Page 1 of 2

Efficacy scale

Posted: Tue Jun 17, 2008 9:25 am
by tony
Hello everyone:

That difference theory and practice are between:

. have activated
. Not be activated Efficacy scale on a EMITTER


Thank you.

Posted: Wed Jun 18, 2008 5:22 am
by tony
Efficacy scale cancels gain?

Posted: Wed Jun 18, 2008 7:55 am
by Bogey Jammer
I never tried to desactivate the efficacy scale, but turning it on let you to set up realistically a lamp depending of the lightning technology.
eg for a 150cm fluo tube of 56Watt, I put 56 for the power and 80lm per watt (standard value for a fluo tube)
An incandescent bulb has 15 lm/W, but to get a realistic simulation, I think the filament must be the emitter and be geometrically accurate...
Sometime I take a look in lightning gear catalogs to get some interesting data.

Posted: Wed Jun 18, 2008 8:15 am
by Borgleader
Here are some other values which may come in handy. If you want take the same mesh and try different values to see what result you get. It may give you a better idea of how this affects your render.
Blendigo Manual wrote:candle - a nominal 40 W --- 0.3 lm/w
5 W tungsten incandescent --- 5 lm/w
40 W tungsten incandescent --- 12.6 lm/w
100 W tungsten incandescent (110V) --- 16.8 lm/w
100 W tungsten incandescent (220V) --- 13.8 lm/w
quartz halogen (12-24V) --- 24 lm/w
5–24 W compact fluorescent --- 45–60 lm/w
34 W fluorescent tube (T12) --- 50 lm/w
32 W fluorescent tube (T8 ) --- 60 lm/w
36 W fluorescent tube (T8 ) --- up to 93 lm/w
28 W fluorescent tube (T5) --- 104 lm/w
5 W white LED --- 22 lm/w

Posted: Wed Jun 18, 2008 8:16 am
by CTZn
tony wrote:Efficacy scale cancels gain?
Yes it does.

Bogey Jammer, you are setting power_drawn correctly and are right about the required realism in modeling. But efficacy_scale is the percentage of power_drawn that is actually emitting light. As an example, light bulbs spend most of the wattage into heat (around 80% !), so for that technology you want to set efficacy_scale to (around) 20 (%): that's the quantity of the total iniyial power this technology actually translates into light.

edit: now that you are two guys showing Lm/W I have doubts, but I shouldn't I guess ;)

Posted: Wed Jun 18, 2008 8:24 am
by Bogey Jammer
efficacy_scale::overall_luminous_efficiacy
The overall luminous efficiacy is the luminous flux per Watt of power drawn.
type: real scalar
units: Lumens per Watt (lm/W)
restrictions: > 0
source: manual
:wink:

Posted: Wed Jun 18, 2008 8:31 am
by CTZn
mesaidnothingooops

Posted: Wed Jun 18, 2008 10:21 am
by PureSpider
CTZn
me LOL'd...

Borgleader
thanks for the data, this comes in realy handy!

Posted: Wed Jun 18, 2008 9:38 pm
by tony
Thank you all for your responses.

I think I understand the theoretical foundations of simulation of real lights through the Efficacy scale parameters.

But I have some doubts:

1. blackbody mode, so that the simulation is real Is that value has to take temperature? 5000 (default)?

2. that way the issuer must have: spherical, cylindrical, flat, etc..

3. that size should have the issuer. He found what he says the wiki

Note: The ‘power drawn’ is over the whole emitter surface, not per unit area. So a bigger mesh area or a scaled one with the same setting means comparatively less light output as seen by the observer. For ‘uniform’ emitters the spectral radiance values to use will be quite large - say 10^8.


For example:

In my house there is a 3mx3m room with a lamp on the table. The lamp has a bulb 40-Watts.

What shape and size I have to give the issuer to enlighten like my lamp?

Tonemap: Reinhard, Linear, Camera?

Thank you.

Posted: Thu Jun 19, 2008 12:31 am
by Kram1032
blackbody is a temperature in Kelvin, yes. Though where ever you see "light temperature", that's not a perfect match, just the closest you can get ;)
(For example, our sun has a temperature of about 5777K)

2. ??? - do you mean, Indigo needs more primitives? That would be nice indeed :)

3. ???


A 40W emitter mat on a 9m² plane...
and the Lm/W should be found on the bulb itself and/or on it's package...

Posted: Thu Jun 19, 2008 1:43 am
by CTZn
Hey tony, i think you are making yourself to much worries.

1. there is no wrong temperatyure, it depends on the light source you want to design. Very hot iron will have a "cool" temperature (redish colour), while for instance a camera flash will have a high temperature (blueish tint).

2. it's up to you. Probably it will have the shape of the real emiter. Take a real object as reference if this is of concern.

3. a realistic size. For a given power, the result will be different if a bulb emits light, or if a filament within the buld emits light. Obviously the power consomption is from the filament, not the bulb. :arrow: Shape and size: a filament.

You do not have to use this or that tonemapper, it's up to you, again.

Posted: Thu Jun 19, 2008 2:10 am
by tony
True, myself to much worries.

Because I think I can make an artistic rendering playing with all the values to achieve what I want.

But to make a render fotorrealístico use exact measurements. Probably this is a entelequia.

Thank you for your responses.

Posted: Thu Jun 19, 2008 4:06 am
by CTZn

Posted: Thu Jun 19, 2008 4:24 am
by Kram1032
Yeah, an amazing render on both Yafray (or was it Blender Internal?...) (in the Blender Gallery) and Indigo :D

Posted: Thu Jun 19, 2008 4:48 am
by Borgleader
Hahaha I LOVE that render. I actually showed this in an oral at school. I was doing an oral on Blender and I talked about the different renderers that could be used with it. And in the misc image portion I showed that one :lol: