Celestia Forums

In case you ever wondered what eclipse shadows would look like on a planet orbiting a close binary pair:



--Chris

presumably it's not supposed to have that sharp split down the middle of it?

Evil Dr Ganymede wrote:presumably it's not supposed to have that sharp split down the middle of it?

Oops. I messed up the image . . . Try refreshing to see an image without that artifact.

--Chris

Better now

Eclipses must be somewhat wacky in such systems - you'd have partial eclipses of either star, total eclipses of either star, or total eclipses of both (or very rarely, partial eclipses of both if the separations and moon size are just right)

Chris,

I suppose Grant can easily explain those sharp dark rims around both relatively light gray shadow disks.

I failed, unfortunately...

Bye Fridger

Evil Dr Ganymede wrote:Better now

Eclipses must be somewhat wacky in such systems - you'd have partial eclipses of either star, total eclipses of either star, or total eclipses of both (or very rarely, partial eclipses of both if the separations and moon size are just right)

In the image above, you can see points on the planet that are experiencing all of those situations. A viewer in the black convex lens shaped region would see a double total eclipse. In the areas northeast and southwest of the points of the lens, it's a double partial eclipse.

--Chris

t00fri wrote:I suppose Grant can easily explain those sharp dark rims around both relatively light gray shadow disks.

isn't the fuzzy outline the penumbra of the shadow?

Evil Dr Ganymede wrote:

t00fri wrote:I suppose Grant can easily explain those sharp dark rims around both relatively light gray shadow disks.

isn't the fuzzy outline the penumbra of the shadow?

Fridger is referring to the sharp edge between the umbra and the penumbra of each shadow.
It's ... um ... non-physical.

Grant

granthutchison wrote:Fridger is referring to the sharp edge between the umbra and the penumbra of each shadow.
It's ... um ... non-physical.

Some work remains to be done on the falloff function in the penumbra. The actual circle-circle intersection calculation is too expensive to place in the pixel shader. I may try using a 1D texture as a lookup table, though I need to sort our how to make that work with a variety of sun/occluder apparent size ratios. Perhaps a 2D texture, indexed by overlap distance and apparent size ratio . . . Hmmm.

--Chris