Jim Charter <jrc### [at] msncom> wrote:
> Kirk Andrews wrote:
> > I'm working on a scene of some ancient Greece-style ruins.  It has proved to
> > be very difficult to get everything to look sufficiently weathered and
> > "ruined".  But here is one of the columns as it stands at the moment.  If
> > anyone has experience or suggestions in ruining things in POV, I'd love to
> > hear them.
> >
> > - Kirk
> >
> >
> > ------------------------------------------------------------------------
> >
> I don't have anything brilliant to add nor can I provide a magic bullet.
> But you've made a beautiful graceful column so far with some nice
> texturing.  As others have said, most find that they reach a point of
> diminishing returns when trying to simulate distress by differencing out
> from primitives with primitives, and integrating those effects with
> perturbed normals,... though there seems no reason that it can't be done
> if sufficient patience and ingenuity is applied to the task. So most
> turn to isosurfaces for real displacement effects from procedural
> patterns.  When render times get large the next recourse is
> programmatically generating meshes where again procedural patterns can
> be incorporated to displace the surface.  Such an approach for
> brickshapes is a macro recently published by Bill Pragnell.  This could
> be adapted for cylinders too.   There are also some heightfield macros
> that might also be useful.


Thank you, those are some useful ideas--I had not thought of using
heightfields, and is Bill's macro posted in p.b.s.f.?  I didn't see it
there.


> Your textures are quite beautiful, the next stage would be to make the
> coloring also responsive to the surface so that grime seems to reside in
> the recesses and so on.  Remember that a factor in the weathering of
> building is the role of bird poop and the like.  Recesses get perched
> on, biology takes its course, then rain causes grime to run and leech
> down the surface.


Ahhh, there's the goal (and I hadn't thought about bird poop, either!).  The
trick is getting that grime *just* in the recesses.  Samuel Benge has been
trying to help me figure out how to do that.  If you have any suggestions
on how that could be done in POV I'd love to hear it.

-- Kirk

Post a reply to this message

"Kirk Andrews" <kir### [at] hotmailcom> wrote:
> Thank you, those are some useful ideas--I had not thought of using
> heightfields, and is Bill's macro posted in p.b.s.f.?  I didn't see it
> there.

It's in p.g. but that's a good point, I should probably put it in p.b.s.f.
too.

> Ahhh, there's the goal (and I hadn't thought about bird poop, either!).  The
> trick is getting that grime *just* in the recesses.  Samuel Benge has been
> trying to help me figure out how to do that.  If you have any suggestions
> on how that could be done in POV I'd love to hear it.

If you're using an isosurface or other pigment function-based perturbation,
you should be able to use the same pattern to provide texturing - limit
your grime textures to the same regions of the colour map as are used for
the erosion. The other thing you could try is using the un-weathered base
object to define an object pattern - only areas inside the object could
then be textured, i.e. just the weathered sections. I don't know if these
are practical for your purposes...

By the way, that's a very nice image. I thought the periodic horizontal gaps
were supposed to be the edges of individual blocks making up the column, but
the whole aging effect is subtle but realistic.

Good luck with further experiments!
Bill

Post a reply to this message

Kirk Andrews nous apporta ses lumieres en ce 2007/06/24 07:03:
>> Looks good. Did you use an ambient occlusion shader for the surface?
>>
>> ~Sam
> 
> Thanks!
> 
> No, I just used a couple slope maps to get a bit of the effect.
> 
> Perhaps I'm not quite bright enough, but I didn't quite understand how your
> suggestion in the advanced users forum would work, the one with jittered
> object pigment.  By "jitter", do you mean a slight amount of turbulence and
> rotation/translation each time, averaged together?
> 
> Thanks,
> 
> Kirk
> 
That jitter means some small to medium, relative to the dimentions of the 
object, translation of the object pattern, using many instances all averaged 
together. You need at least 10 averaged patterns for this to work, and usualy 
over 100 to get good results. As you don't use reflection, it can still be 
relatively fast, it become painfuly slow if you have any reflection, 
particularly if you have self reflection.

-- 
Alain
-------------------------------------------------
Wars are not paid for in wartime, the bill comes later.
Benjamin Franklin

Post a reply to this message

Beautiful image.
I am following this thread closely because I have been thinking about the 
same thing since my image <450bcef5@news.povray.org>. I have no solution to 
offer, but the cogs are turning in my brain  ;-)

Thomas

Post a reply to this message

"Kirk Andrews" <kir### [at] hotmailcom> wrote:
> Ahhh, there's the goal (and I hadn't thought about bird poop, either!).  The
> trick is getting that grime *just* in the recesses.  Samuel Benge has been
> trying to help me figure out how to do that.  If you have any suggestions
> on how that could be done in POV I'd love to hear it.
>
> -- Kirk
Hmm, the way I would do it, is with absorbing media applied to a cylinder
with a radius just a little bit smaller than the column's. Its density
might be (emi-)spherical so the vast majority of the grime would be located
near the bottom.

Nice pic, BTW. I really like the sky, it looks like a "simple" procedural
sky, but simple procedural skies are anything but simple to get right, IMO.

Post a reply to this message

Alain wrote:
> Kirk Andrews nous apporta ses lumieres en ce 2007/06/24 07:03:
>> By "jitter", do you mean a slight amount of 
>> turbulence and
>> rotation/translation each time, averaged together?
>>
>> Thanks,
>>
>> Kirk
>>
> That jitter means some small to medium, relative to the dimentions of 
> the object, translation of the object pattern, using many instances all 
> averaged together. You need at least 10 averaged patterns for this to 
> work, and usualy over 100 to get good results. As you don't use 
> reflection, it can still be relatively fast, it become painfuly slow if 
> you have any reflection, particularly if you have self reflection.

And the pattern you derive from this will need to be encapsulated in a 
pigment_pattern{} block, in order to use pigment_maps with it:

pigment{
  pigment_pattern{ my_predeclared_jittered_object_pigment }
  pigment_map{
   [0 some_pigment]
   [1 some_other_pigment]
  }
}

~Sam

Post a reply to this message

Bill Pragnell wrote:

> 
> If you're using an isosurface or other pigment function-based perturbation,
> you should be able to use the same pattern to provide texturing - limit
> your grime textures to the same regions of the colour map as are used for
> the erosion. 

Agreed.  Pretty much the standard approach.


>The other thing you could try is using the un-weathered base
> object to define an object pattern - only areas inside the object could
> then be textured, i.e. just the weathered sections. I don't know if these
> are practical for your purposes...

I have had some success with this approach also, basically I think I 
jittered the object pattern through several minor scale changes.  I used 
that approach on this chalice image.  In the test shot the green shows 
where the grime colour would fall.

Post a reply to this message

Attachments:
Download 'img.0132.sm.jpg' (106 KB) Download 'img.0210.jpg' (40 KB)

Preview of image 'img.0132.sm.jpg'

Preview of image 'img.0210.jpg'

"Kirk Andrews" <kir### [at] hotmailcom> wrote:
> I'm working on a scene of some ancient Greece-style ruins.  It has proved to
> be very difficult to get everything to look sufficiently weathered and
> "ruined".  But here is one of the columns as it stands at the moment.  If
> anyone has experience or suggestions in ruining things in POV, I'd love to
> hear them.
>
> - Kirk

I once did broken columns by clipping a regular column with an
iso-landscape.  It took some tweaking, but it looked fairly good.

ADB

Post a reply to this message

Jim Charter <jrc### [at] msncom> wrote:

> I have had some success with this approach also, basically I think I
> jittered the object pattern through several minor scale changes.  I used
> that approach on this chalice image.  In the test shot the green shows
> where the grime colour would fall.


That's a beautiful chalice Jim! Did I miss the original full size image?

Did you use multiple versions scaled differently for the jittered object
pattern, or one version scaled multiple times?

Mark Slone

Post a reply to this message

Renderdog wrote:

 >
 > That's a beautiful chalice Jim! Did I miss the original full size image?


Thanks Mark.  No, nothing missed.  That test render was as far as I took 
things.  In fact I don't think the render even includes the grunge 
effect, that was going to be for a second cut which would have also had 
a nice fabric for the display surface.  But I lost interest as usual. 
Funny, Jaime and others have put fabric capabilities into our hands in a 
way that it wasn't when I did the model.  Really should finish this one 
off. Actually the other necessary component, lighting for such a model, 
has also been improved by the efforts of others including Jaime, and I 
am learning it on my current project, so maybe I can roll it all 
together and make a good final render of that chalice.

 >
 > Did you use multiple versions scaled differently for the jittered object
 > pattern, or one version scaled multiple times?
 >

That latter I think.

pigment {
   average
   pigment_map {
     #local I=0;#while(I<90)
       [1
         object {
           object {
             k13_texture_
             #local J = rand(S);
             scale <.9+.2*J,1,.9+.2*J>
           }
           pigment { rgb  CHSL2RGB( <70,.12,.29> ) }
           pigment { rgb  CHSL2RGB( <65,.24,.08> ) }
         }
       ]
       [1
         object {
           object {
             k13_texture_
             rotate y*(-10+20*rand(S))
           }
           pigment { rgb  CHSL2RGB( <70,.12,.29> ) }
           pigment { rgb  CHSL2RGB( <65,.24,.08> ) }
         }
       ]
       #local I=I+1;#end
    }
  }

Post a reply to this message