POV-Ray : Newsgroups : povray.advanced-users : fade_distance for intensity specified in Lumens : Re: fade_distance for intensity specified in Lumens Server Time: 29 Oct 2020 22:43:46 GMT
  Re: fade_distance for intensity specified in Lumens  
From: Mr
Date: 4 Oct 2017 20:45:00
clipka <ano### [at] anonymousorg> wrote:
> Am 04.10.2017 um 13:53 schrieb Mr:
> > Hi, In order to create good lamp presets for Blender exporter I am trying to set
> > up a few example light sources and chose to use the lumens value as light
> > intensity multiplier, so that the relative power of every light has the proper
> > order of magnitude.
> > Chosing 1 candle at 12.57 lumens and a 100w tungsten lightbulb around 900/1000
> > lumens, then 250W halogen is already burnt with at least 5000 lumens, not to
> > mention the 1000w high pressure sodium... I need to compensate these high values
> > with a very low fade_distance such as 0.0000025, right? but then candle  or 40w
> > get blacked out, or is there any other parameter to use?
> > I don't belive I could use different fade values otherwise all proportions will
> > be lost, won't they?
>
> Caution -- you're doing it wrong!
>
>
> First, let's talk about `fade_distance`:
>
> While that parameter does indeed affect brightness, you should /not/
> think of it as being invented for that purpose.
>
> Instead, think of it as a parameter specifying something about the light
> source's geometry. It accounts for the fact that if your light source
> isn't a perfect point, the falloff will deviate from the ideal
> inverse-square-law as you get close to the source; and the value should
> have the same order of magnitude as the assumed radius of the light source.
>
>
> Second, let's talk about measures of "brightness":
>
> In POV-Ray, a light source's brightness parameter does /not/ specify
> /luminous flux/ (i.e. the total output of the light source, measured in
> lumen); this is most apparent in two phenomena:
>
> (1) Spot lights -- If POV-Ray's parameter was luminous flux, reducing a
> spot light's opening angle should increase the light source's apparent
> intensity, as the same total output would be focused onto a smaller
> area. This is not the case.
>
> (2) Fading lights -- If POV-Ray's parameter was luminous flux, changing
> the `fade_distance` shouldn't change the brightness at a reasonably
> large distance.
>
> Both phenomena rule out the parameter representing luminous flux;
> phenomenon (2) also rules out luminous intensity (i.e. the luminous flux
> per unit of solid angle).
>
> I think the most useful interpretation of POV-Ray's parameterization is
> /luminous exitance/ (i.e. the light emitted per unit of surface), with
> `fade_distance` specifying the radius of the light source.
>
> Luminous exitance is measured in lux, i.e. lumen per square meter.
>
> (Beware that the parameterization for light sources may or may not
> differ from that for emissive media and emissive surfaces by a constant
> factor such as 2, pi, or 2*pi.)
>
>
> Finally, let's talk about overly bright images:
>
> If your image is too bright (and there's nothing else wrong with it),
> /the/ one and /only/ reasonable way to solve the issue is to reduce the
> brightness of everything emissive (light sources, emissive media,
> emissive surfaces, and either the global `ambient_light` parameter or
> all `ambient` parameters).
>
> You can do this essentially two ways:
>
> (1) Specify all values in a different base unit. -- If for instance a
> choice of 1 unit = 1 lux turns out too bright, then choose e.g. 1 unit =
> 10 lux.
>
> (2) Introduce a scale factor. -- Specify a global constant, e.g.
> `#declare BrightnessFactor = 0.1`, and use this wherever you specify
> brightness.


Most values I could find from manufacturers were specified in Lumens; do you
think there would be a few ways to make a coherent conversion into lux or lm/m²?
like choosing a 1m distance from light source and considering the section of the
illuminated area cut by a square meter plane, in proportion to surface of that
illuminated volume (sphere for a point light), as a factor by which to divide
the initial lumens value? with a formula for most cylinders, for bulbs, for
cones?


Post a reply to this message

Copyright 2003-2008 Persistence of Vision Raytracer Pty. Ltd.