Hello fellow raytracers,

A friend of me has send me this link:

http://www.saarcor.de/

I looks rather promissing, so i can't wait to see more improvement of this
project!

Greetings,

JWV

PS: I am not connected to this project (not even close :-) )

Post a reply to this message

JWV <jwv|at|planet.nl> wrote:
> http://www.saarcor.de/

> I looks rather promissing, so i can't wait to see more improvement of this
> project!

  So far the example images don't look any impressive to me. They don't seem
to contain anything that couldn't be done quite easily with current
scanline-rendering 3D cards.
  If you can't have smooth triangles (which you clearly can't at least
in the current prototype), raytracing is pretty useless.

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

Warp wrote:
> JWV <jwv|at|planet.nl> wrote:
> 
>>http://www.saarcor.de/
> 
> 
>>I looks rather promissing, so i can't wait to see more improvement of this
>>project!
> 
> 
>   So far the example images don't look any impressive to me. They don't seem
> to contain anything that couldn't be done quite easily with current
> scanline-rendering 3D cards.

This was my thought, too.
But the idea to implement a raytracer in FPGAs is
interesting.

Sebastian H.

Post a reply to this message

"Warp" <war### [at] tagpovrayorg> wrote in message 
news:416a3fa9@news.povray.org...
> JWV <jwv|at|planet.nl> wrote:
>> http://www.saarcor.de/
>
>> I looks rather promissing, so i can't wait to see more improvement of 
>> this
>> project!
>
>  So far the example images don't look any impressive to me. They don't 
> seem
> to contain anything that couldn't be done quite easily with current
> scanline-rendering 3D cards.
>  If you can't have smooth triangles (which you clearly can't at least
> in the current prototype), raytracing is pretty useless.
>

Well, they are clearly using true reflections which are impossible with 
current scanline
accelerators.  One more thing, is that scenes with upwards of 100 million 
triangles
(like the sunflowers) may be approaching the point where raytracing becomes 
more
efficient than scanline rendering - because of the log(n) cost of finding 
intersecting
objects for a pixel may be faster than rasterizing all the polygons.

Post a reply to this message

Andrew Clinton <ajc### [at] uwaterlooca> wrote:
> Well, they are clearly using true reflections which are impossible with 
> current scanline
> accelerators.

  max_trace_level 2 reflections are possible with accelerators up to a
point. However, none of the example images really showed the advantages
of being able to have deeper reflection levels. This is because there
aren't smooth triangles.

-- 
#macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb M()}}
N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}//  - Warp -

Post a reply to this message

On Wed, 13 Oct 2004 04:57:10 -0400, Warp wrote:

> Andrew Clinton <ajc### [at] uwaterlooca> wrote:
>> Well, they are clearly using true reflections which are impossible with 
>> current scanline
>> accelerators.
> 
>   max_trace_level 2 reflections are possible with accelerators up to a
> point. However, none of the example images really showed the advantages
> of being able to have deeper reflection levels. This is because there
> aren't smooth triangles.

Could you point me to an accelerator that has such support?  I know it can
be done by stretching the shading language to its limit and rendering in
passes like:
http://graphics.stanford.edu/papers/rtongfx/
This is not architectural support.

But I don't know of any mainstream graphics hardware that has support for
reflections.  Maybe you are thinking of reflections maps?  This is not a
true reflection, because it is only correct at a single point.  Also the
mirror reflection you see in games are also tricks.

Andrew

Post a reply to this message

Andrew Clinton <ajc### [at] uwaterlooca> wrote:
> Maybe you are thinking of reflections maps?  This is not a true reflection

  What is a "true reflection" in your opinion?

  If a reflection looks good, then it's as "true" as it gets.
  If we start argumenting about what is "true" and what isn't, then nothing
POV-Ray or any other 3D engine does is "true". Everything is just an
approximation.

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

On Wed, 13 Oct 2004 09:20:02 -0400, Warp wrote:

> Andrew Clinton <ajc### [at] uwaterlooca> wrote:
>> Maybe you are thinking of reflections maps?  This is not a true
>> reflection
> 
>   What is a "true reflection" in your opinion?
> 
> 
Okay, I define it to be one that obeys the law of reflection at all points
(not just one).  Of course you can define it any way you'd like, but I
don't think you'd argue that this definition is "more true" than the way
reflection maps work :-)

Andrew

Post a reply to this message

Andrew Clinton <ajc### [at] uwaterlooca> wrote:
> Okay, I define it to be one that obeys the law of reflection at all points
> (not just one).  Of course you can define it any way you'd like, but I
> don't think you'd argue that this definition is "more true" than the way
> reflection maps work :-)

  Reflection maps work pretty well assuming that none of the reflected
scene is too close to the reflecting object.
  If the rest of the scene is far enough from the reflecting object,
there's practically no difference between reflection maps and reflection
by raytracing (give or take some pixelation).

  Reflection maps have the advantage that they are very fast to render
(it can be done in real-time) and many effects can be done for free with
them (eg. blurred reflection, phong highlights, etc).
  They have the disadvantage that they don't work well if an object
being reflected is too close to the reflecting object, and inter-reflection
between two or more reflecting surfaces is not trivial.
  (One could think that an additional disadvantage is that it only works
with flat polygons, but in fact that's not a limitation of reflection
maps: Nothing stops you from using reflection maps with any surface you
are eg. raytracing. Of course it often doesn't make sense, specially in
raytracing, but it's perfectly possible.)

-- 
plane{-x+y,-1pigment{bozo color_map{[0rgb x][1rgb x+y]}turbulence 1}}
sphere{0,2pigment{rgbt 1}interior{media{emission 1density{spherical
density_map{[0rgb 0][.5rgb<1,.5>][1rgb 1]}turbulence.9}}}scale
<1,1,3>hollow}text{ttf"timrom""Warp".1,0translate<-1,-.1,2>}//  - Warp -

Post a reply to this message