation values for planets, as there was much room for improvement here. Most of the changes were made to increase realism.

  • Random hue and saturation shifting of atmosphere color is more subtle now.
  • Unrealistic “fun” colored atmospheres (like green, red, pink, yellow, etc.) are now much more rare. This makes planets with realistic atmospheric colors more common, while also making the “fun” colors more enjoyable to find. A few additional colors have also been added.
  • Gas giant atmosphere assignment rules are now the same for all gas giants (previously there were different rules for different temperature classes, which is realistic for clouds but not for atmospheres). Gas giants will now generally have blue skies, like all four of Sol’s gas giants do.
  • Terrestrial planet atmospheres are assigned primarily based on how much surface liquid a planet has, thanks to the new planet classification system. Planets with no liquid, or only small amounts of liquid, will generally have skies colored by particles suspended in the air, like Mars does. Planets with large amounts of surface liquid will generally have blue skies, as most suspended particles in the lower atmosphere would be removed by precipitation.
  • Secondarily, terrestrial planet atmospheres are assigned based on atmospheric density (previously there was assignment based on pressure, but with a lot of overlap in model selection between different pressure ranges). Dense atmospheres will use models made to simulate denser atmospheres, while thin atmospheres will use models made to simulate thinner atmospheres, all with appropriate colors. To this end, I created some additional atmosphere models to fill gaps in the existing selection of models. Previously, there were two “clear air” atmosphere models which simulated atmospheres dominated by Rayleigh scattering: one called “Earth” (which simulates Earth’s atmosphere, as the name suggests), and one called “Thick” (which I created for an older version of SE). I created two additional models to continue this pattern, one called “Thin” for relatively thin atmospheres (less than 1/10th of Earth’s atmospheric density), and one called “Ethereal” for extremely thin atmospheres (less than 1/1000th of Earth’s atmospheric density). You can see a comparison of these four models below:

The four “clear air” atmosphere models

Planetary Systems

I have also modified several elements of planetary system generation, again mostly in an attempt to increase realism.

  • Asteroids and comets have more realistic rotation periods now – previously, they often had extremely long rotation periods due to using a formula that was intended for much larger objects.
  • Planets can now sometimes have moderately eccentric orbits.
  • The spacing between planetary orbits is now more varied, see image below:
  • An example of orbital spacing and eccentricity in a planetary system (click for full resolution)

  • Dwarf planets now have a greater mass range, making it possible to find objects the size of Ceres.
  • Asteroids now have more variation in orbital eccentricity.
  • Asteroid orbital inclination is now a function of orbital speed, rather than being random as before. This means that asteroid belts very close to a star will be flat disks, while extremely distant asteroid belts will form spherical shells similar to the Oort Cloud. Asteroid belts at intermediate distances from their host stars will have a shape between that of a disk and a shell, like Sol’s main asteroid belt and Kuiper Belt. See below:

The various shapes asteroid belts can have


I’ve worked on various other items as well:

Comet tails: I’ve added a lower limit to the size a comet tail can have, so that comets – especially sungrazers – don’t have absurdly short tails when not near their periastrons.

Spacecraft: Spacecraft are now much easier to roll during atmospheric flight. I also added another engine effect group tied to the warp drive, so it is now possible to add engine exhaust effects to warp engines/hyperdrives/etc.

Interface: The caption pop-up delay (delay between mousing over an interface element with an embedded caption, like a toolbar icon, and the caption appearing) has been greatly reduced, so it is much easier to mouse over an icon and see its function.

Planetary rings: I made some additions and refinements to the backlit dust rings SpaceEngineer developed for “E-rings” are now rare, and can only be generated by relatively small moons; new “G-rings” were added, which are dust rings in the orbits of dwarf moons – unlike the “E-rings”, these can be either bluish or reddish in color (see below).

Dust rings in the orbits of dwarf moons, similar to Saturn’s G-ring

Gas giants: I have attempted to increase the realism of gas giant appearance by changing the brightness of their atmospheres and clouds. Work in progress, see below:

Old (top) vs. new (bottom) gas giant appearance

That’s all for this update! I will have more to share in the future.