Bug or not bug?

[ElChristou]

This is front today build, default package. This effect in the atmosphere is normal?

mars.jpg

[Cham]

This isn't new. I always had this effect since the first implementation of the new atmosphere (the black or transparent "layer" on the dark side).

[ElChristou]

So, not a bug? then is the effect correct? (I find it odd but perhaps it is correct?)

[Cham]

I can't tell if it's a bug. It's too obvious to be missed since all that time, so I guess it isn't a bug. Chris ?

[Hungry4info]

I get the exact same thing, I don't think it's a bug.

[BobHegwood]

I think that this might be dependent upon your particular graphics drivers.
On my machine I do NOT see a clearly-defined band like the one shown in the picture above.

Click on my image here to have a high-resolution look.

EDIT: Also, my Mars definition (in solarsys.ssc) may be different from the norm since I modified it according to atmospheric data garnered from the web. See code below.

Code: Select all

   Atmosphere
   {
   Height         100
   Lower         [ 0.8 0.6 0.6 ]
   Upper         [ 0.7 0.3 0.3 ]
   Sky            [ 0.83 0.75 0.65 ]
   Sunset         [ 0.7 0.7 0.8 ]   # Slightly bluish sunset, as seen in true color pictures from Pathfinder.
   Mie            0.0024
   MieAsymmetry      -0.15
   Rayleigh         [ 0.0010 0.0006 0.0003 ]
   Absorption      [ 0 0 0 ]
   MieScaleHeight      20
   CloudHeight      80            # Height data from http://www.space.com/scienceastronomy/060828_mars_clouds.html.
   CloudSpeed      32
   CloudMap         "MarsClouds.png"
   }

[selden]

Too often graphics bugs can be seen from only a limited number of viewpoints. Here's a Cel:// url corresponding to the picture below. To see this bug, Celestia also must be using the OpenGL 2.0 render path. This image was created using Celestia r4382 built from svn today.

cel://Follow/Sol:Mars/2008-07-20T22:12: ... 2048&ver=2

[BobHegwood]

Too often graphics bugs can be seen from only a limited number of viewpoints. Here's a Cel:// url corresponding to the picture below. To see this bug, Celestia also must be using the OpenGL 2.0 render path. This image was created using Celestia r4382 built from svn today.

cel://Follow/Sol:Mars/2008-07-20T22:12: ... 2048&ver=2

Got it, Selden. Never saw it before, but - as you stated - it only occurs in OpenGL 2.0 mode.
That's why I never saw it I guess. I never use that mode anyway, because of various other rendering problems.
Thanks, Bob

[ElChristou]

After many days I'm still not sure if this is a bug, a known feature that will be corrected later or if there is nothing to do at his level...

[chris]

After many days I'm still not sure if this is a bug, a known feature that will be corrected later or if there is nothing to do at his level...

The short answer is that it's something that will be fixed in a future version of the atmospheric scattering shader.

The long answer is that the scattering atmosphere shader works by calculating a ray from through the atmosphere at every pixel. The amount of scattered light is integrated along this ray by summing the scattering at several points along the ray. Because of performance concerns, a very limited number of samples are used. Artifacts such as the dark band can result. We've done some work on developing alternate approaches to calculating light from atmospheric scattering, but there's no replacement shader ready yet.

--Chris

[ElChristou]

Tx Chris. No need to report this as a bug, right?

[chris]

Tx Chris. No need to report this as a bug, right?

No. There are already four separate bugs filed for atmospheres.

--Chris