Just testing the mix of dynamic cloth and wind effect. Same character as in
the previous tests (Judy on the left is slightly younger),same radiosity
parameters (400/0.05),  more or less default poses, same light system by
Jaime Vives Piqueres (different filter).


G.


--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters

Post a reply to this message

Attachments:
Download 'judy_test4-5.jpg' (83 KB)

Preview of image 'judy_test4-5.jpg'

On Sat, 05 Oct 2002 22:25:21 +0200, Gilles Tran wrote:

> Just testing the mix of dynamic cloth and wind effect. Same character as
> in the previous tests (Judy on the left is slightly younger),same
> radiosity parameters (400/0.05),  more or less default poses, same light
> system by Jaime Vives Piqueres (different filter).

Don't know why I much more prefer these ones. The lightning seems better.
You should have applied wind to hair as well, shouldn't you ? :)

-- 
Jul### [at] cnedraorg
http://www.cnedra.org

Post a reply to this message

pan### [at] cnedraorg...
> Don't know why I much more prefer these ones. The lightning seems better.
> You should have applied wind to hair as well, shouldn't you ? :)

Actually there's wind on the hair too, but somehow it didn't show up due to
the viewing angle, but I got lazy and didn't want to go through the
posing/exporting/rendering process again !

The lighting seems better perhaps because the filter I used does a better
job of compensating for both the fluorescent (left) and incandescent (right)
lights, so the color is more neutral. On the other pics, the filter
compensated only for one of them (in fact the pics are lighting tests too!).
See Jaime's macro for more detail about this.

G.


--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters

Post a reply to this message

Gilles Tran wrote:
> 
> Just testing the mix of dynamic cloth and wind effect. Same character as in
> the previous tests (Judy on the left is slightly younger),same radiosity
> parameters (400/0.05),  more or less default poses, same light system by
> Jaime Vives Piqueres (different filter).

In all pictures you posted the cloth material seems unnaturally stiff. 
Usual material for such a dress would form very small ripples etc.  From
the shadow lines it seems the polygon count is really low which would
partly explain this.

BTW clothes are one thing i intend to make possible for the first release
of my mechanics simulation system.  I already made some experiments with
collision calculation based on intersection tests rather than requiring an
isosurface function.  

The way how to design clothes with this would be generating the geometry
in an unrelaxed and heavily expanded form at some distance from the body
so arms and legs are at the correct 'hole' position etc. and running some
gradient descent steps to relax the geometry (starting simulation right
away would result in heavy movement and problems at the beginning.  

Christoph

-- 
POV-Ray tutorials, IsoWood include,                 
TransSkin and more: http://www.tu-bs.de/~y0013390/  
Last updated 13 Aug. 2002 _____./\/^>_*_<^\/\.______

Post a reply to this message

3DA1AC8F.A2BFFA2D@gmx.de...
> In all pictures you posted the cloth material seems unnaturally stiff.
> Usual material for such a dress would form very small ripples etc.  From
> the shadow lines it seems the polygon count is really low which would
> partly explain this.

Actually the cloth material was done with the default parameters. There are
lots of them friction, damping, cloth density etc.) and the doc is sparse,
so a lot of testing is required. I guess there will be a nice niche market
for cloth sim parameters and tutorials in the months to come !
The clothes in the P5 library are quite low res and this is probably a
compromise between quality and simulation time. All the simulations were
quite fast, 3-4 minutes each and to be frank I'm quite fascinated by the
cloth wrapping itself around the character (or between the character and the
chair it sits on). Note that it's preferable to start the sim with
everything in a default, simple and symmetrical pose with few
angles/corners. The cloth fits itself on the character and then sort of
follows it until the final pose, though body parts like breasts can poke
through even in the final stages.

I tried a hi-res drape too but quit after 10 minutes or so, and I've read
that people had run hour-long simulations. Possibly, it should be possible
to get the cloth models through a separate app to increase the number of
vertices by surface subdivision and get more flowing models. I suppose that
Amapi could do the trick, or Wings3D. It opens lots of possibilities,
because any 1-sided, non-capped surface can be clothified.

> The way how to design clothes with this would be generating the geometry
> in an unrelaxed and heavily expanded form at some distance from the body
> so arms and legs are at the correct 'hole' position etc. and running some
> gradient descent steps to relax the geometry (starting simulation right
> away would result in heavy movement and problems at the beginning.

One of the links I gave in my first posts tries to describe some of
parameters used by Poser. I repost it below because the previous URL I
posted was mangled.
http://www.attrition.org/~demonika/enmeshed/archives/000053.html#000053
Particularly, there's a variable called Collision Offset that looks like
what you have in mind.

In any case, good luck for your cloth experiments!

G.


--

**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters

Post a reply to this message