"jfmiller" <jfm### [at] hotmailcom> wrote in message
news:3dee7323@news.povray.org...
> Boy, I wish I could help you out here.  I have been trying for a different
> effect but having simular problems.  the difficultie is how to make
> different cells of the crackle pattern different valuse, but to have a
> consistent value throughout the cell.  My ultimant conclusion was that to
do
> this one whould have to hand write the crackle function and abandon the
POV
> version.  If you can find a more elegant solution please let me know.

Yeah, a different color within each of the interiors of the cells is
probably not possible since they are all of the same index of that pattern.

Speaking of cells, the cells pattern would make a good stained glass by
itself, just add turbulence. But then you still can't gather color
information into each segment as though an underlying image were used for
the color placements. Also, not really a way to get the leaded borders added
in I guess.

Post a reply to this message

"hughes, b." <omn### [at] charternet> schrieb im Newsbeitrag
news:3dee8c98@news.povray.org...
> "jfmiller" <jfm### [at] hotmailcom> wrote in message
> news:3dee7323@news.povray.org...
> > Boy, I wish I could help you out here.  I have been trying for a
different
> > effect but having simular problems.  the difficultie is how to make
> > different cells of the crackle pattern different valuse, but to have a
> > consistent value throughout the cell.  My ultimant conclusion was that
to
> do
> > this one whould have to hand write the crackle function and abandon the
> POV
> > version.  If you can find a more elegant solution please let me know.
>
> Yeah, a different color within each of the interiors of the cells is
> probably not possible since they are all of the same index of that
pattern.

Isn't this, what the "solid"-keyword is meant for? "Causes the same value to
be generated for every point within a specific cell." This way you get one
index for one cell - but you still have to find out a way to map the wanted
colors to the color-indices. Perhaps some "eval_pigment" in a loop? Evaluate
the index of the cell, evaluate the color for this cell, safe that color in
an array at the specified index. Use that array to create the color_map for
the cracle pattern. Might work or not :-)

Regards,

Marc-Hendrik

Post a reply to this message

"Marc-Hendrik Bremer" <Mar### [at] t-onlinede> wrote in message
news:3def0a88@news.povray.org...
>
> "hughes, b." <omn### [at] charternet> schrieb im Newsbeitrag
> news:3dee8c98@news.povray.org...
> >
> > Yeah, a different color within each of the interiors of the cells is
> > probably not possible since they are all of the same index of that
> pattern.
>
> Isn't this, what the "solid"-keyword is meant for? "Causes the same value
to
> be generated for every point within a specific cell." This way you get one
> index for one cell - but you still have to find out a way to map the
wanted
> colors to the color-indices. Perhaps some "eval_pigment" in a loop?
Evaluate
> the index of the cell, evaluate the color for this cell, safe that color
in
> an array at the specified index. Use that array to create the color_map
for
> the cracle pattern. Might work or not :-)

Oops, yes, right you are. And the lead framing can be added with the regular
crackle in a texture layer. I still don't see a way you could ever get the
values into the right positions. That would require some kind of
paint-by-number method plus placement. Crackle randomly positions the cells
so you would only get random colors. You need a paint tool filler sort of
thing to do what is needed for this. Unless there's something I'm
overlooking.

Post a reply to this message

IME, crackle will produce a passable stained glass-ish effect and MIGHT 
be good for scenes where you're not looking right at the window, but the 
only way to get a really authentic-looking stained glass window is to 
use a photo as an image map, or make your own in an image editor or 
something similar.

The reason is that most stained glass windows aren't truly broken up 
into random cells; instead, they use thick leading to create heavy 
outlines, like a picture in a coloring book, and then fills in those 
outlines with colored glass in a way that may OR MAY NOT be random. 
There's no simple way to do this algorithmically, and in order to even 
try, you would need to create a texture with some kind of edge-finding 
ability, to create the heavy outlines.

(If you wanted to cheat halfway, I suppose you could create an image map 
with just the outlines, and then use a crackle or turbulated cells 
pattern to fill in the colored areas.)

As long as I'm on this subject, I should point out that a lot of stained 
glass isn't crystal clear or consistently colored. A good deal of 
stained glass is milky-looking, or has milky-looking swirls in it; some 
is clear but too dark to really see anything through, and small 
variations in color, color intensity, and transparency/milkiness are 
common. If you don't already have a pretty good idea of what stained 
glass is really like, it would be wise to hunt down some photos to study.

