issue on ies light intensity

[nocivo]

Hi,
first of all, thanks for your hard work. I am very enthusiastic of this restart of lux. Below I report a bug regarding the IES.
I have found an issue in IES emission intensity. I put two spot light source in a room, and I tested with luxrender classic and with luxcore alpha4 in blender. the left spot light is made by a material emitter and a glass lens to concentrate the light (very hard to compute). The right spot light is doing by a point light with applied an IES file from ERCO. The intensity of both light is the same, but the IES light seam that are about 15 time less intense. For this I test the ies light in luxclassic to check if there is the same issue, but seem that in lux classic the result is correct. please take a look on image below.

Thanks
nocivo

luxcorerender alpha4 - 100spp

lux classic - 100spp

luxcorerender alpha4 - 100spp - IES gain set to 14

[Dade]

first of all, thanks for your hard work. I am very enthusiastic of this restart of lux. Below I report a bug regarding the IES.
I have found an issue in IES emission intensity. I put two spot light source in a room, and I tested with luxrender classic and with luxcore alpha4 in blender. the left spot light is made by a material emitter and a glass lens to concentrate the light (very hard to compute). The right spot light is doing by a point light with applied an IES file from ERCO. The intensity of both light is the same, but the IES light seam that are about 15 time less intense. For this I test the ies light in luxclassic to check if there is the same issue, but seem that in lux classic the result is correct. please take a look on image below.

It looks like a bug, can you post the test scene to reproduce the problem ?

[nocivo]

hi,
below the blender scene and the IES file.

the scene attached has the ies light gain set to 14. to recreate the bug, please down to 1. in this way you have the same intensity of the left light.

thanks

[Dade]

below the blender scene and the IES file.

the scene attached has the ies light gain set to 14. to recreate the bug, please down to 1. in this way you have the same intensity of the left light.

Thanks, going to look into the problem.

[Dade]

the scene attached has the ies light gain set to 14. to recreate the bug, please down to 1. in this way you have the same intensity of the left light.

I should have fixed the problem: https://github.com/LuxCoreRender/LuxCor ... 48fcda8ce1

There was a missing "x 4 x PI" (12.56) factor that is pretty near to the "x 14" you had to use to much the outputs.

[Dade]

This is a rendering from latest sources with the fix:

[lighting_freak]

Hi,

Thank you for fixing the IES bug.

How does lux interpret the light distributions that are stored in that file?
Do you read the luminous flux from file or do you compute a relative value?
Are there any angular limits for IES interpretation?

To explain my application... I'd like to visualize car's headlamp beam patterns. They are usually very unsymmetrical. In some cases a resolution of 0.02 x 0.02 degrees is used describe bright/dark cut off lines.

Therefore the description locally very limited. Only +- 40deg horizontal and +10deg and -30deg vertical area.

In classic lux I had some troubles, but it was working issuing less detailed data.
Other renderers promise full IES support but they fail on that task quite often.

@B.Y.O.B.
Do you see any chance to integrate a low poly preview of IES lighting cone in Blender? Especially for unsymmetrical distributions it may simplify orientation a lot.

BR

[B.Y.O.B.]

@B.Y.O.B.
Do you see any chance to integrate a low poly preview of IES lighting cone in Blender? Especially for unsymmetrical distributions it may simplify orientation a lot.

You mean in the OpenGL viewport?
It could be done with an operator: https://blender.stackexchange.com/a/61712/29401
However it sounds like a bit of work.
The viewport render should provide you with a much more accurate preview of the light distribution.
I do not consider this high priority for these reasons. You can open an issue for it on github and I can flag it as "good first issue", maybe in the future someone with Python knowledge decides to implement it.

[Dade]

How does lux interpret the light distributions that are stored in that file?
Do you read the luminous flux from file or do you compute a relative value?
Are there any angular limits for IES interpretation?

To explain my application... I'd like to visualize car's headlamp beam patterns. They are usually very unsymmetrical. In some cases a resolution of 0.02 x 0.02 degrees is used describe bright/dark cut off lines.

Therefore the description locally very limited. Only +- 40deg horizontal and +10deg and -30deg vertical area.

In classic lux I had some troubles, but it was working issuing less detailed data.
Other renderers promise full IES support but they fail on that task quite often.

Old Lux and LuxCore are baking the IES goniometric profile with a "width x height" array of values (a baked map of the IES profile). The default values is 512 x 256 and it is probably too low for your application. I don't think it was possible to change the resolution in the old Lux but it is in LuxCore. Not sure if the parameters are exposed in BlendLuxCore but, in the worst case, you can save the scene in text format and edit the scene to change the resolution.
You probably need a far higher resolution than 512 x 256 to capture fine/small details.

There shouldn't be any other problem in your case but it is hard to say without doing some testing.

[B.Y.O.B.]

Not sure if the parameters are exposed in BlendLuxCore

They were not, they are now (next alpha release).