Hi all:

I've started my own "Landscape of the week" project, following the idea 
   and general concepts of the LOTW project by Christoph Hormann. I'm 
really amazed of what he achieved with isosurfaces, so I tried too...

My isolandscapes are not so exotic, mainly because I can't figure how to 
obtain such spectacular shapes, but also because I don't have his 
patience for testing such slow isos. I tried mutiplying only two pigment 
functions (f_ridge and f_crackle), keeping the max_gradient between 3 
and 5... at least it renders faster than the sky!

For the sky I combined a pigment for the high-level stratus-like clouds, 
and media for the low-level cumulus clouds. The media density for the 
cumulus uses the object pattern, taking as objetc a random clonglomerate 
of randomly sized spheres. It's slow, but not as slow as I expected. The 
idea arrived while looking at the latest cloud technique by Gilles, wich 
uses density files.

There are other things too small to see on this test, as fields on the 
most plain parts, with little iso-trees and simple csg houses placed 
with trace() and eval_pigment(), but this is still WIP.

Of course, following with fidelity Christoph idea, the entire scene is 
randomly parametrized, so changing a few seeds results in a different 
landscape. The render time of this one is 9h, so I think that I can 
produce one landscape per week when finished. There are still some 
problems, the most obvious one being the interaction between cloud 
containers and fog (I'm having a hard time isolating the problem).

Well, I hope to not abandon also this project, as I do with 99% of my 
proyects lately. I've a sort of sinusoidal inspiration these days...

Regards,

--
Jaime

Post a reply to this message

Attachments:
Download 'tierra14.jpg' (145 KB)

Preview of image 'tierra14.jpg'

In article <4090d486@news.povray.org> , Jaime Vives Piqueres 
<jai### [at] ignoranciaorg>  wrote:

> I've started my own "Landscape of the week" project, following the idea
>    and general concepts of the LOTW project by Christoph Hormann. I'm
> really amazed of what he achieved with isosurfaces, so I tried too...

Really nice! Any chance there will be a higher-resolution render (despite
the render time) that allows a closer look at all the little details that
are in the image?

BTW, how complex is the scene (objects, parsing time, etc)?

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

Thorsten Froehlich wrote:
> Really nice! Any chance there will be a higher-resolution render (despite
> the render time) that allows a closer look at all the little details that
> are in the image?

   Thanks! And yes, I will render this and the future LOTW images at 
bigger resolution, but not much more. There are details that look better 
if they are only barely visible and do not show their simplistic nature. 
  This is the reason also for the aerial view.

> BTW, how complex is the scene (objects, parsing time, etc)?

   Not really much, but enough to be a nightmare for me:

   + Single isosurface terrain with max_gradient 3
   + Slope-based procedural texture for the terrain.
   + 300 clouds (object pattern for each cloud uses 24 spheres).
   + 30000 isosurface trees with max_gradient 1.
   + 1000 CSG houses out of some boxes.
   + 5000 vertices for Skylight mesh.
   + No radiosity, only sun and two fill lights (opposed and zenith).

   Parsing time was 3 minutes, mainly because trees and houses both 
obtain their positions randomly, and this takes a while even with the 
2.1Ghz CPU I have now.

   Regards..

--
Jaime

Post a reply to this message

Oops.. I forgot to include the statistics:

Statistics for tierra.pov, Resolution 720 x 480
----------------------------------------------------------------------------
Pixels:          346320   Samples:         3116880   Smpls/Pxl: 9.00
Rays:          12188136   Saved:               205   Max Level: 30/32
----------------------------------------------------------------------------
Ray->Shape Intersection          Tests       Succeeded  Percentage
----------------------------------------------------------------------------
Box                            7658443           10932      0.14
CSG Intersection               1128682             961      0.09
Isosurface                    36538884         3499165      9.58
Isosurface Container          37652233        36985184     98.23
Isosurface Cache               4909174         1668509     33.99
Mesh                         302162795       239164538     79.15
Plane                        606582954         4618707      0.76
Sphere                       584649585       396131296     67.76
Bounding Box               21627336696      6881520552     31.82
Vista Buffer                 110928289        75143350     67.74
----------------------------------------------------------------------------
Isosurface roots:         36538884
Function VM calls:      1001034392
----------------------------------------------------------------------------
Calls to Noise:         6354301481   Calls to DNoise:    90048139972
----------------------------------------------------------------------------
Media Intervals:         106119393   Media Samples:       1390661285 (13.10)
Shadow Ray Tests:        297618867   Succeeded:            164005911
Reflected Rays:             135901
Refracted Rays:             135901
Transmitted Rays:          8799454
----------------------------------------------------------------------------
Smallest Alloc:                  9 bytes   Largest:           262448
Peak memory used:        161710692 bytes
----------------------------------------------------------------------------
Time For Parse:    0 hours  3 minutes  41.0 seconds (221 seconds)
Time For Trace:    9 hours  2 minutes   1.0 seconds (32514 seconds)
     Total Time:    9 hours  5 minutes  35.0 seconds (32735 seconds)

