Images: 2 suns + rings

[chris]

Here are a few images of a Saturn-like world in a binary system. The interesting feature is the pattern of overlapping ring shadows . . . The rings themselves are not lit by multiple light sources yet.





Here's another image showing ring shadows interacting properly with ambient light:



--Chris

[Evil Dr Ganymede]

Huh. I hadn't thought of that - this will take some getting used to .

I notice that the shadows aren't different colours yet, will that be added later on?



This is admittedly an extreme example with a red light and a blue light, but note what the shadows on the rings look like - each shadow is illuminated by the other light source that can still illuminate it, and the only time you get a black shadow is in the part that is behind Saturn and cannot "see" either light source.

The same should apply to the shadow of the rings on the planet itself, which I can't render clearly enough to illustrate here (I'm using POVRay for this).

[chris]

Huh. I hadn't thought of that - this will take some getting used to .

I notice that the shadows aren't different colours yet, will that be added later on?

They are different colors, but it's hard to tell because the color of light from the stars is subtle. The G dwarf has white light, and the M giant has a slightly pinkish light. In the second image, you can see that one of the shadows is a dark red. I have no idea if the contrast would be more dramatic to an observer actually orbiting this world.

If after inspecting the second image, you still don't believe that Celestia is rendering colored shadows, perhaps this image will convince you:



(To produce this one, I hacked Celestia to use pure red for stars with a surface temperature < 5000K and blue for the rest.)

--Chris

[Evil Dr Ganymede]

OK, it's more obvious on the last image there - I believe you now
I guess it is quite subtle without extreme colour differences in the lights though.

[kikinho]

Chris, what I can do to hack the star color? It's because I want to add green and violet stars in my Celestia.
Example: E5V for green star in main sequence and H8VI for a subdwarf violet star.

[Evil Dr Ganymede]

Right now, I think todo that you'd have to take the source code for Celestia, manually change the right part of the code, and compile the altered program.

[kikinho]

And can you say to me what files I have to use and what lines I have to mofify? And how I can compile? Using a special program?

[Evil Dr Ganymede]

And can you say to me what files I have to use and what lines I have to mofify? And how I can compile? Using a special program?

I can't, I dunno if anyone else can. You'd have to go through the code yourself (which is available from somewhere, but I dunno where) and find the relevant parts and know what to change. Basically, I'd say that if you don't know anything about programming (and you sound like you don't), you're better off not even trying to do it til you learn a language because otherwise you won't have a clue what you're doing.

And if you want to learn programming, start learning . (I don't even know what programming language Celestia is written in... I think it's C++ ?)

That said, I don't know if you could make a new star texture that's green or violet and just replace the existing texture with that. The light wouldn't be green or violet, but at least it'd look like how you want it...

[Evil Dr Ganymede]

BTW, chris - why are the rings in your exaggerated image not purple? I'm guessing it's something to do with them being illuminated by the red star from above and the blue star is underneath them?

[chris]

BTW, chris - why are the rings in your exaggerated image not purple? I'm guessing it's something to do with them being illuminated by the red star from above and the blue star is underneath them?

No--it's because the rings themselves aren't rendered with both light sources yet. I'll get to that soon, though.

--Chris

[chris]

More images showing complex shadows being cast on a planet in a binary system:





Note that there are now two shadows being cast on both the planet and the rings.

--Chris

[t00fri]

More images showing complex shadows being cast on a planet in a binary system:

...
Note that there are now two shadows being cast on both the planet and the rings.

--Chris

Chris,

nice progress!

Saturn's ring illumination looks fine as of this morning (CVS, UTC 10:00). Note that your new code has
again the familiar min (....) gcc compile error in line
4061 of render.cpp.

Here are two remaining issues:

1) I had emphasized the new "hyper specular feature"
of the "OpenGL 2.0" Render Path already yesterday in
my listmail to you:

With a mild
SpecularColor [ 0.23 0.256 0.195 ]
SpecularPower 15.0

all other render paths produce just a slight specular
enhancement of Saturns clouds, while now with "OpenGL 2.0" :



This appears highly unrealistic.

2) Your new "bi-luminal" shadow patterns above look
interesting, though I am convinced that the overlap
regime between the two geometrical ring shadows
exhibits significantly unphysical features.




Bye Fridger

[chris]

Chris,

nice progress!

Saturn's ring illumination looks fine as of this morning (CVS, UTC 10:00). Note that your new code has
again the familiar min (....) gcc compile error in line
4061 of render.cpp.

Ugh . . . not again. I'll fix it with the usual typecast . . .

Here are two remaining issues:

1) I had emphasized the new "hyper specular feature"
of the "OpenGL 2.0" Render Path already yesterday in
my listmail to you:

With a mild
SpecularColor [ 0.23 0.256 0.195 ]
SpecularPower 15.0

all other render paths produce just a slight specular
enhancement of Saturns clouds, while now with "OpenGL 2.0" :

Certainly a bug . . . It's likely that the shaders are ignoring the specular color and just using white, resulting in an overbright specular color.

2) Your new "bi-luminal" shadow patterns above look
interesting, though I am convinced that the overlap
regime between the two geometrical ring shadows
exhibits significantly unphysical features.

I think that the shadows are correct, but in order to convince yourself of that, it's best to look as the planet moves relative to its suns in accelerated time. The lighting calculation is quite straightforward:

[code]
Color = Ambient
For each light source
Shadow = 1
For each object occluding the light source
Shadow = Shadow * (1 - blockedFraction)
End
Color = Color + Diffuse * Shadow * LightColor
End
[/quote]

Here, 'Diffuse' is the usual Lambertian factor, the cosine of the angle between the surface normal and the light direction.

--Chris

[/quote]

[Evil Dr Ganymede]

I think the shadows look fairly accurate too. The suns in that image are separated by at least 120 degrees? The two ring shadows would go between the terminators made by each star, and no further.

Fridger could have a point in that where the shadows overlap they're not black (because if they overlap, no light is getting through at all from either star). It doesn't look like there was any ambient light there, but to me they don't look black where they overlap in Fridger's thumbail. (they do look darker where they overlap though)

[granthutchison]

Fridger could have a point in that where the shadows overlap they're not black (because if they overlap, no light is getting through at all from either star).

Where these shadows overlap, light is getting through from both stars ... remember the rings are transparent, and light is scattered through them.

Grant

[Evil Dr Ganymede]

Fridger could have a point in that where the shadows overlap they're not black (because if they overlap, no light is getting through at all from either star).

Where these shadows overlap, light is getting through from both stars ... remember the rings are transparent, and light is scattered through them.

Good point

They're not always transparent though are they? I thought there could be some parts that were thick enough to block out all the light?

[granthutchison]

They're not always transparent though are they? I thought there could be some parts that were thick enough to block out all the light?

Saturn's rings let some light through everywhere, if you include light that is diffracted on its way through thinner sections, and reflected on its way through even thinner areas. So by the time the shadows are projected on Saturn, even the densest shadows are slightly illuminated. Maybe there are planetary rings out there in the Universe that are dense enough to produce a completely black shadow across a substantial width, but you'd have to wonder how long they could last: being physically very thin and composed of discrete particles that interact gravitationally, it seems to me that optically thick rings would spread and become less dense quite quickly. Maybe two stonking great shepherd moons would serve to confine a massive thin ring?

Grant

[chrisr]

wow chris, these new developments are very exciting. When will they be available. I suspect the next pre-release? When will that be?

[Guckytos]

These things look really awesome.
I guess i will have to get a new graphic card to get all the eye candy. Anyway i am looking forward to the next release.

But i have one question: Will this new version also be able to create more than one ring per planet?