So, I'm not the first... :-)

Attached is a sample of grass I cooked up. It's built from smaller squares
rotated randomly to create this patch. I tried a larger size but my computer
(64Mb) refused to swallow it on account of the memory it needed. I'll try it on
another system, but the limitations are already obvious. I don't really
understand this, but I'll do some experimenting.

No flowers (yet).

Remco

Post a reply to this message

Attachments:
Download 'patch.jpg' (64 KB)

Preview of image 'patch.jpg'

Pretty good. How long did it take to render and what kind of machine was it?
What does the code look like?

Jon

Remco de Korte wrote in message <36AEE62C.7F983890@xs4all.nl>...
>So, I'm not the first... :-)
>
>Attached is a sample of grass I cooked up. It's built from smaller squares
>rotated randomly to create this patch. I tried a larger size but my
computer
>(64Mb) refused to swallow it on account of the memory it needed. I'll try
it on
>another system, but the limitations are already obvious. I don't really
>understand this, but I'll do some experimenting.
>
>No flowers (yet).
>
>Remco

Post a reply to this message

Remco de Korte wrote:

> So, I'm not the first... :-)
>
> Attached is a sample of grass I cooked up. It's built from smaller squares
> rotated randomly to create this patch. I tried a larger size but my computer
> (64Mb) refused to swallow it on account of the memory it needed. I'll try it on
> another system, but the limitations are already obvious. I don't really
> understand this, but I'll do some experimenting.

Nicest grass yet, as most real grass lays down, instead of standing up perfectly...

Only problem is that you can see the lines of grass, so it looks like planted sods,
rather than settled natural grass....

Simon
http://home.istar.ca/~sdevet

Post a reply to this message

A good grass example, I've seen and made far worse..

But... isn't grass a bit more glossy ????? Just thinking, depends on the sort, of
course..

There seems to eb a bit too much regularity in this one(lines). and I don't think It'd
look that great in a distance..

Perhaps if you scaled all the plants a bit in a random way ?? and then placing them in
a
spiral pattern, should be possible.... and it'd remove some of the lines, and instead
add
others ;-)


//Spider
Remco de Korte wrote:
> 
> So, I'm not the first... :-)
> 
> Attached is a sample of grass I cooked up. It's built from smaller squares
> rotated randomly to create this patch. I tried a larger size but my computer
> (64Mb) refused to swallow it on account of the memory it needed. I'll try it on
> another system, but the limitations are already obvious. I don't really
> understand this, but I'll do some experimenting.
> 
> No flowers (yet).
> 
> Remco
> 
>  
------------------------------------------------------------------------------------------
>  [Image]

Post a reply to this message

(In reply to the other posts as well:)

The image posted previously was a small patch of grass tiled to file a larger
area. I rotated the grass at random, but still there seems to be some sort of
pattern. I saw that and don't know what is causing it. Anyway, you could easily
avoid this in more then one way. The most simple way is to translate each line
of patches a little bit. The code is not a macro (yet) but all parameters are
variables that define the length and width of the patch. So, you could, for
instance make a larger patch. I don't know how this would affect the amount of
required memory (I thought at first that the patch would be declared once, and
after that you only needed some pointers for each time you use it, but this
obviously isn't the case - but that's more something for the
preogramming-newsgroup).

As for the 'glossiness': the colour might be totally off, I just put in some
random green, no finish or other subtleties, but that shouldn't be a problem.
I've seen very nice grass-images on this group already, but I needed some grass
for a couple of scenes I'm making and what I'm looking for is something that
looks like it could be real but very much isn't. Sort of like a cartoon, but
then not too simplistic.
I guess I could add some nice extras ore refine the grass-texture, but that
shouldn't affect the render time too much.

The previously posted image took 15 seconds to parse and about five hours to
render.
I've attached a new image, which is a close-up of one larger patch, which took 3
minutes to parse and 3:20 hours to render.
I also just finished a new 'pasture' with about 50 patches which took 31 minutes
to parse and 7 hours to render (109Mb peak memory usage).
This all on a Pentium 266 with 64M, so there's room for improvement there.

Regards,

Remco

Spider wrote:
> 
> A good grass example, I've seen and made far worse..
> 
> But... isn't grass a bit more glossy ????? Just thinking, depends on the sort, of
course..
> 
> There seems to eb a bit too much regularity in this one(lines). and I don't think
It'd
> look that great in a distance..
> 
> Perhaps if you scaled all the plants a bit in a random way ?? and then placing them
in a
> spiral pattern, should be possible.... and it'd remove some of the lines, and
instead add
> others ;-)
> 
> //Spider
> Remco de Korte wrote:
> >
> > So, I'm not the first... :-)
> >
> > Attached is a sample of grass I cooked up. It's built from smaller squares
> > rotated randomly to create this patch. I tried a larger size but my computer
> > (64Mb) refused to swallow it on account of the memory it needed. I'll try it on
> > another system, but the limitations are already obvious. I don't really
> > understand this, but I'll do some experimenting.
> >
> > No flowers (yet).
> >
> > Remco
> >
> >  
------------------------------------------------------------------------------------------
> >  [Image]

Post a reply to this message

Attachments:
Download 'grass.jpg' (35 KB)

Preview of image 'grass.jpg'

Remco,

Great Grass!

