New light power system

[OnoSendai]

As part of the move to standardise the Indigo units, and in general tie the various parameters to common units and values, I've added a system that scales the exitant radiance of light sources so that light sources of a given wattage and efficiacy can be simulated.

Here's a 100W incandescent light in a sunlit room:



And here's the light code:

Code: Select all

	<rectanglelight>
		<pos>0.0 0 1.995</pos>
		<width>0.2</width>
		<height>0.2</height>

		<spectrum>
			<blackbody>
				<temperature>3500</temperature>
				<gain>1</gain>
			</blackbody>
		</spectrum>
		
		<efficiacy_scale>
			<power_drawn>100</power_drawn>
			<overall_luminous_efficiacy>17.5</overall_luminous_efficiacy>
		</efficiacy_scale>
		
	</rectanglelight>

I think this should help people struggling to combine meshlights/rectangle lights and sky lighting

edit:

Four 150W light sources, with a lower sun direction.
edit: replaced with later render:

[CTZn]

Excellent initiative Ono, seems very constructive to me, especially for arch design, but better said for overall coherency !

-----------

Some questions arose tho, regarding the code:

- Would <gain> in <spectrum> still be operational ? More precisely, is it working as a multiplier for the overall luminous "efficiacy" (<-mispelling ?), or initial wattage ?

Luminous efficacy measures the fraction of electromagnetic power which is useful for lighting.

I didn't get a clue what this is about till I found this page, and the subject seems to be prone to confusion already (between luminous efficiency and efficacy):

The distinction between efficacy and efficiency is not always carefully maintained in published sources, so it is not uncommon to see "efficiencies" expressed in lumens per watt, or "efficacies" expressed as a percentage.

where I assume

Code: Select all

<overall_luminous_efficiacy>17.5</overall_luminous_efficiacy>

is a percentage of the (electrical) power the light is fed with, right ? Manufacturers should provide such data I saw.

Erf, I'm lost, thx Ono that's cool, bulbs So the skylight code has been modified to be expressed in watts I guess ?

Nevermind, that's overally luminous

[OnoSendai]

The efficiacy_scale element is optional, if it is used, it overrides the gain. Otherwise the gain works as normal.

The overall luminous efficiacy is the luminous flux per Watt of power drawn.
There are some values on that wikipedia page.

Sunlight already has the correct spectral radiance units, so it should work well with the sunlight.

[CTZn]

Ooof, the answer is way simpler than the question, sorry I spread out my confusion people, next time I'll wait a bit (I came back to erase half the post, but now it has been answered...)

Great then !

Navigating the forum is hell now ! You can tell people are at work >lol< in Europa, and back home in NZ !

Cheers !

[Big Fan]

edited to remove all

[Wedge]

It looks great!

It should definately help with the sun and light mismatches.

[IMSabbel]

Code: Select all

	<rectanglelight>
		<pos>0.0 0 1.995</pos>
		<width>0.2</width>
		<height>0.2</height>

		<spectrum>
			<blackbody>
				<temperature>3500</temperature>
				<gain>1</gain>
			</blackbody>
		</spectrum>
		
		<efficiacy_scale>
			<power_drawn>100</power_drawn>
			<overall_luminous_efficiacy>17.5</overall_luminous_efficiacy>
		</efficiacy_scale>
		
	</rectanglelight>

That doesnt make sense, imho. With blackbody temperature and power in W, the luminous efficiacy is already defined by the planck formula. There is no reason for an additional option. In fact setting a different overall luminous efficiacy like in that example HAS to break something, somewhere (or your code is wrong).

[OnoSendai]

IMSabbel:
It's true that a 'blackbody' light source, at a particular temperature, has a certain luminous efficacy, which would be the integral of the spectral power density weighted by the luminosity curve, divided by the integral of the SPD.

The main reason I'm using *overall* luminous efficacy, as opposed to luminous efficacy, is that it is a more convenient value.
Because I'm using overall luminous efficacy, light sources that waste power due to heating etc.. can be used. I think this is a more useful measurement, because it allows us to use the power consumption that we are familiar with (e.g. 100 Watts), as compared to using the total radiated power.

Also for stuff like an RGB spectral emissive power density, the spectrum is really only defined over the visible wavelengths. As such, there is no way to recover the overall emitted power. So the power drawn by the light source is a more convenient value to use (and one value that may be known by the scene author).