I have been working on a scene modeling the bar and glassware cabinet in the
corner of my living room.  When testing the addition of new elements to this
scene I render it at various anti-alias levels with radiosity turned off, and
I'm pretty happy with it.

When I turn radiosity on (using Rad_Settings(Radiosity_IndoorLQ, off, off)), I
get all of these extra highlight spots on the curtains, the rug, and the frames
around the window.  I verified that if I leave the glasses out of the cabinet,
those highlights disappear, so I'm pretty sure it has something to do with the
reflective glass surfaces interacting with radiosity.

The problem is that I'm not sure how to fix this.  The typical finish on the
glassware is something like this:

        finish {
            specular albedo 0.05
            roughness 0.0005
            reflection {
                0.1
                fresnel
            }
            conserve_energy
        }

and the interior:

    interior {
        ior 1.6
        caustics 1.0
        fade_distance 2*Hurricane_glass_base_radius
        fade_power 1001
    }

There is a sun-based light-source outside of the room, but it is positioned such
that it is probably not adding any highlights to the scene.  The lamp on the bar
itself is very low power, but does have an emission value in the finish of the
lampshade.  The main light source is a cluster of 4 bulbs, each modeled as a 2x2
area light behind, above, and to the left of the camera.

Any thoughts on what I am doing wrong that is causing these to appear?  No
matter how bright a light I add to the actual room, I do not see these
highlights on the curtains or the rug.

Thanks,

-- Chris R.

Post a reply to this message

Attachments:
Download 'scene_1.5_2021-09-13.png' (1064 KB)

Preview of image 'scene_1.5_2021-09-13.png'

"Chris R" <car### [at] comcastnet> wrote:
> I have been working on a scene modeling the bar and glassware cabinet in the
> corner of my living room.  When testing the addition of new elements to this
> scene I render it at various anti-alias levels with radiosity turned off, and
> I'm pretty happy with it.
>
> When I turn radiosity on (using Rad_Settings(Radiosity_IndoorLQ, off, off)), I
> get all of these extra highlight spots on the curtains, the rug, and the frames
> around the window.  I verified that if I leave the glasses out of the cabinet,
> those highlights disappear, so I'm pretty sure it has something to do with the
> reflective glass surfaces interacting with radiosity.
>
> The problem is that I'm not sure how to fix this.  The typical finish on the
> glassware is something like this:
>
>         finish {
>             specular albedo 0.05
>             roughness 0.0005
>             reflection {
>                 0.1
>                 fresnel
>             }
>             conserve_energy
>         }
>
> and the interior:
>
>     interior {
>         ior 1.6
>         caustics 1.0
>         fade_distance 2*Hurricane_glass_base_radius
>         fade_power 1001
>     }
>
> There is a sun-based light-source outside of the room, but it is positioned such
> that it is probably not adding any highlights to the scene.  The lamp on the bar
> itself is very low power, but does have an emission value in the finish of the
> lampshade.  The main light source is a cluster of 4 bulbs, each modeled as a 2x2
> area light behind, above, and to the left of the camera.
>
> Any thoughts on what I am doing wrong that is causing these to appear?  No
> matter how bright a light I add to the actual room, I do not see these
> highlights on the curtains or the rug.
>
> Thanks,
>
> -- Chris R.

My apologies, I mixed up the various experiments I had done and was rerunning to
try some other things out and discovered that it doesn't matter if I include the
glasses in the hutch or not.  In both cases I get the extra highlights on the
curtains.  So, it is either something else on the bar, or its something with the
main light source in the room.

-- Chris R.

Post a reply to this message

> I have been working on a scene modeling the bar and glassware cabinet in the
> corner of my living room.  When testing the addition of new elements to this
> scene I render it at various anti-alias levels with radiosity turned off, and
> I'm pretty happy with it.
> 
> When I turn radiosity on (using Rad_Settings(Radiosity_IndoorLQ, off, off)), I
> get all of these extra highlight spots on the curtains, the rug, and the frames
> around the window.  I verified that if I leave the glasses out of the cabinet,
> those highlights disappear, so I'm pretty sure it has something to do with the
> reflective glass surfaces interacting with radiosity.
> 
> The problem is that I'm not sure how to fix this.  The typical finish on the
> glassware is something like this:
> 
>          finish {
>              specular albedo 0.05
>              roughness 0.0005
>              reflection {
>                  0.1
>                  fresnel
>              }
>              conserve_energy
>          }
> 
> and the interior:
> 
>      interior {
>          ior 1.6
>          caustics 1.0
>          fade_distance 2*Hurricane_glass_base_radius
>          fade_power 1001
>      }
> 
> There is a sun-based light-source outside of the room, but it is positioned such
> that it is probably not adding any highlights to the scene.  The lamp on the bar
> itself is very low power, but does have an emission value in the finish of the
> lampshade.  The main light source is a cluster of 4 bulbs, each modeled as a 2x2
> area light behind, above, and to the left of the camera.

