William Tracy wrote:
> Wow, I like that.
> 

Thanks! :D I'm new to stonework.

Post a reply to this message

William Tracy wrote:
> William Tracy wrote:
>> Wow, I like that.
> 
> I concur.

Hehe

> 
> Something that I only learned after a lot of fiddling: Povray can build 
> heightfields from textures generated procedurally in-memory. I can 
> actually save parse time that way if the texture is detailed (large 
> images take a while to load from disk).
> 
> Something like so:
> 
> height_field {
>     function 400, 400 { fn_pattern(x, y, 0)}
> }
> 

Yes, that's what I had in mind when I mentioned 'keeping-it-in-POV'. 
After actually trying it though, on this particular machine it's taking 
longer to re-generate it than it is to load an image so I may end up 
sticking to images.  For some reason I had it in my head that a 
non-image_map based height_field worked more like an isosurface without 
a specific resolution. (Another thing I haven't learned a lot about.)

Charles

Post a reply to this message

Mark Birch wrote:
> Have you tried 'crackle solid'?
> 
> I think that pretty much does what you want.
> 
> 

Thanks, that was exactly what I needed if you don't count going 
backwards in time and fully reading the section on crackle. :D I'd 
missed that one.
Charles

Post a reply to this message

Edouard Poor wrote:
> Looks great!
Thanks!

> I may be misunderstanding what you are trying to do, but can you use the 
> "solid" keyword with the crackle pattern? That sets each crackle to be a 
> solid colour chosen from random from the colour map.

That's exactly what I needed.  I've been able to get pretty similar 
results without having to resort to image-editing!  I'm still tweaking 
though.
Charles

Post a reply to this message

Tek wrote:

> Looks good to me! I think the wrinkle looks just right, nice and subtle. I 
> think the edges look a bit odd, I'd expect them to be more vertical in 
> places (obviously awkward in a height_field).

 From the work I've seen of yours, that's something.  :)  Yeah, the 
insides of the cracks seem a bit 'round' to me.  I can narrow the cracks 
and I noticed that the amount of turbulence seems to have a pretty big 
affect on the size of the rubbley areas.  But I kinda like having 
rubbley areas...   I'll try your code to see what it looks like 
(Thanks!), although I do generally like to write my own.  I've been 
playing with 'form' some today which seems to be a quicker way of 
shaping the stones than what I'd been doing before with more pigment_map 
entries.

Charles

Post a reply to this message

It's a little darker, but now it's pure POV albeit still using two 
images to avoid re-generate it each time.  Thanks again everyone.
Charles

Post a reply to this message

Attachments:
Download 'crackle flagstone floor without gimp.jpg' (171 KB)

Preview of image 'crackle flagstone floor without gimp.jpg'

"Charles C" <"nospam a nospam.com"> wrote in message 
news:47e5db26@news.povray.org...
>
> I'll try your code to see what it looks like (Thanks!), although I do 
> generally like to write my own.

Good for you! I was just writing that 'cause I felt I could explain more 
clearly in code :)

>  I've been playing with 'form' some today which seems to be a quicker way 
> of shaping the stones than what I'd been doing before with more 
> pigment_map entries.

metric is also fun to play with. High values like metric 100 give  more 
organized patterns, not suitable for flagstones but useful for other stuff.

-- 
Tek
http://evilsuperbrain.com

Post a reply to this message

Nice work, it looks practically identical!

-- 
Tek
http://evilsuperbrain.com

"Charles C" <"nospam a nospam.com"> wrote in message 
news:47e5e7b9@news.povray.org...
> It's a little darker, but now it's pure POV albeit still using two
> images to avoid re-generate it each time.  Thanks again everyone.
> Charles
>

Post a reply to this message

Charles C wrote:
> After actually trying it though, on this particular machine it's taking 
> longer to re-generate it than it is to load an image so I may end up 
> sticking to images.

I did some experiments a while back with patterns that were around one 
gigabyte when written as a TGA. At that point, doing everything in 
memory was significantly faster. ;-)

-- 
William Tracy
afi### [at] gmailcom -- wtr### [at] calpolyedu

"Once again, the world is saved by the inscrutible machinations of 
contract law."
     -- Ubersoft.net

Post a reply to this message

Charles C wrote:
> This is a height_field (multiple actually, tiled, but not tile-able) 
> made from a nine* megapixel .png made of course from an orthogonal view 
> of crackle pattern pigment on a plane.  The pigment used for the 
> height_field is a *different* .png which is basically the original after 
> a few bucket-fills in GIMP with different colors to different 'stones.'
> 
> I could keep-it-in-POV-Ray (no image based height_fields or image_maps) 
> if I knew how to give individual stones different colors using crackle. 
>  Unfortunately it looks like there's a choice between patterns that can 
> give borders to cells, or patterns which can give different colors to 
> different cells.

It's simple:  Use the solid cell pattern as the coloring for the centers 
of the regions in the mortared cell pattern.

pigment { crackle
   pigment_map {
     [.1 rgb .75]
     [.1 crackle solid
       color_map { [0 rgb <.9,.8,.7>][1 rgb <.7,.6,.5>] }
     ]
   }
//  scale, rotate, and transform to taste
}

If you base the height_field off of a pigment instead of an image file, 
then you can use a suitable solid pattern to color the stones.  Then use 
another pigment_based height_field to generate the mortar.

Hope this helps,
John

Post a reply to this message