I fixed the spray problem.  I had the particle distribution set way
too high, so all I was getting was a fine mist.
the Isonsurfaces, it turns out, aren't the culprits.  Most of the 36 hr
render time was taken up by the water texture and the glass.
    The glass texture is straight out of glass.inc, I use the F_Glass6
finish and the I_Glass_Exp interior with a Clear pigment

    The water is a difference (I cut out the shapes of sumberged objects
to prevent co-incident surfaces and to keep the water out of places it
wasn't supposed to be.  And, as I didn't want to use the one out of
textures inc, I created my own water texture:

#declare MyWater =
texture {
    pigment { Clear }
    finish {
        reflection { 0.3, 0.6 fresnel }
        ambient 0.05
        diffuse 0.2
        conserve_energy
        }
    }

    Additionally, I use an interior:

interior {
    fade_power 1001
    fade_distance 17.5*ft // ft = 12
    caustics 0.625
    }

    Not surprisingly, the Spray is what takes the largest ammount of
time to render. This may have something to do with the somewhat high
max_trace_level I specified to avoid black spots (max_trace_level 10, I
couldn't figure out what the default is.)

    In keeping with my meticulous attention to detail, the fountain
tower is fastened together in several places with bolts, and there are
glass light covers, such as might be found covering lights in a swimming
pool, in the basin.  I intend to place small gas flames behind the
ornamental grill at the base with light sources to illuminate the water
from undernieth, as the final render will probably be a Night or early
morning scene.

    I also added my water texture to the fountain spray, as I wasn't
happy with the one built into the macro. (I'm currently using the
somewhat outdated, but adequite, Liquid Spray include file by Chris
Colefax)

I get odd results when rendering with photons, though.  It seems to
treat the transparent surfaces as though they were opaque (that is, the
surfaces are still transparent, but the light doesn't seem to get
through.  I get a solid shadow on the ground to the left instead of
having a lighter area for the dome, and a colored shadow for the
crystal.), and nothing I do fixes the problem.  If anyone has an Idea
about what I can do to fix this, I'd appreciate suggestions.

ADB

Post a reply to this message

Attachments:
Download 'fountain6.jpg' (169 KB)

Preview of image 'fountain6.jpg'

Hello.  Long time lurker, saw some of the same issues I have had to deal
with in the past:

"Anthony D. Baye" <ant### [at] sdsmtedu> wrote in message
news:40E99166.75AEBA73@sdsmt.edu...
>     I fixed the spray problem.  I had the particle distribution set way
> too high, so all I was getting was a fine mist.
> the Isonsurfaces, it turns out, aren't the culprits.  Most of the 36 hr
> render time was taken up by the water texture and the glass.

Unfortunately I missed what hardware you were using, even so that seems like
a long time for this scene.  What is the render time without photons?  Have
you tuned the scene for speed (ie added collect off as needed)?

>     The water is a difference (I cut out the shapes of sumberged objects
> to prevent co-incident surfaces and to keep the water out of places it
> wasn't supposed to be.

I am not sure that is required.  Co-incident surfaces are not an issue if
one object is completely within another.  Artifacts only appear where 2
surfaces are occupying the same space.  Intersections between objects and
objects inside of objects do not cause artifacts.

>     Additionally, I use an interior:
>
> interior {
>     fade_power 1001
>     fade_distance 17.5*ft // ft = 12
>     caustics 0.625
>     }

If you are using photons correctly, you should not use caustics.  Caustics
only simulates the same effect photons produce.  I highly reccomend adding
an ior of 1.33 for your water object.  The values for fade_distance and
fade_power are a little unusual, I prefer to use an absorption media to
simulate light attenuation in water, but what ever works.

>
>     Not surprisingly, the Spray is what takes the largest ammount of
> time to render. This may have something to do with the somewhat high
> max_trace_level I specified to avoid black spots (max_trace_level 10, I
> couldn't figure out what the default is.)

If the texture for the spray is partially transparent, it will take the
scene a lot longer.  I think the default max_trace_level is 5 or so.

> I get odd results when rendering with photons, though.  It seems to
> treat the transparent surfaces as though they were opaque (that is, the
> surfaces are still transparent, but the light doesn't seem to get
> through.  I get a solid shadow on the ground to the left instead of
> having a lighter area for the dome, and a colored shadow for the
> crystal.), and nothing I do fixes the problem.  If anyone has an Idea
> about what I can do to fix this, I'd appreciate suggestions.

