Features I'd Like to See (Go Ahead! Tell Me These are Just Fluff!)


Sure, everyone wants those. I am sure the devs do too. Thing is, overlooking is not the problem. The devs just have to much on their plate. There are way more requests then man power. Btw, The official place to post suggestions is blenderstorm.org

  1. is just not important eneough i guess with “open recent” and folder bookmarks already available.
  2. afaik someone made a patch for this which is already waiting for being included
  3. As there is no realistic camera model in blender to set a shutter time / F-Stop value setting the scale for real world units would be arbritary as well. And there is imho no need for it. DCC is not about simulating the real world but approximating it.



Aren’t watts just an arbitrary scale and not an actual measure of brightness, just power consumption?

Lumens would be “more correct” I think.

The scale isn’t really arbitrary and is used by many many packages…

for a diffuse surface then If your surface normal is directly facing A WHITE light and there’s no falloff attenuation then a light value of 1 will perfectly illuminate that point… so the pure surface colour* diffuse reflectance is what you see…

brighter values “over expose”, falloffs attenuate with distance… colours modulate the result… etc etc…

I know. E.g. in Maxwell you can specify the light intensity in Lumens. But in Maxwell you can also scale the brightness of your scene in general with real world camera settings like iso, f-stop and shutter speed. In combination these parameters make sense as you can shoot a foto in similiar conditions and compare to a rendered result in maxwell and both images will look similar as well.
I’m sure this would be ideal for adding VFX stuff to live action footage as you can take the exact values from the gaffer(I asked leo.org what “Oberbeleuchter” is in english. Is it realy the right term?) and the cameraman and would get perfect fitting image in far less time.
But in BI setting a light to e.g. 200 lumen would most likely not give you the result you’d expect.


Aren’t watts just an arbitrary scale and not an actual measure of brightness, just power consumption?

Yep, like a 12 watt compact flourescent = a 60 watt incandescent bulb. W=j/s, or (N x M)/s, basically a rate of force on an area per second.

As jay says, brightness per watt depends on bulb technology, so it’s a very arbitrary scale…

Depends on falloff type, which is why i said “with no falloff” you get perfect illumination…(that’s at any distance)

With linear falloff it falls of from full intensity to 0 over the “range”, in a linear way so if your surface is half the range distance away from the lightspource it gets half the intensity…

there are other falloff types that are more realistic … quadratic is the most natural which stays strong close to the light then rapidly decays as it approaches the range setting.

  1. What aspect of the physics of light behaviour is the scale based upon?

It’s not physically correct, it’s just simple fakes like lambert shading (dot product of surface normal with light incidence angle) summed for each light source and multiplied by the light intensity after whatever distance falloff that has been chosen.
…whatever, these hacks have been around for ages…

  1. How do I convert from the Blender lighting scale to (OK) lumens, or candles, or watts?

You don’t, it’s not physically correct. as frigge mentioned there’s no aperture, no fstops… no shutter speed… no simulation of film response… so it’s all a"relative" scale based on your lightsources and falloff choices…

You just “eyeball” all of that (usually with a fair chunk of compositing colour corrections to fake film response etc…)

Thanks for jumping into the discussion, Michael. I’m really hoping to learn something here.

-Ron T.

I’m not sure i’m much help :wink:

Cycles is a more physically correct renderer but very early. At least there it will be possible to use actual data if you know the scales… (it doesn’t have fstops, index of refraction, shutter speed etc etc either but because it’s physical then the maths is such that those could be added!

1)If you switch off attenuation the distance doesn’t matter, so a surface perfectly facing the light source will always get a perfect illumination.
For example:
If you have a red surface (a value of (1, 0, 0)) and your light source has an intensity of 1.0, what you’ll see on your rendered image is the same red color (1.0 * (1, 0, 0) = (1, 0, 0)).
The formula for diffuse lambertian illumination is just the dot product of the (normalized) light vector and the (normalized) surface normal.

What I mean with this is that illumination in digital rendering is just an approximation based on visual phenomena observed in the real world. It has nothing to do with how lighting in the real world works.
2)it is not based on the physics of light behaviour and
3)there is no scale and you cannot convert it to lumens/kelvin or whatever.

All you can do is adjust your lighting so that it looks good.

Someone in the 3delight forums once asked a similar question and the developers explained it quite well and also recommended a good book on that topic, if you’re interested in how this whole rendering stuff works:

EDIT: i guess 2 answers are bette than none :slight_smile:

As someone how spends his time designing lighting circuits for builds its not arbitrary. Its how much power , which is what watts is a measure of, a light bulb will consume regardless of how much of that power it will convert to light. for an engineer that is more important as it determines how many light bulbs you can put on one circuit if for example you were using a 10A circuit breaker on each circuit.

Lumens is the measure of how bright a light is, so you use that.