AGI32 and Relux? What are they?

[gus_webb]

Hey everybody-

 

So... a guy here in the office came across two lighting programs that I know nothing about, and it seems maybe I should. I've looked at both of their websites, and I still don't quite get what they are, or what they specifically do.

Anybody know these two?

Thanks!

[STRAT]

i've never heard of them either until you mentioned them. but i'm sure theres many others just like these aswell he havent heard of. they lighting analasis s/w.

 

and by judging from the galleries in both sights you certainly aint missing anything

[gus_webb]

lol... yeah, I can't say I'm blown away by the renderings, that's for sure. But the real-world luminaire data, specifically with Relux, looks somewhat interesting.

[Devin Johnston]

There lighting analysis programs that are designed to give you foot candles or lumen readouts and also double as rendering engines. Basically there like Lightscape.

[candlegirl]

i've never heard of them either until you mentioned them. but i'm sure theres many others just like these aswell he havent heard of. they lighting analasis s/w.

 

and by judging from the galleries in both sights you certainly aint missing anything

 

wow, that was less than professional...Given your position as moderator, I'm sure you could have found a more tactful way to make your statement

 

AGI32 and Relux are photometrically accurate lighting software programs. They predict lighting metrics using physically based simulations (calculated by radiosity and ray tracing algorithms) and IES photometric files.

 

disclaimer - I work for Lighting Analysts which develops AGI32 -

 

Lighting practitioners (engineers, lighting consultants, manufacturers, etc.) use these packages to evaluate real world luminaire performance and daylighting performance using renderings and point by point evaluations such as illuminance, luminance, glare metrics, daylight factor, roadway metrics, etc.

 

Yes, these software programs may not produce the pretty picture images you see on this site, but they do produce photometrically accurate imagery and quantifications (as close as computer algorithms, hardware limitations and luminaire testing allow).

 

And- the visualization capabilities, at least for our software, improve with each release, so photorealistic imagery is attainable for those who want it. However, the intent of these software programs is to produce accurate lighting simulations, not fakiosity imagery.

 

I am happy to answer any technical questions you have about AGI32, capabilities, etc.

 

You are entitled to your opinion, as is everyone else. However, I think that before you discredit other software packages, you should, at least, find out what it does and not just comment off the top of your head, or else resign your position as moderator.

 

AGI32 product information can be found here:

http://www.agi32.com/Products/AGI32/agi32.htm

 

and luminaire photometric data formatted for AGI32 can be found here (but you can use any manufacturer photometric data in IES format)

http://www.agi32.com/Extra/photometrics.htm

 

Leora Radetsky

Training and Technical Marketing

Lighting Analysts, Inc.

[Devin Johnston]

Wow, that was very defensive of you given that STRAT doesn’t work for you and he has the right to give his opinion about what ever he wants. Buy the way he's right, the render engine is definitely not what I would consider top of the line as far as image quality is concerned, which is probably what gus_webb is going to use it for.

[candlegirl]

There lighting analysis programs that are designed to give you foot candles or lumen readouts and also double as rendering engines. Basically there like Lightscape.

Hi Maxer

Yes, in many ways these programs are similar to Lightscape, but with some differences as well.

They provide their own modeling tools so you don't have to have 3rd party modeling software. AGI32 can also import 3D CAD models in DXF/DWG format so 3rd party software can be used (AutoCAD, SolidWorks, Microstation, etc.)

The metrics AGI32 calculates exceeds what Lightscape can do, and the daylighting calculations more accurate - like Radiance, if you have heard of this software package.

 

There are some rendering features in Lightscape (and other software packages too) that we can't replicate, at this time. Our software is evolving and we have many goals and ideas to achieve.

 

Leora Radetsky

Lighting Analysts, Inc.

http://www.agi32.com

[Jeff Mottle]

Thanks for your post Leora. Having been a Lightscape user since 1996 I have always liked the idea of getting "real" results from a rendering engine. Although I only just skimmed the surface of Lightscape's lighting analysis tools (admitedly quite limited), when I did use it, I'm really curious to know your opinion on the various applications out there. Obviously AIG32, Radiance, Inspirer, Relux etc. How do they all compare? I never worked as a lighting designer, so I'm curious to hear your thoughts. Do they all use the same core technology and the difference between them is the tools sets they come with?

[candlegirl]

Wow, that was very defensive of you given that STRAT doesn’t work for you and he has the right to give his opinion about what ever he wants. Buy the way he's right, the render engine is definitely not what I would consider top of the line as far as image quality is concerned, which is probably what gus_webb is going to use it for.

You're right; STRAT is entitled to his own opinion. However, I expected STRAT, as a moderator, to moderate this discussion and not just "blow off" a software program he himself claims to know nothing about.