Inside of transparent or reflective objects, you need to set the object
photon block, something like:
object{
    myobject
photons{
    target 1.0
    reflection on
    refraction on
    collect off // prevents caustics from being calculated onto the surface
of the glass
        }
}

This causes povray to calculate the photons for that object.  By default,
all objects in a given scene behave as though they will be lit by photons.
You can significantly speed rendering by putting collect off in objects that
will not be struck by photons.

// doesn't get hit by photons, speeds rendering by not calculating photons
for this object
object{
    myobject
photons{ collect off }
}



HTH,
Reactor

Post a reply to this message

On Mon, 05 Jul 2004 11:35:34 -0600, Anthony D. Baye  

<ant### [at] sdsmtedu> wrote:
>
> I get odd results when rendering with photons, though.  It seems to
> treat the transparent surfaces as though they were opaque (that is, th
e
> surfaces are still transparent, but the light doesn't seem to get
> through.  I get a solid shadow on the ground to the left instead of
> having a lighter area for the dome, and a colored shadow for the
> crystal.), and nothing I do fixes the problem.  If anyone has an Idea
> about what I can do to fix this, I'd appreciate suggestions.

Do you have photons { pass_through } on the dome and the crystal?

Reactor recommended using an ior for the water.  In fact, i don't believ
e  

you'll get any results at all without an ior.  Fresnel reflection,  

refraction, and refractive caustics are dependent on the material's inde
x  

of refraction, and POV-Ray will not calculate these effects without one.


-- 

Using Opera's revolutionary e-mail client: http://www.opera.com/m2/

Post a reply to this message

I simply forgot to type in the ior, it is 1.33.  I didn't copy the block
directly.
I'll try collect off.

apologies and thanks,
ADB

Richard Callwood III wrote:

> On Mon, 05 Jul 2004 11:35:34 -0600, Anthony D. Baye
> <ant### [at] sdsmtedu> wrote:
> >
> > I get odd results when rendering with photons, though.  It seems to
> > treat the transparent surfaces as though they were opaque (that is, the
> > surfaces are still transparent, but the light doesn't seem to get
> > through.  I get a solid shadow on the ground to the left instead of
> > having a lighter area for the dome, and a colored shadow for the
> > crystal.), and nothing I do fixes the problem.  If anyone has an Idea
> > about what I can do to fix this, I'd appreciate suggestions.
>
> Do you have photons { pass_through } on the dome and the crystal?
>
> Reactor recommended using an ior for the water.  In fact, i don't believe
> you'll get any results at all without an ior.  Fresnel reflection,
> refraction, and refractive caustics are dependent on the material's index
> of refraction, and POV-Ray will not calculate these effects without one.
>
> --
> Using Opera's revolutionary e-mail client: http://www.opera.com/m2/

Post a reply to this message

One other thing:  I may not have needed to be as complex as I was with regards
to the water object, but with the lighting scheme I intend to use, some shape
adjustment was necessary.  For instance; there are fifteen ports around the
inside of the basin that are covered with glass caps, without clipping the
water would extend through the caps into the flame cavities behind the grates
at the base.  This may seem a little over-zealous but I was concerned with
unexpected affects this might have on the lighting.

I also didn't want the water object to extend through the tower itself in case
this presented a problem with any particle system I might employ in the
future.

ADB

"Anthony D. Baye" wrote:

> I simply forgot to type in the ior, it is 1.33.  I didn't copy the block
> directly.
> I'll try collect off.
>
> apologies and thanks,
> ADB
>
> Richard Callwood III wrote:
>
> > On Mon, 05 Jul 2004 11:35:34 -0600, Anthony D. Baye
> > <ant### [at] sdsmtedu> wrote:
> > >
> > > I get odd results when rendering with photons, though.  It seems to
> > > treat the transparent surfaces as though they were opaque (that is, the
> > > surfaces are still transparent, but the light doesn't seem to get
> > > through.  I get a solid shadow on the ground to the left instead of
> > > having a lighter area for the dome, and a colored shadow for the
> > > crystal.), and nothing I do fixes the problem.  If anyone has an Idea
> > > about what I can do to fix this, I'd appreciate suggestions.
> >
> > Do you have photons { pass_through } on the dome and the crystal?
> >
> > Reactor recommended using an ior for the water.  In fact, i don't believe
> > you'll get any results at all without an ior.  Fresnel reflection,
> > refraction, and refractive caustics are dependent on the material's index
> > of refraction, and POV-Ray will not calculate these effects without one.
> >
> > --
> > Using Opera's revolutionary e-mail client: http://www.opera.com/m2/

Post a reply to this message