-Xplo

Post a reply to this message

"hughes, b." <omn### [at] charternet> schrieb im Newsbeitrag
news:3def7958@news.povray.org...
> Oops, yes, right you are. And the lead framing can be added with the
regular
> crackle in a texture layer. I still don't see a way you could ever get the
> values into the right positions. That would require some kind of
> paint-by-number method plus placement. Crackle randomly positions the
cells
> so you would only get random colors. You need a paint tool filler sort of
> thing to do what is needed for this. Unless there's something I'm
> overlooking.
>

I don't know, if it is possible. It would be, if every cell gets an unique
number. The docs don't tell anything about the range the solid-values are
in. Probably something between 0 and 255 - but does any cell in the rage of
<0,0,0>, <1,1,1> gets it's own number? Would be nice, but somehow I doubt
it. Would anybody want to take a look at the source code?

Regards,
Marc-Hendrik

Post a reply to this message

"Marc-Hendrik Bremer" <Mar### [at] t-onlinede> wrote in message
news:3def7ef8@news.povray.org...
>
> I don't know, if it is possible. It would be, if every cell gets an unique
> number. The docs don't tell anything about the range the solid-values are
> in. Probably something between 0 and 255 - but does any cell in the rage
of
> <0,0,0>, <1,1,1> gets it's own number? Would be nice, but somehow I doubt
> it. Would anybody want to take a look at the source code?

The source code shows a noise function (Perlin?) is used, to get each of a
possible 256 values I believe, but I get lost in all that programming.

Post a reply to this message

> As long as I'm on this subject, I should point out that a lot of stained
> glass isn't crystal clear or consistently colored. A good deal of
> stained glass is milky-looking, or has milky-looking swirls in it; some
> is clear but too dark to really see anything through, and small
> variations in color, color intensity, and transparency/milkiness are
> common. If you don't already have a pretty good idea of what stained
> glass is really like, it would be wise to hunt down some photos to study.
>
> -Xplo

Thanks, Xplo.  This was very informative, and I'll do some more research.
Apparently my uneducated eye hasn't noticed some of the things you've
mentioned.

- How

Post a reply to this message

How Camp wrote:
> However, I'm not sure if this is really possible, or if it's the easiest way
> to create this stained-glass effect.  Can one come up with a method to
> 'color in' a piece of the crackle pattern?
> 
> My latest sad and sorry attempt is attached.  I need to somehow know what
> the color is behind each of these pieces.  The center color, average color,
> whatever.  Something that allows me to transform each crackle 'piece' into a
> solid color.
> 
Well, here's my version - the idea is based on some old pigment swirling 
code of mine.

I'll post the code in p.b.s-f shortly.

Bye for now,
	Mike Andrews.

Post a reply to this message

Attachments:
Download 'cr_solid_test.jpg' (68 KB)

Preview of image 'cr_solid_test.jpg'

"Michael Andrews" <m.c### [at] readingacuk> wrote in message
news:3df0d6eb@news.povray.org...

> Well, here's my version - the idea is based on some old pigment swirling
> code of mine.

Mike, I failed to reply in a timely manner, however your code is exactly
what I was looking for.  Thanks, I appreciate it.  I've learned some new
techniques.  As for my original quest for stained glass, this thread has
changed my goals, I think.  I'm trying to find a method that yields more
realistic results.

Your code has produced some fun (albiet amateur-ish) results for me, though!
It actually solved a totally different problem I was having with textures.
Thanks for that, too.  :)

- How

Post a reply to this message

"How Camp" <kro### [at] hotmailcom> wrote in message
news:3df8f54e@news.povray.org...
>
> "Michael Andrews" <m.c### [at] readingacuk> wrote in message
> news:3df0d6eb@news.povray.org...
>
> > Well, here's my version - the idea is based on some old pigment swirling
> > code of mine.
>
> Mike, I failed to reply in a timely manner, however your code is exactly
> what I was looking for.  Thanks, I appreciate it.  I've learned some new
> techniques.  As for my original quest for stained glass, this thread has
> changed my goals, I think.  I'm trying to find a method that yields more
> realistic results.
>
> Your code has produced some fun (albiet amateur-ish) results for me,
though!
> It actually solved a totally different problem I was having with textures.
> Thanks for that, too.  :)
>
> - How
>
>


I have yet to find a good way to programmatically carve up an image into
stained glass panes, for my own renderings I've resorted to drawing what I
want and scanning it in.

For realism I use masks of the original scanned image to 'carve' isosurfaces
into colored glass and lead objects I combine into a window. You can see two
my images that used this technique as well as some explanations and
in-progress renderings here: http://www.angelfire.com/weird2/gonk/ You can
get a good close up view of the window in the first one (Ahh Sunlight) near
the bottom of its page.

If you want to look at the code you can hunt up the IRTC entries and
download the ZIP files or e-mail me at jef### [at] twilightfaircom and I'll send
them along.

- Jeff

Post a reply to this message