[gmenzel]

I want to simulate the illumination of a surface with a spotlight with known properties (beam spread angle and intensity in Lumen; sadly, no IES data). From that, I want to be able to roughly judge the intensity of surface illumination when photographed with certain camera parameters. My goal is to find out what type/strength of spotlight I will need for a real-world scenario.

I know, Maxwell Render is not actually intended to be used for such a purpose. But, in principle, it should be possible, right? My question is: What would be a practical way to do this in MR?

Emitters in MR do not have a directionality parameter (does anyone know, why not?) and seem to emit light in all directions equally (hemisphere around face normal). How would I go about focusing emitted light into a beam (other than actually emulating the optics of a spotlight). The 3DS Max plugin offers MxLight with a ‘spot’ mode. But there, by the way the beam is formed (simple cone-shade), it seems to be difficult to relate the stated intensity to any real-world value.

Any hint will be greatly appreciated.

[tom]

Do you think this could be useful?
http://www.maxwellrender.com/forum/view ... 17#p313717

[RichG]

Hi Tom

I created something similar to that tube spotlight but I made the disc circular to avoid intersecting the walls (having heard a long time ago that you should avoid emitters intersecting with other geometry. If you mask the emitter to make a disc shape, can the geometry intersect with no adverse effects? Obviously a simple 2 triangle plane is better than a disc for render efficiency.

[tom]

Intersection is not a problem when masked. Because masked out parts will not be emitting light and you can also texturize your lamp if needed. Polycount of emitter is not a big issue as in the past but it's always good not engaging the engine with unnecessary detail. Masking will also slow it down a little so a few poly flat geometry could be the fastest. In this perspective you could consider a low poly disc.

[gmenzel]

Do you think this could be useful?
http://www.maxwellrender.com/forum/view ... 17#p313717

Thanks, Tom. Actually, I was hoping to be able to avoid having to build reflector geometry like this. A directional emitter seem to be such a simple concept. I keep wondering why MR doesn't offer this functionality, which seems to be perfectly consistent with an unbiased approach.

Anyway, if a parabolic reflector really is the only way currently available, I have to at least make sure that the result still has some significance. If I place a spherical emitter within such a reflector with a perfectly reflecting material, does the total flux in lumen through the opening of the reflector equal the flux as set in the emitter material? Theoretically, it should.

[tom]

Well, if you don't have IES/EULUMDAT data for your spotlight, you should model it and I think this is fair enough. And yes, the flux will be the same only if you coat the interior of the spot cone with 100% reflective material.

Additionally, here's an editor suggested by Brian. I'm not quite into IES but you may find it interesting...
http://www.maxwellrender.com/forum/view ... 37#p336037

[feynman]

We actually uses MR to test light fixtures that are yet to be built. You get reliable results, but you must 1. use the lamp's IES data (practically all LED, CFL, FL manufacturers provide these for download) and you must 2. use correct reflector geometry (many reflector manufacturers provide these for download in IGS file format or at least as 1:1 2D DWG file). You can model your own reflector - great to actually test if it works and eliminates banding at the perimeter of the light cone. Without a lamp's IES data, your light cannot be simulated correctly and you would be back in the old days of building a physical prototype of your light fixture.

[Bubbaloo]

you must 1. use the lamp's IES data (practically all LED, CFL, FL manufacturers provide these for download) and you must 2. use correct reflector geometry

I thought the purpose of using IES was to NOT need a reflector. Modelling an accurate reflector defeats the purpose of using IES in the first place, because IES simulates the light distribution of the fixture. If you have an accurate reflector, all you need is a regular emitter in the correct location and size.