honnza nous apporta ses lumieres en ce 2007/10/07 11:05:
> "Rune" <aut### [at] runevisioncom> wrote:
>> "Warp" wrote:
>>> Rune wrote:
>>>> Actually no, you just said "regarldess of distance", not mentioning any
>>>> maximum distance at all.
>>>  What I said was "are you sure that shadow tests are always done with a
>>> fading light regardless of distance". Or in other words, "are you sure
>>> POV-Ray doesn't have a check that if a fading light source is far enough,
>>> it doesn't skip doing shadow tests".
>>>
>>>  I don't really understand how you misunderstood what I said.
>> I didn't misunderstand anything, but are using expressions that are not well
>> defined.
>>
>>>> How would you define "the maximum distance the
>>>> light source would have an effect from"?
>>>  With a simple mathematical formula. From the light source parameters
>>> you can calculate what is the distance at which the effect of the light
>>> source drops below negligible.
>> You didn't answer the question, so I ask again: What is negligible? Anything
>> below the ADC-bailout threshold? Or the
>> floating point precision error threshold? Or the internal EPSILON threshold?
>>
> IMHO ADC ;-)
>> And what if there are many light_sources close to each other, so that any
>> single one is negligible at distance X, but combined their effect is
>> significant?
>>
> these should be made area. Or you can use a lower bailout value (maybe a
> light should accept the ADC_bailout keyword).
>> Rune
> 
> 
> 
> 
Making several "close together" light is not always advisable. Just imagine that 
those lights are placed irregularly, and only visualy close while been actualy 
far appart from each other. An example would be a line of street lights placed 
every 100m, placed 2 Km away in a line close to the line of sight. One would not 
contribute to your shadow, but 100 of those can very well cast a notable shadow.

-- 
Alain
-------------------------------------------------
'Hurhurhur, 2400 baud sucks' - V.bis and Baudhead

Post a reply to this message