This scene took quite a long time to assemble and tweak--not only because of the
complexity and detail, but because it raised a host of issues and interesting
problems (especially in shapes.inc) that needed to be understood and dealt with.
Whenever I ran into some unexpected POV behavior (or my own POV ignorance!), I
took time to pursue the cause with lots of experimentation. As always, a great
learning experience.

The cave is a HFCylinder macro, very hi-resolution, with inside_vector added
(inside shapes.inc) for CSG.

No radiosity for this scene (yet); rad and photons together have some issues
that I haven't solved.

Comments, criticisms, suggestions all welcome.

For those intrepid souls who want to see some of the changes I made to
shapes.inc., take a look here:

http://news.povray.org/povray.advanced-users/thread/%3Cweb.48e0b5a74b96f64378dcad930%40news.povray.org%3E/

Ken W.

Post a reply to this message

Attachments:
Download 'ancient_cave.jpg' (339 KB)

Preview of image 'ancient_cave.jpg'

"Kenneth" <kdw### [at] earthlinknet> wrote:
> This scene took quite a long time to assemble and tweak--not only because of the
> complexity and detail, but because it raised a host of issues and interesting
> problems (especially in shapes.inc) that needed to be understood and dealt with.
> Whenever I ran into some unexpected POV behavior (or my own POV ignorance!), I
> took time to pursue the cause with lots of experimentation. As always, a great
> learning experience.
>
> The cave is a HFCylinder macro, very hi-resolution, with inside_vector added
> (inside shapes.inc) for CSG.
>
> No radiosity for this scene (yet); rad and photons together have some issues
> that I haven't solved.
>
> Comments, criticisms, suggestions all welcome.
>
> For those intrepid souls who want to see some of the changes I made to
> shapes.inc., take a look here:
>
>
http://news.povray.org/povray.advanced-users/thread/%3Cweb.48e0b5a74b96f64378dcad930%40news.povray.org%3E/
>
> Ken W.

The picture is very suggestive.
I like the presence of POV logo.

Post a reply to this message

Kenneth wrote:

> The cave is a HFCylinder macro, very hi-resolution, with inside_vector added
> (inside shapes.inc) for CSG.

The cave is very impressive. What did you use for the texture?

> No radiosity for this scene (yet); rad and photons together have some issues
> that I haven't solved.

Even without radiosity, the lighting might look more convincing if
your light sources used fade_power 2 (or a reduced fade_distance
in case they already do).

Post a reply to this message

"Kenneth" <kdw### [at] earthlinknet> wrote:
> This scene took quite a long time to assemble and tweak...
..
..
..
> No radiosity for this scene (yet); rad and photons together have some issues
> that I haven't solved.
>

I, too, have had a large amount of problems using radiosity and photons at the
same time.  As a result, I typically render an image in 3 or more passes - a
radiosity pass using save_file, a seperate photon pass also saving the results,
then a final pass when the radiosity and photon data is loaded.  Often, when I
do the pre-final passes, I'll usually use a plain white pigment, unless it will
affect the outcome significantly (for instance, if the gray_threshold is very
low while using radiosity, keeping the colors might be really important).

A possible issue with this work around is that certain scenes may require
interaction between radiosity and photons, such as an indirectly lit corner
under water but near to the surface.  Experimentation with such scenes has lead
to unexpected results, and I've not yet been able to figure out why.

-Reactor

Post a reply to this message

"Carlo C." <nomail@nomail> wrote:


> The picture is very suggestive.
> I like the presence of POV logo.

Thank you! The logo is actually a height_field, made from a sort of 'depth
render' of the actual logo 'object' in logo.inc. I originally wanted some tool
'chisel marks' on it--as if ancient sculptors had made it--so I set it up as
part procedural HF/part crackle function. Hard to describe; but the result
looked nice. In the end, though, it just looked too 'busy' for the scene. So as
it turned out, I could have used the much-simpler POV logo object itself!

Ken W.

