On 23-9-2015 16:27, clipka wrote:
> Am 23.09.2015 um 09:17 schrieb Thomas de Groot:
>> On 18-9-2015 17:07, clipka wrote:
>>> Unless using an area_light with area_illumination on, fade_distance
>>> should typically be roughly equal to the hypothetical radius of whatever
>>> light source you are trying to simulate. So for instance, to simulate a
>>> classic frosted E27 incandescent light bulb, and presuming you're using
>>> a scale of 1m per POV-Ray unit for your scene, a fade_distance of 0.03
>>> would be about the right value.
>>
>> Could you elaborate a bit on (the use of) /area_illumination/ please? I
>> have tried to look up info in the docs but there is nothing
>> comprehensive there (or is there?). Thanks!
>
> Normally, using an area_light only affects brightness attenuation due to
> occlusion - in other words, soft shadows. For all other purposes - most
> notably the angle of incidence to compute highlights and diffuse
> brightness (but also the distance to compute distance-based falloff) -
> the light source still behaves like a spotlight.
>
> Consider you want to simulate a fluorescent light tube mounted close to
> a wall. If you do this with an array of, say, 1000 tiny point light
> sources, the illumination of the wall will be computed correctly, giving
> you a long patch of diffuse light and (if the wall is somewhat shiny) a
> distinctively line-shaped highlight.
>
> If you try the same feat with a 1x1000 area_light, even if you don't use
> the "adaptive" keyword you'll find that both the patch of diffuse light
> as well as the highlight will /not/ have the expected linear shape, and
> instead look like you had only one bright point light source at the center.
>
> Using "area_illumination" forces POV-Ray to do the diffuse and
> highlights computations for each "lightlet" separately, to give you the
> proper illumination effect.
>
> (I don't know off the top of my head whether area_illumination also
> affects distance-based falloff; I guess so, but I'm not sure.)
>

Excellent! Thanks indeed.

-- 
Thomas

Post a reply to this message