ambient occlusion map ?

Hi

I wanted to make ambient occlusion map of the scene.....
what i did is

1. Perfectly black plastic material for avoid interreflection
2. define glow color to surface.
3. and a surface light source pointing straight up
4. used rpict with -i for irradiance map and set set ambient bounce to 0;

e.g.

void plastic meshmaterial
0
0
5 0.0 0.0 0.0 0 0

void glow irrad
0
0
4 0.4 0.4 0 .4 0

irrad source irradsphere
0
0
4 0 0 1 180.0

# Mesh details ......all are using meshmaterial

when i render then rendered image does not show any scene (i.e. black scene image) ??
What i am missing conceptually ??

My camera position is correct as i get other maps such as depth or normal map correct.

Best Regards, Brajesh Lal

···

________________________________

You will need at least one ambient bounce to calculate the contribution of an ambient "glow" source. Zero ambient bounces means only direct sources like "light" would be traced.

Also, I'm not sure, but the rpict manual says the -i option actually estimates the irradiance using the radiance value and the lambertian reflectance. I've never delved into the code, so I'm not sure how accurate that description is. After testing -ab 1, you might also try assigning a very small reflectance to your black surface.

···

____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

Hi Christopher

Actually i am computing per pixel visibility ( or say each point visibility of the surface) in the up direction. I suppose if we set ambiance bounce more than zero, then it will have inter-reflection effect. and to avoid interreflection contribution i have done surface material perfectly black and bounces zero.

Yes we do need light source, for that i am using glow ( self luminous surface ) and a unity hemispherical light source for a point.

and logically i suppose -completely visible point is shaded as unity, and zero will be treated as occluded. so value of rgb determines the visibility factor.

-i option produces irradiance map.

so basically i render the scene
rpict -x resx, -y resy -i -ab 0 -vf camera.vp scene.oct > scene.hdr

Best Regards, Brajesh Lal

···

________________________________
From: Christopher Rush <[email protected]>
To: Radiance general discussion <[email protected]>
Sent: Friday, March 15, 2013 3:57 PM
Subject: Re: [Radiance-general] ambient occlusion map ?

You will need at least one ambient bounce to calculate the contribution of an ambient “glow” source. Zero ambient bounces means only direct sources like “light” would be traced.

Also, I’m not sure, but the rpict manual says the -i option actually estimates the irradiance using the radiance value and the lambertian reflectance. I’ve never delved into the code, so I’m not sure how accurate that description is. After testing -ab 1, you might also try assigning a very small reflectance to your black surface.

____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

Are you tracing the pixels from a view point looking at the black surface, to show the irradiance caused by the glow source, which is behind your view point contributing onto the black mesh surface? And the view point is a parallel projection?

···

____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

I render the image from camera point of view as i captured the real world scene.

and if i put any distant liht source e.g. sun - then i can see the scene. but can see the scene if we use object as light source...

Best Regards, Brajesh Lal

···

________________________________
From: Christopher Rush <[email protected]>
To: Radiance general discussion <[email protected]>
Sent: Friday, March 15, 2013 4:57 PM
Subject: Re: [Radiance-general] ambient occlusion map ?

Are you tracing the pixels from a view point looking at the black surface, to show the irradiance caused by the glow source, which is behind your view point contributing onto the black mesh surface? And the view point is a parallel projection?
____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

Brajesh,

This is expected behavior. As Chris pointed out glow sources are sampled
using the ambient calculation. This is because glow was intended to model
the sky, a very large area source. It is much more efficient and accurate
to sample the sky using the stochastic ambient calculation compared of the
deterministic direct calculation.

With -ab 0 you get no light in your image from the glow (unless you can see
the glow geometry in your view). Setting -ab 1 when using a glow source is
equivalent to a direct only calculation (ie without any
inter-eflection effects).

Andy

···

On Fri, Mar 15, 2013 at 9:39 AM, Brajesh Lal <[email protected]> wrote:

I render the image from camera point of view as i captured the real world
scene.
and if i put any distant liht source e.g. sun - then i can see the
scene. but can see the scene if we use object as light source...

Best Regards, Brajesh Lal
  ------------------------------
*From:* Christopher Rush <[email protected]>
*To:* Radiance general discussion <[email protected]>
*Sent:* Friday, March 15, 2013 4:57 PM
*Subject:* Re: [Radiance-general] ambient occlusion map ?

Are you tracing the pixels from a view point looking at the black surface,
to show the irradiance caused by the glow source, which is behind your view
point contributing onto the black mesh surface? And the view point is a
parallel projection?
____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

Thanks Andrew,

yes if i give -ab 1 is direct calculation only then it is working.....

Best Regards, Brajesh Lal

···

________________________________
From: Andrew McNeil <[email protected]>
To: Brajesh Lal <[email protected]>; Radiance general discussion <[email protected]>
Sent: Friday, March 15, 2013 5:53 PM
Subject: Re: [Radiance-general] ambient occlusion map ?