Should be at least 5x5, preferably 9x9, with adaptive 0.

> 
> Any thoughts on what I am doing wrong that is causing these to appear?  No
> matter how bright a light I add to the actual room, I do not see these
> highlights on the curtains or the rug.
> 
> Thanks,
> 
> -- Chris R.
> 

I often see similar highlights IRL. So, they look natural.

Whenever you have reflective or refractive surfaces in a radiosity 
scene, they act the same way photons do : Cause highlights and caustics. 
The mechanism causing them is different, but the end result can be similar.

Post a reply to this message

Alain Martel <kua### [at] videotronca> wrote:
>
> > I have been working on a scene modeling the bar and glassware cabinet in the
> > corner of my living room.  When testing the addition of new elements to this
> > scene I render it at various anti-alias levels with radiosity turned off, and
> > I'm pretty happy with it.
> >
> > When I turn radiosity on (using Rad_Settings(Radiosity_IndoorLQ, off, off)), I
> > get all of these extra highlight spots on the curtains, the rug, and the frames
> > around the window.  I verified that if I leave the glasses out of the cabinet,
> > those highlights disappear, so I'm pretty sure it has something to do with the
> > reflective glass surfaces interacting with radiosity.
> >
> > The problem is that I'm not sure how to fix this.  The typical finish on the
> > glassware is something like this:
> >
> >          finish {
> >              specular albedo 0.05
> >              roughness 0.0005
> >              reflection {
> >                  0.1
> >                  fresnel
> >              }
> >              conserve_energy
> >          }
> >
> > and the interior:
> >
> >      interior {
> >          ior 1.6
> >          caustics 1.0
> >          fade_distance 2*Hurricane_glass_base_radius
> >          fade_power 1001
> >      }
> >
> > There is a sun-based light-source outside of the room, but it is positioned such
> > that it is probably not adding any highlights to the scene.  The lamp on the bar
> > itself is very low power, but does have an emission value in the finish of the
> > lampshade.  The main light source is a cluster of 4 bulbs, each modeled as a 2x2
> > area light behind, above, and to the left of the camera.
>
> Should be at least 5x5, preferably 9x9, with adaptive 0.
>
> >
> > Any thoughts on what I am doing wrong that is causing these to appear?  No
> > matter how bright a light I add to the actual room, I do not see these
> > highlights on the curtains or the rug.
> >
> > Thanks,
> >
> > -- Chris R.
> >
>
> I often see similar highlights IRL. So, they look natural.
>
> Whenever you have reflective or refractive surfaces in a radiosity
> scene, they act the same way photons do : Cause highlights and caustics.
> The mechanism causing them is different, but the end result can be similar.

I am currently running a portion of the scene where I used Radiosity_IndoorHQ to
see if better radiosity settings would eliminate the blotches.  It's been
running for 2 hours and is about to start tracing, but I think I am still seeing
the highlights, so that probably isn't the problem either.

One thing I failed to mention earlier is that the indoor light source was
created using LightSys, using its color calculator and fading.  This always
results in very bright light colors to make the fading work.  I am wondering if
that means I should be using a "max_sample" setting in radiosity?

-- Chris R.

