Here's the scene I was talking about in the other thread where an value 
of 1 for assumed_gamma was resulting in the image being too bright and 
washed out. Either I have to modify TerraPOV's atmospheric media, or I 
have to modify every other object in my scene.

-Mike

Post a reply to this message

Attachments:
Download 'gh_scene_outdoors_cubemap_gamma_on.png' (1546 KB)

Preview of image 'gh_scene_outdoors_cubemap_gamma_on.png'

SharkD schrieb:
> Here's the scene I was talking about in the other thread where an value 
> of 1 for assumed_gamma was resulting in the image being too bright and 
> washed out. Either I have to modify TerraPOV's atmospheric media, or I 
> have to modify every other object in my scene.

To me this looks like you're using an awful lot of ambient.

Post a reply to this message

clipka wrote:
> To me this looks like you're using an awful lot of ambient.

Wow! That made a big difference! However, the colors still seem sort of 
washed out. I've attached the new render as well as an old one from 
before I lowered the gamma. I prefer the crisp, saturated colors of the 
older image.

- Mike

Post a reply to this message

Attachments:
Download 'gh_scene_outdoors_cubemap.png' (1455 KB) Download 'gh_scene_outdoors_saturated.png' (1507 KB)

Preview of image 'gh_scene_outdoors_cubemap.png'

Preview of image 'gh_scene_outdoors_saturated.png'

This last one looks alright to me.

Thomas

Post a reply to this message

Here's my latest render. The scene uses two media for the sky: one for 
the clouds and one for the haze. Either one of them by itself renders 
pretty fast, but together they cause the render to take a *really* long 
time.

-Mike

Post a reply to this message

Attachments:
Download 'gh_scene_outdoors_cubemap_atmos_gamma_b.png' (533 KB)

Preview of image 'gh_scene_outdoors_cubemap_atmos_gamma_b.png'

SharkD a écrit :
> Here's my latest render. The scene uses two media for the sky: one for 
> the clouds and one for the haze. Either one of them by itself renders 
> pretty fast, but together they cause the render to take a *really* long 
> time.
> 
> -Mike
> 
> ------------------------------------------------------------------------
> 
Are they overlaping or sharing the same container?

I'd try with containers that are distinct, with some gap between them. 
That way, you compute one media, them, somewhat further, you compute the 
other one. The haze could easily use a low samples value, like samples 
3, while the clouds will use a larger samples value, like samples 30.



Alain

Post a reply to this message

Alain wrote:
> Are they overlaping or sharing the same container?
> 
> I'd try with containers that are distinct, with some gap between them. 
> That way, you compute one media, them, somewhat further, you compute the 
> other one. The haze could easily use a low samples value, like samples 
> 3, while the clouds will use a larger samples value, like samples 30.

They each use their own container. The smaller container is inside the 
larger. If it makes the scene render faster, I would like to change them 
so they use the same container. Not sure how to go about doing this 
though. I also don't want to make them use distinct volumes. Their media 
should "add".

-Mike

Post a reply to this message

SharkD a écrit :
> Alain wrote:
>> Are they overlaping or sharing the same container?
>>
>> I'd try with containers that are distinct, with some gap between them. 
>> That way, you compute one media, them, somewhat further, you compute 
>> the other one. The haze could easily use a low samples value, like 
>> samples 3, while the clouds will use a larger samples value, like 
>> samples 30.
> 
> They each use their own container. The smaller container is inside the 
> larger. If it makes the scene render faster, I would like to change them 
> so they use the same container. Not sure how to go about doing this 
> though. I also don't want to make them use distinct volumes. Their media 
> should "add".
> 
> -Mike
This forces both medias to use the same number of samples.
Also, the sampling for the clouds probably continue into the haze media.

To use only one container, you'll need to add another density pattern.

One media with a single density for the haze.
One media with two densities for the clouds. The new pttern can be 
"planar" with a map similar to: (added in the cloud's media block)
density{planar {color_map{ [0 rgb 0][0.02 rgb 1] }}}
To have the clouds go sharply from invisible to full density.

Two media in a container add.
A media with two densities multiply the densities.


Alain

Post a reply to this message

Alain wrote:
> This forces both medias to use the same number of samples.
> Also, the sampling for the clouds probably continue into the haze media.
> 
> To use only one container, you'll need to add another density pattern.
> 
> One media with a single density for the haze.
> One media with two densities for the clouds. The new pttern can be 
> "planar" with a map similar to: (added in the cloud's media block)
> density{planar {color_map{ [0 rgb 0][0.02 rgb 1] }}}
> To have the clouds go sharply from invisible to full density.
> 
> Two media in a container add.
> A media with two densities multiply the densities.
> 
> 
> Alain

Actually I am using concentric spheres. The media attached to the 
smaller sphere uses an onion pattern, so I'll need to figure out how to 
limit it to just the assigned area instead of repeating.

-Mike

Post a reply to this message

Here is the media for the cloud sphere, which is the inner of the two 
volumes. What I need is to be able to "clip" the media to "InnerRadius" 
and "OuterRadius" so that I can then simply copy/move the media to the 
larger atmosphere sphere.

-Mike

//BEGIN

	#local InnerRadius = EarthRadius + CloudMinRad;
	#local OuterRadius = InnerRadius + CloudMaxRad;
	#local BoundRadius = sin(acos(EarthRadius/OuterRadius)) * OuterRadius;
	#local ScaleAmount = Width/2;
//	#local testpattern = function {mod(sqrt(Sqr(X) + Sqr(Y) + Sqr(Z)), 1.0)}
	#local CloudPigment = pigment
	{
		average
		pigment_map
		{
			[2 onion	scale CloudMaxRad/ScaleAmount	phase 
-mod(OuterRadius,CloudMaxRad)/CloudMaxRad/ScaleAmount]
			[2 granite	scale 1024 * 32/ScaleAmount]
		}
	}
	#local CloudMedia = media
	{
		scattering	{1, 0.5 * 1/ScaleAmount / 320}
		method		3
		intervals	1
//		samples		30	//30, 100
		samples		4
		density
		{
			pigment_pattern {CloudPigment}
			density_map {[0.60 rgb 0][0.61 rgb 1]}
		}
	}
	#local CloudSphere = difference
	{
		sphere {0, OuterRadius}
		sphere {0, InnerRadius}
		bounded_by {cylinder {<0,EarthRadius,0,>, <0,OuterRadius,0,>, 
BoundRadius}}
		hollow
		material
		{
			texture {pigment {rgbt 1}}
			interior
			{
				media {CloudMedia}
			}
			scale ScaleAmount
		}
		translate -y * EarthRadius
	}

//END

Post a reply to this message