Jim Charter wrote:
> 
> John VanSickle wrote:
> 
> > Well, maybe not a silverback gorilla, but one could easily
> > hypothesize an isolate group of hominids that survived.
> 
> BTW I know you are working on a mesh modeller.  I am generating the
> hair direction in POV by tracing splines along the surface of the
> figure.  In this picture I used the trace function.  In my next try I
> want to use the fact that Wings allows you to assign materials to
> individual polys.
>   So when I export to .obj, I can use this to parse out a hand drawn
> sequence of vertices and their normals.  It would be easier to be able
> to merely export selected vertices and their normals to a text file,
> then POV it from there.

The fellow who develops Wings3D has been in contact with me about my
surface subidvision code, so there is some cross-pollinization going
on.  Once I can get uv-mapping working right, my own modeller will
probably appear in public.  I used it for one of the models in one of
my IRTC entries for the current round (I did a Rusty and a spin-off).

Regards,
John

Post a reply to this message

Jim Charter wrote:
> 
> Now that I have a book to look at, the hair seems to vary a lot, even
> between individuals in a species.  The source photo I was using did
> seem to show pretty scraggy hair, which I took to mean "old" also.
> Other chimps have more tuffed hair, in others its softer, more aligned
> in waves.  None look quite as wiry as in my picture though.

But it looks good.  If this were a species of intelligent life, I
could easily believe one of their old men (or women) having hair like
this.

Regards,
John

Post a reply to this message

John VanSickle wrote:

> 
> The fellow who develops Wings3D has been in contact with me about my
> surface subidvision code, so there is some cross-pollinization going
> on.  Once I can get uv-mapping working right, my own modeller will
> probably appear in public.  

I'll be looking foreward to trying it.

-Jim

Post a reply to this message

Jeremy M. Praay wrote:

> 
> Amazing work, Jim!  This is breath-taking, in my opinion.
>
Thanks Jeremy

> I would like to hear more elaboration on how you did the hair, and how CPU
> and memory intensive it was, etc., if you wouldn't mind.
> 
Yeah I know you have been attacking the hair problem too. That's partly 
why I posted this example.  Give me a day or two and I'll put something 
together.  I am actually working with a new way to do the hair which 
looks quite promising but is buggy at the moment.  Then maybe I can show 
you both methods.  I was using mesh, but not instancing, and that render 
took around 40 min I think with Jaime's lighting which includes radiosity.

I never could have done it before I fixed the case fan which seems to 
have solved my overheating problems.  I am beginning to think that while 
my processor was heating up, so was everything else and it may be the 
video card that was actually flaking out.  That might explain why it 
seems to be heat related, even though I was only taking the palomino up 
to 60 degrees before crashing.

I have half a gig of memory installed and no where near all of it gets 
used.  But that guy going to need a lot more hair before he's finished.

Post a reply to this message

Gilles Tran wrote:

> 
> IIRC, Jim has a formal art training and is an accomplished traditional
> painter (see his website), which is a great advantage in 3D from what I've
> seen from other traditional media artists who turned to 3D.

It is also true that I was first introduced to mesh modelling in an 
instructor-lead, Maya class.  So that when I came to Wings, I had the 
concept.  I knew what I wanted to do.  I just had to learn how.

-Jim

Post a reply to this message

wow!
-- 
Rick

Kitty5 NewMedia http://Kitty5.com
POV-Ray News & Resources http://Povray.co.uk
TEL : +44 (01270) 501101 - ICQ : 15776037

PGP Public Key : http://pgp.kitty5.com

Post a reply to this message

Hillary!

Jim

"Jim Charter" <jrc### [at] msncom> wrote in message
news:400485b8@news.povray.org...
> There is more wrong with this than right, but I wanted to show the
> direction I have being working in.  This actually started as an entry
> for "future" but has turned into a more measured project to teach myself
> organic modelling, texturing, hair generation etc.  It's going to take a
> while to get something really showable.
> No sss in place yet.
> Modelled in Wings
> Lighting by Jaime
> Bumpmaps and image maps generated in PS
> Uvmapped in uvmapper
>


----------------------------------------------------------------------------
----

Post a reply to this message

Jeremy M. Praay wrote:


> 
> I would like to hear more elaboration on how you did the hair, and how CPU
> 
Okay.

1. I used the trace macro to find a lines of points on the surface of 
the figure to describe the direction of the hair.  Splines are generated 
  from these points.  One spline alone the surface for the hair roots. 
One spline from points along the normals to place the end points of the 
hairs

2. The splines are sampled to get start, end, and midpoints, for each 
hair.  These points are use to create a spline.  Each hair is then built
by placing mesh along the spline

3. The hair spline can be modulated by sampling a smooth pattern such as 
bozo and using the values to rotate the midpoints slightly around the normal

4. The placement of each hair is also jittered

But the original method of tracing points to establish the hair 
direction was time consuming and cumbersome.  So going forward I plan to 
use the following method to establish the direction splines

5. In Wings, nine different materials can be assigned to nine selected 
points

6. With all nine points selected, they can be exported as a .obj mesh 
file.  In the mesh file, each polygon or 'face' is tagged with a 
separate material. e.g.

g cube1_1a
usemtl 1a
f 19//19 31//31 10//10 1//1
g cube1_1b
usemtl 1b
f 11//11 34//34 24//24 3//3
g cube1_1c
usemtl 1c
f 12//12 33//33 21//21 2//2
g cube1_2a
usemtl 2a
f 6//6 18//18 30//30 17//17
g cube1_2b
usemtl 2b
f 8//8 22//22 29//29 23//23
g cube1_2c
usemtl 2c
f 16//16 36//36 15//15 4//4
g cube1_3a
usemtl 3a
f 20//20 35//35 27//27 7//7
g cube1_3b
usemtl 3b
f 9//9 26//26 32//32 25//25
g cube1_3c
usemtl 3c
f 13//13 28//28 14//14 5//5

7. The obj mesh file can be parsed in a separate program and one of the 
four vertices and normals fo each face can be retrieved

8. These points can be used to generate the direction splines.

Post a reply to this message

Attachments:
Download '1.jpg' (35 KB) Download '2.jpg' (43 KB) Download '5.jpg' (72 KB) Download '8.jpg' (29 KB)

Preview of image '1.jpg'

Preview of image '2.jpg'

Preview of image '5.jpg'

Preview of image '8.jpg'

"Jim Charter" <jrc### [at] msncom> wrote in message
news:400485b8@news.povray.org...

Jesus, Jim, you sure know how to play it cool. A few weeks ago, you
posted a link in p.o-t to the work of a modeler whom you admired. Now
you post this, which is twice as good as any of his stuff.

I have wondered if something like this:
http://news.povray.org/povray.binaries.images/23494/
might produce better looking, less memory eating hair at the expense of
greater rendering time.

 -Shay

Post a reply to this message

Shay wrote:


> posted a link in p.o-t to the work of a modeler whom you admired. Now
> you post this, which is twice as good as any of his stuff.
>

Er..., not yet, but when I get it to where I want it to be,... maybe, 
definitely maybe ;)  Thanks for the boost.  Since I know you say exactly 
what you think, I means a lot to me.

> I have wondered if something like this:
> http://news.povray.org/povray.binaries.images/23494/
> might produce better looking, less memory eating hair at the expense of
> greater rendering time.
> 

I do plan to try that.  I do expect the render time to hurt.  I will see 
how far I can take the mesh first.

-Jim

Post a reply to this message