Post a reply to this message

"Chris R" <car### [at] comcastnet> wrote:
> Alain Martel <kua### [at] videotronca> wrote:
> >
> > > I have been working on a scene modeling the bar and glassware cabinet in the
> > > corner of my living room.  When testing the addition of new elements to this
> > > scene I render it at various anti-alias levels with radiosity turned off, and
> > > I'm pretty happy with it.
> > >
> > > When I turn radiosity on (using Rad_Settings(Radiosity_IndoorLQ, off, off)), I
> > > get all of these extra highlight spots on the curtains, the rug, and the frames
> > > around the window.  I verified that if I leave the glasses out of the cabinet,
> > > those highlights disappear, so I'm pretty sure it has something to do with the
> > > reflective glass surfaces interacting with radiosity.
> > >
> > > The problem is that I'm not sure how to fix this.  The typical finish on the
> > > glassware is something like this:
> > >
> > >          finish {
> > >              specular albedo 0.05
> > >              roughness 0.0005
> > >              reflection {
> > >                  0.1
> > >                  fresnel
> > >              }
> > >              conserve_energy
> > >          }
> > >
> > > and the interior:
> > >
> > >      interior {
> > >          ior 1.6
> > >          caustics 1.0
> > >          fade_distance 2*Hurricane_glass_base_radius
> > >          fade_power 1001
> > >      }
> > >
> > > There is a sun-based light-source outside of the room, but it is positioned such
> > > that it is probably not adding any highlights to the scene.  The lamp on the bar
> > > itself is very low power, but does have an emission value in the finish of the
> > > lampshade.  The main light source is a cluster of 4 bulbs, each modeled as a 2x2
> > > area light behind, above, and to the left of the camera.
> >
> > Should be at least 5x5, preferably 9x9, with adaptive 0.
> >
> > >
> > > Any thoughts on what I am doing wrong that is causing these to appear?  No
> > > matter how bright a light I add to the actual room, I do not see these
> > > highlights on the curtains or the rug.
> > >
> > > Thanks,
> > >
> > > -- Chris R.
> > >
> >
> > I often see similar highlights IRL. So, they look natural.
> >
> > Whenever you have reflective or refractive surfaces in a radiosity
> > scene, they act the same way photons do : Cause highlights and caustics.
> > The mechanism causing them is different, but the end result can be similar.
>
> I am currently running a portion of the scene where I used Radiosity_IndoorHQ to
> see if better radiosity settings would eliminate the blotches.  It's been
> running for 2 hours and is about to start tracing, but I think I am still seeing
> the highlights, so that probably isn't the problem either.
>
> One thing I failed to mention earlier is that the indoor light source was
> created using LightSys, using its color calculator and fading.  This always
> results in very bright light colors to make the fading work.  I am wondering if
> that means I should be using a "max_sample" setting in radiosity?
>
> -- Chris R.

I'm mostly talking to myself here, but it turns out that setting max_sample to a
non-default value has removed the unwanted lighting artifacts.  I am currently
using 10 for max_sample in the scene and rendered it with the Radiosity_Fast
settings and the artifacts went away.  I'm rerendering the scene now with better
indoor radiosity settings and will repost the image here to show the difference.

Post a reply to this message

"Chris R" <car### [at] comcastnet> wrote:
> "Chris R" <car### [at] comcastnet> wrote:
> > Alain Martel <kua### [at] videotronca> wrote:
> > >
> > > > I have been working on a scene modeling the bar and glassware cabinet in the
> > > > corner of my living room.  When testing the addition of new elements to this
> > > > scene I render it at various anti-alias levels with radiosity turned off, and
> > > > I'm pretty happy with it.
> > > >
> > > > When I turn radiosity on (using Rad_Settings(Radiosity_IndoorLQ, off, off)), I
> > > > get all of these extra highlight spots on the curtains, the rug, and the
frames
> > > > around the window.  I verified that if I leave the glasses out of the cabinet,
> > > > those highlights disappear, so I'm pretty sure it has something to do with the
> > > > reflective glass surfaces interacting with radiosity.
> > > >
> > > > The problem is that I'm not sure how to fix this.  The typical finish on the
> > > > glassware is something like this:
> > > >
> > > >          finish {
> > > >              specular albedo 0.05
> > > >              roughness 0.0005
> > > >              reflection {
> > > >                  0.1
> > > >                  fresnel
> > > >              }
> > > >              conserve_energy
> > > >          }
> > > >
> > > > and the interior:
> > > >
> > > >      interior {
> > > >          ior 1.6
> > > >          caustics 1.0
> > > >          fade_distance 2*Hurricane_glass_base_radius
> > > >          fade_power 1001
> > > >      }
> > > >
> > > > There is a sun-based light-source outside of the room, but it is positioned
such
> > > > that it is probably not adding any highlights to the scene.  The lamp on the
bar
> > > > itself is very low power, but does have an emission value in the finish of the
> > > > lampshade.  The main light source is a cluster of 4 bulbs, each modeled as a
2x2
> > > > area light behind, above, and to the left of the camera.
> > >
> > > Should be at least 5x5, preferably 9x9, with adaptive 0.
> > >
> > > >
> > > > Any thoughts on what I am doing wrong that is causing these to appear?  No
> > > > matter how bright a light I add to the actual room, I do not see these
> > > > highlights on the curtains or the rug.
> > > >
> > > > Thanks,
> > > >
> > > > -- Chris R.
> > > >
> > >
> > > I often see similar highlights IRL. So, they look natural.
> > >
> > > Whenever you have reflective or refractive surfaces in a radiosity
> > > scene, they act the same way photons do : Cause highlights and caustics.
> > > The mechanism causing them is different, but the end result can be similar.
> >
> > I am currently running a portion of the scene where I used Radiosity_IndoorHQ to
> > see if better radiosity settings would eliminate the blotches.  It's been
> > running for 2 hours and is about to start tracing, but I think I am still seeing
> > the highlights, so that probably isn't the problem either.
> >
> > One thing I failed to mention earlier is that the indoor light source was
> > created using LightSys, using its color calculator and fading.  This always
> > results in very bright light colors to make the fading work.  I am wondering if
> > that means I should be using a "max_sample" setting in radiosity?
> >
> > -- Chris R.
>
> I'm mostly talking to myself here, but it turns out that setting max_sample to a
> non-default value has removed the unwanted lighting artifacts.  I am currently
> using 10 for max_sample in the scene and rendered it with the Radiosity_Fast
> settings and the artifacts went away.  I'm rerendering the scene now with better
> indoor radiosity settings and will repost the image here to show the difference.

