> You are not understanding at all what an image-pattern does!  It does only
> define a pattern of the image in the unit cube.  The pattern (the checker
> pattern makes this easy to see) together with the normal map is part of the
> the surface normal function, and that is then mapped onto the surface (this
> is what the uv-mapping modifier specifies) using uv-mapping, which is a
> 2d-function, rather than specifying 3d-function.  This is exactly like it
> works for every pattern and every other statement using patterns in POV-Ray.
> In fact, it would not make any sense to just pick arbitrary parts of the
> normal statement and use only them for uv-mapping.
> 
> Just imagine what would happen if what you seem to be expecting was the case
> when using pigments.  The image-map would be uv-mapped and a pigment-map
> would not? - This could never work in a predictable manner at all!
> 

OK, I think I get what you are saying now and that does fit what I am 
observing.  So the image pattern is placed in the unit cube then in this 
case acts as a mask to determine which normals go where.  Then this flat 
image of normals is applied to the box.  I was thinking it applied the 
image pattern to the object then masked off the normals.  I got this 
ideal in my head when I read the last part of 6.7.11.20 Image Pattern 
where it talks about using it to create masks.  I just thought the image 
could be uv mapped onto and object and act like a mask for maps.  So my 
new question is how can I use an image as mask to determine where 
different normals will go put will not scale the patterns I am using as 
normals unless I insert scale x.

Post a reply to this message

In article <eric.medlin-1EE335.13520623032004@news.povray.org> , Eric Medlin
<eri### [at] ipapercom>  wrote:

> OK, I think I get what you are saying now and that does fit what I am
> observing.  So the image pattern is placed in the unit cube then in this
> case acts as a mask to determine which normals go where.  Then this flat
> image of normals is applied to the box.  I was thinking it applied the
> image pattern to the object then masked off the normals.  I got this
> ideal in my head when I read the last part of 6.7.11.20 Image Pattern
> where it talks about using it to create masks.  I just thought the image
> could be uv mapped onto and object and act like a mask for maps.  So my
> new question is how can I use an image as mask to determine where
> different normals will go put will not scale the patterns I am using as
> normals unless I insert scale x.

Hmm, everything will work as expected if you use either uv-mapping for
normals and pigments or you don't use it for either.  Otherwise, you will
need to make adjustments anyway as uv-mapping depends on the object type,
and not all objects' uv-vectors are as trivial as those of a box.

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

> Hmm, everything will work as expected if you use either uv-mapping for
> normals and pigments or you don't use it for either.  Otherwise, you will
> need to make adjustments anyway as uv-mapping depends on the object type,
> and not all objects' uv-vectors are as trivial as those of a box.
> 
I only want to use uv mapping for the mask not the patterns that 
generate the normals.  In my example I had one box on top of another 
box.  I would like to be able to keep the same scale for both patterns 
used to generate the normals for both boxes the same.  But, one box 
would use a mask to determine where the normals would show up at.  This 
technique will be used a more completicated model that I have.  It is in 
a mesh2 object with part of the part of the mesh2 uv mapped and the 
other part not mapped at all.  The part that is uv mapped needs to use a 
mask to determine where a normal map will be used and the part that is 
not uv mapped will use the same normal map.  I need to keep the scale 
the same so it looks the same across the model.  From your statement 
that does not sound possible.

P.S.  Thanks for all the help.  I am heading home from work now and will 
pick this back up tomorrow morning.

Post a reply to this message