Atmospheres: old vs new

[chris]

I'm still working on the model of atmospheric scattering. A few comparisons between the old and new atmosphere code:

Earth seascape - old:

Earth seascape - new - notice the ocean fade into the haze toward the horizon:


Sunrise - old:

Sunrise - new:


Whole Earth - old:

Whole Earth - new - note the bluish tint due to Rayleigh scattering:


--Chris

[chris]

One more comparisons . . .

Old:


New:


--Chris

[Kolano]

Nice progress.

For comparison...
The classic Apollo image...


Other earth from space image:


I presume there are a few variables for you to play with in this.

It seems to me that the fade out needs to be more condensed (i.e. pushed further towards the edges of the sphere, having less effect on the center areas).

[Kolano]

I was making my post at the same time as Chris. Surprising how closely the two images I selected line up with his.

[Malenfant]

Looks good, Chris! How does a comparison of other planets look though (ie old mars vs new mars, old venus vs new venus)

[Toti]

Looks great!!

Did you include aerial perspective? In the second image the ground seems to be affected by distance.

Later:
Oops! didn't see the legend accompanying that image.

[rra]

Great work, can't wait to play with it .
One thing I personally didn't like was the new "whole earth" view,
because of Rayleigh scattering the Earth doesn't look so crisp as it used te be.
Is the new rendering an option one can select (and also turn off or select another ?) .

Any idea when we can expect the first beta build
with all of the new eye candy ??

Ren?©

[ElChristou]

Very nice!
Tx for the great work Chris!

[ANDREA]

Chris, the new seascape and sunrise images look absolutely real!
Thank you for this wonderful improvement.
Bye

Andrea

[Christophe]

A very, very nice improvement indeed.

Chris, on the images above you don't seem to have FS AA enabled, any reason for it?

[Malenfant]

Great work, can't wait to play with it .
One thing I personally didn't like was the new "whole earth" view,
because of Rayleigh scattering the Earth doesn't look so crisp as it used te be.

Actually I think one reason the Celestia full earth looks unrealistic is that the scattering on the ocean is too big and the cloud cover not complete enough. I mean, look at how much the clouds cover the Earth on that Apollo pic...

[chris]

Looks good, Chris! How does a comparison of other planets look though (ie old mars vs new mars, old venus vs new venus)

I haven't put as much care into selecting the atmosphere parameters for Mars as for Earth. I need to figure out what's actually going on in Mars's atmosphere. The dust colors the sky reddish and the sunsets have a slight blue color--it's as if the wavelength dependence of scattering is backward version of the Rayleigh 1/wavelength^4. Anyhow, here's an attempt:

Mars - old:


Mars - new:


There are still some tesselation artifacts visible with the new atmosphere code. I plan on taking care of those before release time.

--Chris

[Chuft-Captain]

Chris,
JOOI, what sized Earth texture are you using in these pictures?

There's no chance I'll be able to display this quality with my current graphics, but just interested for the day that I will.

[Malenfant]

I think one big difference with Mars is that there's a lot of very fine-grained particulates in the atmosphere. That's probably going to give you different scattering effects than just molecules of gas or vapour would.

This might be of some use - the second link is more technical:
http://webexhibits.org/causesofcolor/14C.html
http://www.mars-lunar.net/Solar.Energy. ... ities.html

[phoenix]

wow
first the new specular-rendering and now this, incredibly fast and beautiful development chris

[symaski62]

joli !!! image !

chris thanks !

celestia 1.4.2 BETA ou Celestia 1.5.0 BETA ?!?

[Christophe]

Chris, I tried the new code, with the atmosphere settings you posted. Here is what I get:



Turning off clouds or cloud shadows fixes it.

with clouds off:


Also, when going below arround 750m, a gap between the horizon and the atmosphere appears:


And except at sunrise and sunset, the sky appears black.

Code: Select all

Vendor : NVIDIA Corporation
Renderer : GeForce 6600 GT/PCI/SSE2/3DNOW!
Version : 2.0.2 NVIDIA 87.62
Max simultaneous textures: 4
Max texture size: 4096

[granthutchison]

I need to figure out what's actually going on in Mars's atmosphere. The dust colors the sky reddish and the sunsets have a slight blue color--it's as if the wavelength dependence of scattering is backward version of the Rayleigh 1/wavelength^4.

Unfortunately, I think it shows the directional dependence of particulate colours. When you look towards the sun, you're seeing the shadowed side of all the dust particles, so they add little or no colour to the sky. All that's left is Mie and Rayleigh scattering, generating a bluish-white patch around the sun at Martian sunset. Next time you stand next to a (daytime) bonfire, scoot round and look at the sun through the smoke: you'll see brown smoke with a central washed-out blue patch surrounding the sun: the similarity to Martian sunset photos is quite impressive.

I have seen it suggested that the Martian sunset sky is produced by a population of reverse-Rayleigh scatters, as you say. Take particles that are precisely tuned in size to be on the cusp between Rayleigh and Mie scattering. They'll Rayleigh-scatter red, but Mie-scatter blue. Theoretical calculations bear that out: 10-micron particles produce a Rayleigh spectrum mildly weighted to the red end: Craig Bohren's Atmospheric Optics chapter provides a graph (as well as perhaps being a useful reference source for your work, Chris).
Trouble is, what plausible mechanism could produce such a finely tuned population of dust particles in the Martian atmosphere?

Grant

[chris]

Chris, I tried the new code, with the atmosphere settings you posted. Here is what I get:

Turning off clouds or cloud shadows fixes it.

Strange . . . Could you send or post the contents of shaders.log?

Also, when going below arround 750m, a gap between the horizon and the atmosphere appears:

Yes, I know about this. The new atmosphere code has required adjustments to the code that sets up the viewport clipping planes. It should be an easy fix (and soon I will be completely rewriting that code.)

And except at sunrise and sunset, the sky appears black.

It does? One of those images looks like a midday scene with a blue sky.

--Chris

[Christophe]

Strange . . . Could you send or post the contents of shaders.log?

http://epicycle.free.fr/celestia-shaders.log

And except at sunrise and sunset, the sky appears black.

It does? One of those images looks like a midday scene with a blue sky.

You're right, the problem is not linked to the position of the Sun in the sky but to the altitude of the point you're looking at. The sky is always black at the zenith, even at midday. This short video shows the problem.