Le 16-05-01 02:52, Thomas de Groot a écrit :
> On 30-4-2016 18:26, Jaime Vives Piqueres wrote:
>>    That's strange... I'm getting good results with just 16 samples.
>>
>
> This is basically what I do; most other details are Gilles' original
> settings (more or less):
>
>    #declare P_cloud2_1 =
>    pigment {density_file df3 "cloud2_1.df3"
>      lambda 3
>      omega 0.6
>      interpolate 1
>      translate -0.5
>      scale <1, -1, 1>
>    }
>
>    #declare Cloud1 =
>    box {-0.5,0.5
>      texture {pigment {Clear} finish {ambient 0 diffuse 0}}
>      hollow
>      interior {
>        media {
>          scattering {1, C_Sun*0.006 extinction 0.25}
>          intervals 1
>          samples 100, 1000
>          density {
>            gradient y
>            density_map {
>              [0 pigment_pattern {P_cloud2_1}
>                 turbulence Turbulence*0.1
>              ]
>              [1 pigment_pattern {P_cloud2_1}
>                 turbulence Turbulence*2
>              ]
>            }
>          }
>        }
>      }
>      scale <4/3,1,1>
>    }
>
>

samples 100, 1000 is specific to method 1 and 2. As it is now, the 1000 
is ignored.
Also, it's preferable to not set intervals at all and just use the 
default of 1.

Post a reply to this message