JRG <jrg### [at] hotmailcom> wrote:
: I have to say that I did know POV (I mean MegaPov) can make such images.
: That's the reason why I use Pov instead of Lightflow (which is much less
: flexible even if it has some interesting features).

  I wasn't actually answering to your question. I just was curious whether
I could make a similar image with povray or not, and decided to post the
results in the same thread.

: My question was about
: the render time of such features and not about quality. The Lightflow image
: I posted took just one minute or two to render (and, I have to repeat,
: everything but the shadows is raytraced). The same image made with MegaPov,
: which is posted here (I modified Warp's texture to make it more like the
: original), took more than 40 minutes to render with my Athlon 1Ghz

  This can't be true.
  The image I posted took 13 minutes to render in this Ultra5. The rendering
speed of this computer is aproximately the same as of a P-II 400MHz.
  An Athlon would have rendered it in less than 5 minutes.

  You have to also take into account that the version I posted uses an
isosurface while the lightflow version uses probably just a triangle mesh (or
does it?). 
  If I had created a triangle mesh instead, it would have probably doubled
the rendering speed. (Perhaps I should test this.)

: (with an area light 10x10 and adaptive 1).

  Well, of course if we start to add unnecessary quality... Why not use
+a0 +am2 as well? And radiosity?
  Of course this way we can make it slower.

: To Warp only: I'll give you money if you manage to make the second image I
: posted with MegaPov (hope you will). Ah, render time was somewhere between 5
: and 6 minutes...

  I can try.

-- 
#local D=array[6]{11117333955,7382340,3358,3900569407,970,4254934330}
#local I=0;#macro M()<mod(D[I],13)-6,mod(div(D[I],13),8)-3,10>#end
#while(I<6)cylinder{M()#local D[I]=div(D[I],104);M().1
pigment{rgb M()}}#local I=(D[I]>99?I:I+1);#end              /*- Warp -*/

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
:   The image I posted took 13 minutes to render in this Ultra5.

  Reducing the accuracy of the isosurface to 10^-3 reduced the render
time to 8 minutes without any visible difference in the image.

-- 
#local D=array[6]{11117333955,7382340,3358,3900569407,970,4254934330}
#local I=0;#macro M()<mod(D[I],13)-6,mod(div(D[I],13),8)-3,10>#end
#while(I<6)cylinder{M()#local D[I]=div(D[I],104);M().1
pigment{rgb M()}}#local I=(D[I]>99?I:I+1);#end              /*- Warp -*/

Post a reply to this message

Hi,

I've had a good look at JRG's Lightflow image and Warp's POV version. If you
closely examine the shadows on each of the images, the deficits of the
Lightflow zbuffering technique become apparent - all the shadows are
blurred/softened equally. There are no sharp shadows where the object is
close to the plane which isn't as accurate as the POV shadows.

Cheers,
Ian.

http://www.outerarm.demon.co.uk/graphics/graphics.html

Post a reply to this message

JRG <jrg### [at] hotmailcom> wrote:
: To Warp only: I'll give you money if you manage to make the second image I
: posted with MegaPov (hope you will). Ah, render time was somewhere between 5
: and 6 minutes...

  Ok, this one was a bit tougher. I didn't even get a 99% match as in the
first case. This is more like a 80% match:

Post a reply to this message

Attachments:
Download 'lftest2.jpg' (29 KB) Download 'lftest2b.jpg' (30 KB)

Preview of image 'lftest2.jpg'

Preview of image 'lftest2b.jpg'

Warp wrote:

>   Ok, this one was a bit tougher. I didn't even get a 99% match as in the
> first case. This is more like a 80% match:

If I didn't know better I would say these were photographs. Cool!

-- 
Ken Tyler - 1400+ POV-Ray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message