Maxwell Render | Official public forum

I'm not sure that this has been mentioned but this seems like a no-brainer for a renderer that is mostly used for Archviz. It's been in VRay and Max for a long time:

http://www.cgarena.com/freestuff/tutori ... ieslights/

I agree. As LEED comes into all our designs now the ablility to model correct lighting is critical.

I think i said it before and i'm gonna say it again, IES is highly unlikely to come for Maxwell. I have some serious doubts about that, but you're ofcourse free to wish for what you want.

/ Max

I need IES really bad. Even clients now asking if my software can handle ies because they are used to vray renders that supports it.
Please make IES in the next version

this just proves the point: clients don't care how the renderer is designed but when you tell them that it can't do what they expect any other archviz renderer to, then you have a big problem. Trying to convince them that Maxwell looks better than VRay isn't going to sell considering they spent a ton of money hotel lighting that looks a particular way.

rboaz wrote:now the ablility to model correct lighting is critical.

That's just it! You have to model the correct lighting.

I think the programmers should put some time into figuring out how to overcome this.

There is an inherent contradiction between how real and accurate Maxwell can render light, but it can not (easily) distribute light to match a lamp/fixture combination.

Yes, if I put the time into modeling the lamp and fixture I can simulate (or re-create) the real lighting. But that is a lot of work! And the fixture data is not always readily available.

If Maxwell were not SO good at producing beautiful, accurate images this would not be so important (especially for those of us whose clients ask for it).

So, yes, I vote for a "smart" emitter that distribute light based on a IES file.

Gary

+1 for IES support.

I recently had to model a bunch of l.e.d. light fixtures

The tiniest change in the shape/size/position of the emitter and its attendant reflectors/lenses plus color and intensity had huge effects on the light output/shape/color etc. It's practically impossible to make a completely accurate model.

In the end I just get somewhere in the neighborhood, after a lot of back and forth with the client, and finally they say "that's close enough".

It's especially disheartening when I open the same model file in 3ds max and switch to mental ray, assign an IES profile and bam! done in two minutes. I really don't want to use mental ray. (really!) But the next time a rendering job requires specific emitter output, and there are IES files available...

edit: I'd love to hear some feedback from NL about this. I suspect that the whole IES way of doing things just isn't compatible with Maxwell's way.

I looked into this IES stuff a while ago, and it shouldn't really be too difficult to do in Maxwell:

IES define the amount of light in all directions from a point.
So...

1) Convert IES file to spherical transparency map
2) Apply to sphere, hidden from the Maxwell camera
3) place tiny emitter inside the sphere

So unless I'm missing something fundamental, I don't see why this should not work... Obviously there's the limits of reality of the IES data - no matter what the distance is between a surface and the lighsource, the pattern (caustics) will always be the same!

Here's a spec, but take note there's a European and an American version of these files... http://www.kxcad.net/autodesk/3ds_max/A ... ormat.html

Hi.

how to

1) Convert IES file to spherical transparency map

??

Best

Jesper

zak wrote:how to

1) Convert IES file to spherical transparency map

??

This would be the part that requires a little conversion program. I have been meaning to do it for ages, but still never seem to get around to doing it. It would map the points specified in the IES onto a bitmap and interpolate between them to fill the rest of the image. It really can't be a big deal I don't think...

i think a trancparency with a "normal" emitter wouldn't work because you seperate the vector of emitted rays and their intensity into two different components; it would be more the diffuse effect of a lampshade as non of the caustic effects in the light's reflectors are simulated that way.

I agree totally about the importance of this. This is Maxwell's biggest issue and I hope they keep an open mind and take it seriously.

There are two ways to go; Maxwell or all the programs that utilize IES files (these include Viz + actual lighting analysis programs). For everyone except visualization-only pros, the IES functionality is critical for workflow and for accuracy. If Maxwell doesn't implement this, it will simply lose out in time to competing programs that will eventually mimic Maxwell's advantages (multilight, etc), but also be IES-compatible.

IES stands for Illumination Engineering Society (NA, for North America). These files are a compact and realistic way to represent photometric properties of real-world lighting fixtures; they are produced so that you don't have to go to the ridiculous step of modeling actual fixtures (which is virtually guaranteed to be error-prone). These files have standards to which lighting manufacturers adhere, so as to realistically represent their fixture data in a uniform way. Better yet, most manufacturers of actual lights produce these files so designers can use them in lighting analysis programs; there are thousands of these files out there, just go visit your favorite lighting fixture company. This is the professional format used by lighting designers, architects, and the entire design/construction industry. This is the standard to which Maxwell ought to adhere, not vice-versa...that is, if it is to be considered "physically accurate." Certainly from a workflow standpoint, without IES files there is a great deal lost in terms of accuracy and productivity.

Maybe instead of "The Light Simulator" a more appropriate moniker for Maxwell should be "The Really Cool Light Renderer." Simulation implies accuracy, and its hard to imagine being accurate without accurate light distribution.

I did a quick search and it seems there's a simple way to convert ies to bitmaps, IES Viewer has an internal renderer to preview ies files :
http://www.cgarena.com/freestuff/tutori ... ieslights/

Wish solved?