As promised, here is the scene with max_sample set to 10.  The artifacts are
gone.  I also took the suggestion of using a 5x5 area light for the main indoor
light.

-- Chris R.

Post a reply to this message

Attachments:
Download 'scene.png' (1701 KB)

Preview of image 'scene.png'

"Chris R" <car### [at] comcastnet> wrote:
> As promised, here is the scene with max_sample set to 10.  The artifacts are
> gone.  I also took the suggestion of using a 5x5 area light for the main indoor
> light.

Hi Chris,

It did help.

After reading the docs here
(http://www.povray.org/documentation/3.7.0/r3_4.html#r3_4_4_3_3_8) it would seem
that the max_sample feature controls the intensity of reflected/refracted rays.
It would appear that using it is similar to altering the brightness clamping of
reflected and refracted caustics in Blender Cycles... The bright blotches that
your scene was producing are probably analogous to the so-called "fireflies"
that you get in renderers such as Cycles. When you clamp these values, a scene's
lighting becomes a bit less realistic.

I can't think of any other good solution... You can't just add a no_radiosity
flag to the offending objects, because then the lighting would probably be
*really* unrealistic. And if you use a super-high count in the radiosity block,
the scene would take forever to render... The only other option I can think of
it to try Clipka's UberPOV to make use of the randomized radiosity samples (I
think the keyword to enable it is called no_cache). But even then, your problem
goes from blotches to sharp pixel-sized fireflies...

(For the record, I think most of the initial blotches you were seeing were cast
from the brass objects.)

Sam

Post a reply to this message

> "Chris R" <car### [at] comcastnet> wrote:
>> As promised, here is the scene with max_sample set to 10.  The artifacts are
>> gone.  I also took the suggestion of using a 5x5 area light for the main indoor
>> light.
> 
> Hi Chris,
> 
> It did help.
> 
> After reading the docs here
> (http://www.povray.org/documentation/3.7.0/r3_4.html#r3_4_4_3_3_8) it would seem
> that the max_sample feature controls the intensity of reflected/refracted rays.
> It would appear that using it is similar to altering the brightness clamping of
> reflected and refracted caustics in Blender Cycles... The bright blotches that
> your scene was producing are probably analogous to the so-called "fireflies"
> that you get in renderers such as Cycles. When you clamp these values, a scene's
> lighting becomes a bit less realistic.
> 
> I can't think of any other good solution... You can't just add a no_radiosity
> flag to the offending objects, because then the lighting would probably be
> *really* unrealistic. And if you use a super-high count in the radiosity block,
> the scene would take forever to render... The only other option I can think of
> it to try Clipka's UberPOV to make use of the randomized radiosity samples (I
> think the keyword to enable it is called no_cache). But even then, your problem
> goes from blotches to sharp pixel-sized fireflies...
> 
> (For the record, I think most of the initial blotches you were seeing were cast
> from the brass objects.)
> 
> Sam
> 
An option would be to use the importance mechanism.

Use a large count value. Add a #default directive that include something 
like :
#declare PrettyHighCount = 10000000; // 10 millions
#declare Intended_count = 150;
#default {... radiosity{importance Intended_count/PrettyHighCount}}

Then, in the description of the bright objects, add :
radiosity{importance 1}

In other not quite as bright objects, you can set the importance to 
something else, like 0.1.

That allow a lot of samples to be taken in the vicinity of the bright 
objects, while using a reasonable sampling density for the rest of the 
scene. That greatly reduce the rendering time.
Need version 3.7, work better in version 3.8.

Post a reply to this message