LuxCore icon indicating copy to clipboard operation
LuxCore copied to clipboard

irradiance from different light source type doesn't match

Open drs-project opened this issue 4 years ago • 0 comments

Hi,

I ran some tests to verify if the irradiance pass from luxcore can be used for physical simulations. That would make luxcore a very powerfull tool.

I compared the irradiance of a point source, an extended area light source and an area light source fed with an ies file on a matte plane in 1m distance.

The fall-off of the irradiance from the center of the plane to the edges matches exactly what you would expect. So I'm very confident, that the math inside luxcore is precisely correct.

But the absolute values of the three source types don't match. blender_test_files.zip

  • An extended lambertian source with 100lm should yield 31,8lx in 1m distance. The irradiance pass shows the arbitrary value of 3,79.

  • Taking a lambertian distribution from an IES file should deliver the same results. But the irradiance pass shows 1,28 what is a factor of pi. Scaling the lumens by 3,14 delivers matching results, so I believe this is a bug!

  • A point source with 100lm should have 7,9lx in 1m distance. Here the irradiance pass shows 4,1. If the same scale was used as for the area source, I'd expect 0,95.

From my point of view the values of the irradiance pass are incorrect - at least if sources of different types are mixed. But it should be pretty simple to do the same conversion to lux for all sources alike.

Could you help me to find the code lines where this conversion is done?

Regards, Ingo

drs-project avatar Jan 27 '22 11:01 drs-project