Indigo Renderer Forum

http://www.indigorenderer.com/materials/categories/15

And more to come... What would you like in priority?

Etienne

Oh wow those are really nice! I particularly love the Blue LED one.
If you could make some basic every-day type of lights that would be great too. Just your average 25/50/75/100Watt house bulbs would be great for the less experienced users. Name them something basic too

Excuse my ignorance, but shouldn't the candle flame be orange? Or is that tonemapping?

The spectra based on the data from the amici spectroscope are already included as presets in Blendigo 2.3.1

Could we get the presets from sketchup and blendigo on the material database? paste em up!

That's what I'm doing! Most of them come from the Amici page.

By the way, how did you do the interpolation from non-linear scale to linear? If not done properly you will mess up all peak ratios (hence precise color) in spectra with sharp peaks.

Etienne

Yes, the candle looks quite too neutral, I checked multiple times and did no mistakes. That comes from the source spectrum. Next time I have a candle and a spectrometer I redo it!

About the "wattage" request, it's not really possible to do it in Indigo I think, since in the .igs the luminous flux is defined in the model, not the material (am I right? this is what I've understood from the technical doc but I know it's quite outdated)

Etienne

galinette wrote:By the way, how did you do the interpolation from non-linear scale to linear?

I believe it was already done by the developers of LuxRender, who I took the data from.
All I did since then was to linear-interpolate to a different wavelength range.

However, I would like to cross-check the data I have against yours, just to be sure.

Luxrender is able to use arbitrary wavelength scale (the spectrum is a list of wavelength/intensity pairs) so they just calculated one pair per spectrum pixel. It ends up with a step much finer in the blue range (~0.15nm) than in the red range (~5nm).

Danger of simple linear interpolation is that you can dramatically change the peak areas for peaky spectras, which then ends up in a sensible change in color.

Etienne

Hi Galinette.
The wattage should be set in material editor/emitter attributes/power. at least in Skindigo.

galinette wrote:What would you like in priority?

More led lights would be great! Thank you and thanks for the black light answer..

I love you.

Ferolina : Yes, in skindigo, the wattage is set in the material properties. But when you export to igs, it appears that the wattage is set in the <model> element, and not in <material>. In that case, it's not possible to export a igm material with included power.

Any help from the igs xml format appreciated! (Please, update the reference doc!)

Etienne

Here are plots of two of the spectra that I have.

Here's one you can directly compare with your own data:

Mercury Lamp - Osram HQA 80W (HPM2)

Hard to see on a chart.

Here is a close up on the leftmost peaks of the HQA80W.

Interpolation is not linear or cubic interpolation, but equal to the local average of the source spectrum, so that we keep the same peak area (and hence the area ratios of the peaks, which fix the color).

Etienne

galinette wrote:Yes, in skindigo, the wattage is set in the material properties. But when you export to igs, it appears that the wattage is set in the <model> element, and not in <material>. In that case, it's not possible to export a igm material with included power.

Oh, I get it now. Thanks for your clear answers Galinette!