El Wed, 03 Nov 2010 01:48:38 -0500, Tim Cook escribió:

> [..] I seem to recall
> reading somewhere that ideally the only property of intensity,
> fade_power, and fade_distance that would need to be changed to be
> accurate to reality would be /intensity/

  Yes, but for that to be realistic, fade_power should be 2: intensity is 
inversely proportional to the square of the distance from the source.

> what, if any, unit of
> luminosity is appropriate to apply to a light source if 1 POV-unit=1 m
> and fade_power and fade_distance are both set to 1.0?  Does the
> resulting variable of intensity have any real-world analog, or is there
> any defined conversion factor?  Even an approximation would be useful.

  I researched this topic years ago, and I found it too complicated for 
just artistic purposes... all that matters really is that all lights must 
have realistic attenuation. Also, if you have several lights which are 
supposed to have different intensities, the difference in intensity 
should be realistic. So, for example, you can use real lumens data as the 
intensity for your lights, scaling the data to adjust the brightness of 
your image.

  But if you're trying to match a scene lighting level with something 
measured from reality, the best way would be to first perform tests 
against a real scene, "a la" cornell box, to obtain that "conversion 
factor".


-- 
Jaime Vives Piqueres
		
La Persistencia de la Ignorancia
http://www.ignorancia.org

Post a reply to this message