Brajesh,

This is expected behavior. As Chris pointed out glow sources are sampled using the ambient calculation. This is because glow was intended to model the sky, a very large area source. It is much more efficient and accurate to sample the sky using the stochastic ambient calculation compared of the deterministic direct calculation.

With -ab 0 you get no light in your image from the glow (unless you can see the glow geometry in your view). Setting -ab 1 when using a glow source is equivalent to a direct only calculation (ie without any inter-eflection effects).

Andy

On Fri, Mar 15, 2013 at 9:39 AM, Brajesh Lal <[email protected]> wrote:

I render the image from camera point of view as i captured the real world scene.

and if i put any distant liht source e.g. sun - then i can see the scene. but can see the scene if we use object as light source...

Best Regards, Brajesh Lal

________________________________
From: Christopher Rush <[email protected]>
To: Radiance general discussion <[email protected]>
Sent: Friday, March 15, 2013 4:57 PM
Subject: Re: [Radiance-general] ambient occlusion map ?

Are you tracing the pixels from a view point looking at the black surface, to show the irradiance caused by the glow source, which is behind your view point contributing onto the black mesh surface? And the view point is a parallel projection?
____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

Hi Mark,

I suspect you are seeing specular samples from low angles on your black frame. If the frame has a roughness below 0.001 or so and non-zero specularity, the Fresnel coefficient will cause the specularity to go to nearly 1 as you approach grazing. This is correct behavior, but in regions near the inside edge, this will cause stray, bright sample values in your ambient calculation. The only way to fix it is to set roughness above 0.002 or set specularity to zero on your surface.

I haven't downloaded your model to confirm the material parameters, so I'm really just guessing.

Cheers,
-Greg

···

Sent from my iPad

On Mar 15, 2013, at 9:53 AM, Andrew McNeil <[email protected]> wrote:

Brajesh,

This is expected behavior. As Chris pointed out glow sources are sampled using the ambient calculation. This is because glow was intended to model the sky, a very large area source. It is much more efficient and accurate to sample the sky using the stochastic ambient calculation compared of the deterministic direct calculation.

With -ab 0 you get no light in your image from the glow (unless you can see the glow geometry in your view). Setting -ab 1 when using a glow source is equivalent to a direct only calculation (ie without any inter-eflection effects).

Andy

On Fri, Mar 15, 2013 at 9:39 AM, Brajesh Lal <[email protected]> wrote:

I render the image from camera point of view as i captured the real world scene.
and if i put any distant liht source e.g. sun - then i can see the scene. but can see the scene if we use object as light source...

Best Regards, Brajesh Lal
From: Christopher Rush <[email protected]>
To: Radiance general discussion <[email protected]>
Sent: Friday, March 15, 2013 4:57 PM
Subject: Re: [Radiance-general] ambient occlusion map ?

Are you tracing the pixels from a view point looking at the black surface, to show the irradiance caused by the glow source, which is behind your view point contributing onto the black mesh surface? And the view point is a parallel projection?
____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

That was my first suspicion as well, but the halo remained even after dropping specularity to zero.

Computers have been doing extremely weird stuff for me these last few days. Is there a big solar flare or something?

Mark

···

On Fri, 15 Mar 2013, Greg Ward wrote:

Hi Mark,

I suspect you are seeing specular samples from low angles on your black frame. If the frame has a roughness below 0.001 or so and non-zero specularity, the Fresnel coefficient will cause the specularity to go to nearly 1 as you approach grazing. This is correct behavior, but in regions near the inside edge, this will cause stray, bright sample values in your ambient calculation. The only way to fix it is to set roughness above 0.002 or set specularity to zero on your surface.

I haven't downloaded your model to confirm the material parameters, so I'm really just guessing.

Cheers,
-Greg

Sent from my iPad

On Mar 15, 2013, at 9:53 AM, Andrew McNeil <[email protected]> wrote:

Brajesh,

This is expected behavior. As Chris pointed out glow sources are sampled using the ambient calculation. This is because glow was intended to model the sky, a very large area source. It is much more efficient and accurate to sample the sky using the stochastic ambient calculation compared of the deterministic direct calculation.

With -ab 0 you get no light in your image from the glow (unless you can see the glow geometry in your view). Setting -ab 1 when using a glow source is equivalent to a direct only calculation (ie without any inter-eflection effects).

Andy

On Fri, Mar 15, 2013 at 9:39 AM, Brajesh Lal <[email protected]> wrote:

I render the image from camera point of view as i captured the real world scene.
and if i put any distant liht source e.g. sun - then i can see the scene. but can see the scene if we use object as light source...

Best Regards, Brajesh Lal
From: Christopher Rush <[email protected]>
To: Radiance general discussion <[email protected]>
Sent: Friday, March 15, 2013 4:57 PM
Subject: Re: [Radiance-general] ambient occlusion map ?

Are you tracing the pixels from a view point looking at the black surface, to show the irradiance caused by the glow source, which is behind your view point contributing onto the black mesh surface? And the view point is a parallel projection?
____________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general