Celestia Forums

Look at this URL:
cel://Follow/Sol:Jupiter:Io/2003-05-10T ... 36103&lm=0

The time is backwards, but that does not matter. It shows a shadow falling on Io, but in this shadow has two edges - Very unnatural

It sure can be seen on Pentium 4, 1,7 GHz, 256 MB with Nvidia GeForce2 MX 400 64 MB
and also on Notebook Dell Inspiron 8200, Pentium 4, 2,4 GHz, 512 MB with ATI Mobility Radeon 9000, 64 MB.
Both computers running with Windows XP.

Io is in the process of going into eclipse by Jupiter. You're seeing the penumbral effect. My understanding is that that's the way Celestia is designed to draw the penumbra and umbra shadows, although it may be rather unexpected.

Toggle rendering of eclipse shadows off and on, and you'll see what I mean. Also, type "h" followed by "c". You'll see that the sun is starting to go behind Jupiter.

I hope this clarifies things a little.

The edge looks weird because of the limited spatial resolution of eclipse shadows; Jupiter is so much bigger than Io that it's possible to see the effect easily.

It's because of the low resolution of the eclipse shadows... They are generated 256x256. I have made a personal version with 1024x1024 and the shadows are far better !

I am not sure what you folks are seeing, but the attached picture is what I see. Not many edges to complain about here.

Could you folks check out the following script and help me learn whether the shadow in the shot of the moon at the end is in the umbra or the penumbra zone?

http://vernon.is-a-geek.org/jerusalem_00330403_18.cel

Thanks.

vsingleton wrote:I am not sure what you folks are seeing, but the attached picture is what I see. Not many edges to complain about here.

Well, the bug was reported in 2003... Celestia has changed a lot in six years. On graphics cards that support OpenGL 2.0 shaders (i.e. nearly all of them nowadays), Celestia renders perfectly smooth eclipse shadows like the ones that you reported.

--Chris

chris wrote:Well, the bug was reported in 2003...

LOL