There's one thing I'm trying to do this fall. I'll fetch some leafes and
scan them with my little cheap flatbed scanner and then I'll try to make
small meshes from the scanned pictures. Got this idea from Oliver
Deussen. One of the authers of xfrog.

Marc 
-- 
Marc Schimmler

Post a reply to this message

Marc Schimmler wrote:
> 
> There's one thing I'm trying to do this fall. I'll fetch some leafes and
> scan them with my little cheap flatbed scanner and then I'll try to make
> small meshes from the scanned pictures. Got this idea from Oliver
> Deussen. One of the authers of xfrog.
> 
> Marc
> --
> Marc Schimmler

  Making it with meshes might be ok but I don't think you will get as
much detail as you would with my patented HF method. My method allows
you to have the leafs veins too and since the HF is internaly represented
as a mesh object it is highly effecient memory wise. Multiple copies
of a HF are no more memory intensive as multiple copies of a mesh
construct. Plus the HF method is a very easy method to produce the
exact shape of the original plus you can use the original image as
an image map for accurate color rendition.

-- 
Ken Tyler -  1100+ Povray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

Ken <tyl### [at] pacbellnet> wrote:
: much detail as you would with my patented HF method. My method allows
: you to have the leafs veins too and since the HF is internaly represented
: as a mesh object it is highly effecient memory wise.

  Would leaf veins be represented more efficiently with a proper normal map?

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

Nieminen Juha wrote:
> 
> Ken <tyl### [at] pacbellnet> wrote:
> : much detail as you would with my patented HF method. My method allows
> : you to have the leafs veins too and since the HF is internaly represented
> : as a mesh object it is highly effecient memory wise.
> 
>   Would leaf veins be represented more efficiently with a proper normal map?

See P.B.I. for an example of my method.

-- 
Ken Tyler -  1100+ Povray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

Ken <tyl### [at] pacbellnet> wrote:
:>   Would leaf veins be represented more efficiently with a proper normal map?

: See P.B.I. for an example of my method.

  It looks great, of course, but I still think that a proper bump map would
achieve an almost identical result (unless you zoom very near the leaf) and
the rendering time would be quite smaller.
  Of course there have to be a way of representing the leaf itself. Perhaps
a mesh (number of triangles would be a lot smaller than with the heightfield).
The advantage of the mesh would be that you could bend the leaf, which I think
is not possible with the heighfield.

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

The downside of a bump_map is that it isn't real, it just looks real - and only if
you are far away enough or at such an angle that it doesn't look fake. I think the
mesh idea would give more controll, but, actually, thinking about it, a hightfield
is not really a bad idea.

Just throwing my 2 cents in,

Tim Soderstrom
TigerHawk


Nieminen Juha wrote:

> Ken <tyl### [at] pacbellnet> wrote:
> :>   Would leaf veins be represented more efficiently with a proper normal map?
>
> : See P.B.I. for an example of my method.
>
>   It looks great, of course, but I still think that a proper bump map would
> achieve an almost identical result (unless you zoom very near the leaf) and
> the rendering time would be quite smaller.
>   Of course there have to be a way of representing the leaf itself. Perhaps
> a mesh (number of triangles would be a lot smaller than with the heightfield).
> The advantage of the mesh would be that you could bend the leaf, which I think
> is not possible with the heighfield.
>
> --
> main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
> ):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

TigerHawk wrote:
> 
> The downside of a bump_map is that it isn't real, it just looks real - and only if
> you are far away enough or at such an angle that it doesn't look fake. I think the
> mesh idea would give more controll, but, actually, thinking about it, a hightfield
> is not really a bad idea.

  I ran a test for mostly parse time and memory use. With 19,881 height
fields each with an image map the total parse time was 29 sec. and
consumed about 27 megs of memory. The same HF's with a simple pigment
parsed in the same amount of time and comsumed only 22 megs. In either
case the render time was not at all time consuming and was actually
surprisingly fast. The only time I imagine it would take a long time
to render is if you were doing a VERY high resolution leaf and you
were zoomed in tightly to get maximum detail. Otherwise distance has
a way of cancelling out the detail so it does not take as long to render.

-- 
Ken Tyler -  1100+ Povray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

Really? Hmm...well, I'll admit, I was skeptical at first about it because I had burned
into my mind that HF's were for terrains only, but actually they seem to be quite
useful for such a realistic and finite surface. I AM still skeptical about using them
for water, nuless you can tile them or something along those lines :)

One more question to run by you - do you know of a utility that will place the leaves
in locations - or any object, for that matter. I am using a simple leaf from Tree
Designer, but I have to place it by hand and that is not near as fun as I could be :)
Also, I am curious as to if Lparser includes a set of leaves or if you can include
them
in? I am not very familiar with lparser, and my tests have only given me blue
outputs? <shurg>

Happy tracing

Tim Soderstrom
TigerHawk

Ken wrote:

>   I ran a test for mostly parse time and memory use. With 19,881 height
> fields each with an image map the total parse time was 29 sec. and
> consumed about 27 megs of memory. The same HF's with a simple pigment
> parsed in the same amount of time and comsumed only 22 megs. In either
> case the render time was not at all time consuming and was actually
> surprisingly fast. The only time I imagine it would take a long time
> to render is if you were doing a VERY high resolution leaf and you
> were zoomed in tightly to get maximum detail. Otherwise distance has
> a way of cancelling out the detail so it does not take as long to render.
>
> --
> Ken Tyler -  1100+ Povray, Graphics, 3D Rendering, and Raytracing Links:
> http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

TigerHawk wrote:
> 
> Really? Hmm...well, I'll admit, I was skeptical at first about it because I had
burned
> into my mind that HF's were for terrains only, but actually they seem to be quite
> useful for such a realistic and finite surface. I AM still skeptical about using
them
> for water, nuless you can tile them or something along those lines :)
> 
> One more question to run by you - do you know of a utility that will place the
leaves
> in locations - or any object, for that matter. I am using a simple leaf from Tree
> Designer, but I have to place it by hand and that is not near as fun as I could be
:)

 Sounds tedious at best. Actually if you are not already aware of them there
are quite a few tree making macros floating around right now. Couple of them
with the right parameters look just like a branch with leaves on them and they
are configurable enough that you could easily substitute your own leaves into
the process. Take a look at either of the two links in my sig and find the
macro and include sections. The tree macro by Gilles Tran is a killer and
most likely the one you would want to start with. You might also look for
TomTree which is another tree making macro include file with some excellent
output.


> Also, I am curious as to if Lparser includes a set of leaves or if you can include
> them in? I am not very familiar with lparser, and my tests have only given me blue
> outputs? <shurg>

  Haven't messed around with it in a long time. What little bit I did
was simply modifying the existing example files to do my bidding. The
colors of course can be changed by rewriting the color definitions in
the pov files it spits out so it does not necessarily have to be blue.

-- 
Ken Tyler -  1100+ Povray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

Ken wrote:

> The tree macro by Gilles Tran is a killer and
> most likely the one you would want to start with.

Thanks Ken (as always). More specifically, my tree macro contains a leaf macro that
can work
independently from the tree one. It's limited but it can output an *.inc file
containing a
small mesh with the leaf. Because it's a mesh you can use it zillion time in a scene
at
little memory cost.
G.

Post a reply to this message