Post a reply to this message

Oh... my... God.

Fantastic.

(picks up jaw from the floor)

Post a reply to this message

de news: 4090d486@news.povray.org...
> Hi all:
>
> I've started my own "Landscape of the week" project, following the idea
>    and general concepts of the LOTW project by Christoph Hormann. I'm
> really amazed of what he achieved with isosurfaces, so I tried too...

Really great!!
>
> problems, the most obvious one being the interaction between cloud
> containers and fog (I'm having a hard time isolating the problem).
How did you do at the end? ovelapping media containers are a problem I never
got rid of


Cheers

Marc

Post a reply to this message

"Marc Jacquier" <jac### [at] wanadoofr> wrote in message
news:409100c9$1@news.povray.org...

> de news: 4090d486@news.povray.org...
> > problems, the most obvious one being the interaction between cloud
> > containers and fog (I'm having a hard time isolating the problem).
> How did you do at the end? ovelapping media containers are a problem I never
> got rid of

Actually, I think using (very) hight interval-values will get rid of the problem

Excellent image btw, Jaime, but let's be honest, that's to be expected from you
:)

cu!
--
camera{location-z*3}#macro G(b,e)b+(e-b)*(C/50)#end#macro L(b,e,k,l)#local C=0
;#while(C<50)sphere{G(b,e),.1pigment{rgb G(k,l)}finish{ambient 1}}#local C=C+1
;#end#end L(y-x,y,x,x+y)L(y,-x-y,x+y,y)L(-x-y,-y,y,y+z)L(-y,y,y+z,x+y)L(0,x+y,
<.5,1,.5>,x)L(0,x-y,<.5,1,.5>,x)               // ZK http://www.povplace.be.tf

Post a reply to this message

Photograph.

 Excellent.

 ~Steve~   ;)

Post a reply to this message

Jaime Vives Piqueres wrote:
> Hi all:
> 
> I've started my own "Landscape of the week" project, following the idea 
>   and general concepts of the LOTW project by Christoph Hormann. I'm 
> really amazed of what he achieved with isosurfaces, so I tried too...

Thanks. :-)

It's nice to see my idea inspiring something like this.  The image looks 
very good although you can't see much of the objects placed on the 
surface at that size.

> [...]
> 
> Of course, following with fidelity Christoph idea, the entire scene is 
> randomly parametrized, so changing a few seeds results in a different 
> landscape. The render time of this one is 9h, so I think that I can 
> produce one landscape per week when finished.

Without radiosity 9 hours seems quite long but i assume the clouds take 
most of the time.  You might consider using df3 files instead of object 
pattern and meshes for the objects placed on the surface.

How many parameter sets have you tried?  The problem with such a scene 
built with random parameters is to allow a reasonable variation (all 
scenes looking more or less the same is boring, apart from the landscape 
geometry my LOTW scenes suffered quite a lot from this problem).  At the 
same time you have to avoid certain parameter combinations resulting in 
unreasonable or completely uninteresting views.

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 21 Mar. 2004 _____./\/^>_*_<^\/\.______

Post a reply to this message

Gorgeous! Breathtaking technique results in a breathtaking scene. 
Details like the fringe of foam aling the shore leave me shaking my 
head.  I haven't been to your site for awhile; it may be time to do that.

If you are playing with the tension between what is pristine wilderness, 
and what is inhabited wilderness, then I think that one element having 
an exaggerated significance is the existence, or not, of roads.  Not 
that I really think you need me to point that out ;)

Post a reply to this message