Jaime Vives Piqueres wrote:
> Christoph Hormann wrote:
> 
>> 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.
> 
> 
>   Well, they are prety pathetic right now, although the placement idea 
> with eval_pigment() and crackle pigments looks promissing (I've attached 
> a more close view).
> 
>> 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.
> 
> 
>   I noticed later that it was my mistake: I used too high antialias 
> settings... with something decent it tooks now 3 hours. The df3 clouds 
> are fast, but have other inconvenients. And I prefer to wait a bit more 
> but to do it all internally. It's a sort of purism by laziness... :) But 
> I will take your advice about meshes for the houses, thanks.
> 
>> How many parameter sets have you tried?  
> 
> 
>   The sun light and the sky dome depend on random altitude and azymuth, 
> and other things depend on the sun and sky colors, like the fog, the 
> clouds color or the water color and fading.
> 
>   The cloud shape and placement also have their own seeds, as the water 
> level, the trees and houses. The camera is placed randomly too, with 
> trace() to put it above the terrain at the desired height, and looking 
> always to the center of the terrain for practical purpouses.
> 
>   The isosurface function pigment is scaled and translated randmoly, and 
> the color map for it has also random entries. This gives quite diferent 
> terrains, from soft hills to mountains with abrupt walls, although half 
> of the time the results are not very usable.
> 
>   There is only one fixed slope texture at the moment, but I plan to add 
> several choices for different types of terrains (desertic, artic, 
> etc..), as I think this is the most important aspect to get clearly 
> differentiated landscapes (right now all my tests look "mediterranean").
> 
> -- 
> Jaime
> 
> 
> ------------------------------------------------------------------------
> 
first thought, where is it? second - photo, third - has the Loch Ness 
monster moved residence :-)

I looked at the first picture, then the post mentioning objects 
including 1000 houses and went back for another look for houses.
Then saw the second picture .....all became clear.

same sort of reaction to first, if not more so of first and second

some silly questions
what is the range of x,y,z in the pic for the iso-surfaces?
do you do the iso as <-1,-1,-1> to <1,1,1> and scale ?
or different ranges?
how do you do the houses, csg - yes- but what size typically as a 
standalone object for devel, and then scale by some factor, to fit in, 
process of emphirical adjustment until it looks right?

please post source or some tantalising snippets to aid understanding ?

stephen

Post a reply to this message