A couple of ideas to get rid of the lines/regularity. First, declare several
different patches and then use them randomly. Create several patches that
are only half as dense as you require, then when you tile them, overlap by
50%. Again choosing random pairs will get rid of the pattern. Finally, if
you declare your patches as being only 25% as dense as you want, then
overlap by 75 % and rotate them around the vertical axis a random amount,
the edges prety much disappear.

Good luck on this, I have been playing with this for a year or so off and
on. The largest "field" I have done is a kilometre square. This was a field
of wheat with plants spaced 10 cms apart and 8 - 10 blades per plant. That
is a LOT of triangles.

Regards
Gordon
<gbe### [at] birdcameroncomau>

Post a reply to this message

Gordon wrote:
> 
> Remco,
> 
> Great Grass!
> 
> A couple of ideas to get rid of the lines/regularity. First, declare several
> different patches and then use them randomly. Create several patches that
> are only half as dense as you require, then when you tile them, overlap by
> 50%. Again choosing random pairs will get rid of the pattern. Finally, if
> you declare your patches as being only 25% as dense as you want, then
> overlap by 75 % and rotate them around the vertical axis a random amount,
> the edges prety much disappear.
> 
> Good luck on this, I have been playing with this for a year or so off and
> on. The largest "field" I have done is a kilometre square. This was a field
> of wheat with plants spaced 10 cms apart and 8 - 10 blades per plant. That
> is a LOT of triangles.
> 
> Regards
> Gordon
> <gbe### [at] birdcameroncomau>

Thanks, for the comments. I think overlapping is a good way to get even more
randomness in the grass. I was aiming for a low-memory sort of grass, but missed
at that.

Remco

Post a reply to this message

The grass is always greener on the other side of POV ;)
Seems no ones ever satisfied, last several grasses I've seen looked okay
enough. About that grooving in the patch (or is it several individual
patches?), maybe it's a result of a limited range of shifting, either
rotation or translation, you've used. Even if random then it would still
be patterned to those constraints. Think so? Or no...


Remco de Korte wrote:
> 
> So, I'm not the first... :-)
> 
> Attached is a sample of grass I cooked up. It's built from smaller squares
> rotated randomly to create this patch. I tried a larger size but my computer
> (64Mb) refused to swallow it on account of the memory it needed. I'll try it on
> another system, but the limitations are already obvious. I don't really
> understand this, but I'll do some experimenting.
> 
> No flowers (yet).
> 
> Remco
> 
>   ------------------------------------------------------------------------
>  [Image]

-- 
 omniVERSE: beyond the universe
  http://members.aol.com/inversez/POVring.htm
=Bob

Post a reply to this message

I think the best low-mem I've seen is Ken's...
I did it and redid it ;-)
the same way you did with patches, I did with that one. Don't think I have it left
thou...


//Spdier

Remco de Korte wrote:
> 
> Gordon wrote:
> >
> > Remco,
> >
> > Great Grass!
> >
> > A couple of ideas to get rid of the lines/regularity. First, declare several
> > different patches and then use them randomly. Create several patches that
> > are only half as dense as you require, then when you tile them, overlap by
> > 50%. Again choosing random pairs will get rid of the pattern. Finally, if
> > you declare your patches as being only 25% as dense as you want, then
> > overlap by 75 % and rotate them around the vertical axis a random amount,
> > the edges prety much disappear.
> >
> > Good luck on this, I have been playing with this for a year or so off and
> > on. The largest "field" I have done is a kilometre square. This was a field
> > of wheat with plants spaced 10 cms apart and 8 - 10 blades per plant. That
> > is a LOT of triangles.
> >
> > Regards
> > Gordon
> > <gbe### [at] birdcameroncomau>
> 
> Thanks, for the comments. I think overlapping is a good way to get even more
> randomness in the grass. I was aiming for a low-memory sort of grass, but missed
> at that.
> 
> Remco

Post a reply to this message

Bob Hughes wrote:
> 
> The grass is always greener on the other side of POV ;)
> Seems no ones ever satisfied, last several grasses I've seen looked okay
> enough. About that grooving in the patch (or is it several individual
> patches?), maybe it's a result of a limited range of shifting, either
> rotation or translation, you've used. Even if random then it would still
> be patterned to those constraints. Think so? Or no...
> 

Probably, which makes it a 'groovy grass'.

Remco

Post a reply to this message