Another chapter from the tome of POV-Ray Voodoo:

When working with layered textures and pertubed surface normals, only the
topmost layer's normal pertubation is used for reflections and refractions
(although diffuse and highlight components are based on each layer's own normal
pertubation.)

So if for instance you want to model a blurry-reflective sphere (using micro- or
macronormals for the blur effect) partially covered with rust (using normals to
give the rust a rough appearance), be sure to model it with the rust *below*
the reflective texture (as the rust will not rely on reflection), or use a
texture_map to blend the two textures.

Post a reply to this message

"clipka" <nomail@nomail> wrote:
> Another chapter from the tome of POV-Ray Voodoo:
>
> When working with layered textures and pertubed surface normals, only the
> topmost layer's normal pertubation is used for reflections and refractions
> (although diffuse and highlight components are based on each layer's own normal
> pertubation.)
>
> So if for instance you want to model a blurry-reflective sphere (using micro- or
> macronormals for the blur effect) partially covered with rust (using normals to
> give the rust a rough appearance), be sure to model it with the rust *below*
> the reflective texture (as the rust will not rely on reflection), or use a
> texture_map to blend the two textures.

Sorry, I must correct myself - I misread the source code:

Reflection *does* take into account each individual layer's normal pertubations;
only refraction doesn't - so if you want to model a dull rusty *glass* sphere...
>_<.

Post a reply to this message

clipka wrote:
> Reflection *does* take into account each individual layer's normal pertubations;
> only refraction doesn't - so if you want to model a dull rusty *glass* sphere...
>> _<.
> 
> 

Thank $DEITY you noticed that! My next project was a bunch of rusting
glass planetoids :-)

John
-- 
"Eppur si muove" - Galileo Galilei

Post a reply to this message

Another related chapter of Voodoo:

When using radiosity with "normal on", all layered textures in a batch will use
the same normal pertubations for the sake of radiosity computation - but this
may not necessarily be taken from the topmost texture, and may even differ for
adjacent pixels.

Post a reply to this message

Doctor John <joh### [at] homecom> wrote:
> Thank $DEITY you noticed that! My next project was a bunch of rusting
> glass planetoids :-)

Hm... yeah, I guess it's a bit more far-fetched than the rusty steel thing ;)

But I *can* think of earnest use cases for it; for instance, someone might want
to model some "message in a bottle" scene: An old bottle, having grown dull
already (diffuse refraction, we may try antialiasing-driven micronormals here
for instance), sitting partially in the sand somewhere, with a sheet of dry
sand having settled on it (normals here for grainy appearance, governed by a
slope pattern).

In a naive approach we might try to do this using layered textures, with the
sand having a slope-patterned pigment to vary transmittence, sitting on top of
our glass texture with much smaller normals.

Won't work.

We can't even swap the layers around - because the glass is transparent
*everywhere*. If we place it on top to get the normals right for reflection, we
can't hide the sand texture below it anywhere.

So we need to resort to using a slope-patterned texture_map instead,
interpolating between the two textures.

Post a reply to this message

clipka wrote:
> Doctor John <joh### [at] homecom> wrote:
>> Thank $DEITY you noticed that! My next project was a bunch of rusting
>> glass planetoids :-)
> 
> Hm... yeah, I guess it's a bit more far-fetched than the rusty steel thing ;)
> 
> But I *can* think of earnest use cases for it; for instance, someone might want
> to model some "message in a bottle" scene: An old bottle, having grown dull
> already (diffuse refraction, we may try antialiasing-driven micronormals here
> for instance), sitting partially in the sand somewhere, with a sheet of dry
> sand having settled on it (normals here for grainy appearance, governed by a
> slope pattern).
> 

Thinking about it last night I realised that acid-etched or sand-blasted
glass would be among the classes of objected affected.

John
-- 
"Eppur si muove" - Galileo Galilei

Post a reply to this message

Doctor John <joh### [at] homecom> wrote:
> Thinking about it last night I realised that acid-etched or sand-blasted
> glass would be among the classes of objected affected.

For instance, yes - although this alone does not require layered textures. But
imagine such a glass object to be partly decorated with glossy paint after the
sand-blasting, and you're there.

Post a reply to this message

"clipka" <nomail@nomail> schreef in bericht 
news:web.4a60659677801e9a4aed1130@news.povray.org...
> Another chapter from the tome of POV-Ray Voodoo:
>
> When working with layered textures and pertubed surface normals, only the
> topmost layer's normal pertubation is used for reflections and refractions
> (although diffuse and highlight components are based on each layer's own 
> normal
> pertubation.)
>
> So if for instance you want to model a blurry-reflective sphere (using 
> micro- or
> macronormals for the blur effect) partially covered with rust (using 
> normals to
> give the rust a rough appearance), be sure to model it with the rust 
> *below*
> the reflective texture (as the rust will not rely on reflection), or use a
> texture_map to blend the two textures.
>

Thanks indeed. This is something that crops up once in a while in my scenes. 
I shall jot this down in the "Big Book" (i.e. my note book of miscellaneous 
POV-Ray do's and don't's).

Thomas

Post a reply to this message