Post a reply to this message

Christian Froeschlin <chr### [at] chrfrde> wrote:

>
> The cave is very impressive. What did you use for the texture?

An averaged pigment_map, of two turbulated cells patterns, one larger-scaled
than the other. Rather high omega for the turbulence, too. The basic colors are
sort of pastel red, green and blue, plus gray--but the cells patterns create all
sorts of varying colors from those. Plus high phong. (The different colors are
much more discernable in white light; the yellow/orange cave lighting mutes
them down.)

There *was* an extra texture overlay, of water streaks on the cave walls (a
cylindrical function-image, with some transparency)--but because of the
complexity of the cave (a MERGE of cave, logo and the logo's crystal, which is
the photon target), that single extra texture was causing POV to actually
crash, when using photons! I really never found the root cause of the problem,
so I just eliminated the streaks. (I'm sure I could re-work the scene to
separate out some of that merge, which might solve it.)
>
> > No radiosity for this scene (yet); rad and photons together have some issues
> > that I haven't solved.
>
> Even without radiosity, the lighting might look more convincing if
> your light sources used fade_power 2 (or a reduced fade_distance
> in case they already do).

Yes, I agree. It was a real chore, trying to balance *all* the elements in the
scene. (I must have done 200+ test renders--150 of which I still have!) The
lights do have some fade, but need more. As do the photon effects, IMO.

KW

Post a reply to this message

"Reactor" <rea### [at] hotmailcom> wrote:
> "Kenneth" <kdw### [at] earthlinknet> wrote:

> > No radiosity for this scene (yet); rad and photons together have some issues
> > that I haven't solved.
> >
>
> I, too, have had a large amount of problems using radiosity and photons at the
> same time.  As a result, I typically render an image in 3 or more passes - a
> radiosity pass using save_file, a seperate photon pass also saving the results,
> then a final pass when the radiosity and photon data is loaded.

I actually did try some experiments with that idea--out of desperation-- and it
looked very promising. (It was the *only* way I could figure out how to
eliminate the problem.) And I did discover some interesting things along the
way, about how they interact. I didn't pursue it far enough, though. :-[ Yet
another *tangent* to pursue; after all that shapes.inc stuff, I figured I'd
better finish the scene first. I *will* get back to it. Good to know that
others are thinking along the same lines.

KW

Post a reply to this message

Here's one of my test renders.  No photons or radiosity, and no AA, but it does
show the water streaks on the walls, and the 'chisel marks' on the logo.
(They're actually  more visible on the surrounding cutaway 'box' shape--not
part of the logo HF, but a separate isosurface made from the f_rounded_box
function, plus f_crackle.) I tried to place the logo HF so that it looks like
it's blending into the isosurface.

But all too 'busy', IMO.

KW

Post a reply to this message

Attachments:
Download 'cave_test_122.jpg' (315 KB)

Preview of image 'cave_test_122.jpg'

You're on a very good way...

K. (France)


web.48f20165afdf087878dcad930@news.povray.org...
> This scene took quite a long time to assemble and tweak--not only because 
> of the
> complexity and detail, but because it raised a host of issues and 
> interesting
> problems (especially in shapes.inc) that needed to be understood and dealt 
> with.
> Whenever I ran into some unexpected POV behavior (or my own POV 
> ignorance!), I
> took time to pursue the cause with lots of experimentation. As always, a 
> great
> learning experience.
>
> The cave is a HFCylinder macro, very hi-resolution, with inside_vector 
> added
> (inside shapes.inc) for CSG.
>
> No radiosity for this scene (yet); rad and photons together have some 
> issues
> that I haven't solved.
>
> Comments, criticisms, suggestions all welcome.
>
> For those intrepid souls who want to see some of the changes I made to
> shapes.inc., take a look here:
>
>
http://news.povray.org/povray.advanced-users/thread/%3Cweb.48e0b5a74b96f64378dcad930%40news.povray.org%3E/
>
> Ken W.
>

Post a reply to this message