 

And yes, I am defensive about our software. I follow this board regularly and want to stand up to any comments I feel are mistaken. That's all. If anything, I attempted to respond as professionally as possible. It’s a very small world, and integrity is quite important to me.

 

We can discuss top of the line renderers if you want. I agree that the imagery output by Vray, Maya, VIZ, Maxwell, etc. is beautiful and amazing. I am always stunned by the level of imagery created by many users who frequent these boards. However, at times the methods applied to lead to these results are not physically based, they are non GI (and fictitious light source) based.

 

Our software will not allow you to create fakiosity calculations and we do not intend to compete with these renderers. That's all I'm really trying to say.

 

I guess I am just looking for more open discussion. Thanks

 

Leora

[Devin Johnston]

We can discuss top of the line renderers if you want. I agree that the imagery output by Vray, Maya, VIZ, Maxwell, etc. is beautiful and amazing. I am always stunned by the level of imagery created by many users who frequent these boards. However, at times the methods applied to lead to these results are not physically based, they are non GI (and fictitious light source) based.

Leora

 

I agree with what you said with the exception of Maxwell, out of all the render engines you mentioned Maxwell is the only one that produces physically accurate results. In fact they are the only engine that uses this point in their advertising and this is what their entire engine is based upon.

[candlegirl]

Thanks for your post Leora. Having been a Lightscape user since 1996 I have always liked the idea of getting "real" results from a rendering engine. Although I only just skimmed the surface of Lightscape's lighting analysis tools (admitedly quite limited), when I did use it, I'm really curious to know your opinion on the various applications out there. Obviously AIG32, Radiance, Inspirer, Relux etc. How do they all compare? I never worked as a lighting designer, so I'm curious to hear your thoughts. Do they all use the same core technology and the difference between them is the tools sets they come with?

Hi Jeff

First - I love this website and visit it almost daily to see what's new and be inspired. I’ve never posted before today, but felt the need to jump in and speak.

 

AGI32 is lighting design software. We are an independent lighting software developer in a market with other lighting software packages such as Radiance, Lumen Designer, Lumen Micro and manufacturer based packages such as Luxicon, Visual, Relux, Dialux, and others.

 

Our primary interest is lighting analysis – both daylighting and electric lighting. The packages I listed above, with the exception of Radiance, are radiosity driven software. Some only do electric lighting analysis; others are able to computer daylighting contribution as well. The differences between them have to do with modeling abilities and available metrics. For calculations (that is to say point by point analysis), we are “stuck” with the limitations of radiosity (mesh implementation), lambertian surfaces (matte surfaces) and far field photometric testing assumptions (luminaires are tested as point sources and must be modeled as such, using an array of distributed sources if needed).

 

We (lighting software) are also limited to photopic testing and simulation – meaning we apply the photopic luminous efficiency function, which allows us to approximate the perceived brightness of the measured light. That is a completely different ball of wax, one of great interest in our community.

The core technology of these programs is similar; the calculation engine behind each one may have different optimizations. We all use IESNA and CIE metrics for our lighting analysis metrics.

 

Some of these programs provide hybrid ray tracing algorithms as well that allow you to introduce specular surface attributes for visualization (and false color analysis) purposes.

 

Radiance is different because it uses ray tracing algorithms to compute its metrics which allows it some additional functionality (such as caustics) that the others don’t have. Radiosity is employed in commercial lighting software because it is faster and view independent.

 

I can’t speak technically about Inspirer because I am only familiar with it in passing. I am more familiar with Lightscape and can speak to the differences between it and AGI32 if desired.

 

Leora

[candlegirl]

Doesn't Maxwell also allow you to use fictitious light sources if you want?

For example, an omni light source can be applied instead of a CIE/IESNA base sun and sky model (with time/day and latitude/longitude coordinate input)? In these cases, even though the GI light transfer between surfaces may be restricted to physically based algorithms, the rendering produced would not be physically accurate, since the "source" is fake.

Please correct me if I'm wrong. My knowledge of Maxwell at this time is from their literature only, not from real world usage.

 

- leora

[Devin Johnston]

There are several things in place within Maxwell that allows you to create a physically accurate lighting setup. First is the Physical Sky which can only be set by entering Longitude and Latitude coordinates at the moment. Once this is in a Date and Time are chosen as well as the GMT offset. This setup correctly simulates the position and intensity of the sun and sky at any given time and place on earth.

 

The second piece of the Maxwell puzzle is what they call a Maxwell Emitter which gives you the ability to choose the different types of Illuminants that exist like fluorescent, incandescent and so on. After entering an intensity which is measured in Watts this material is applied to an emitter, which can be any object in the scene. In order to be absolutely accurate a model of the emitter can be created by modeling the exact fixture and bulb if necessary. When the final version of Maxwell is released it will also support IES data.

 

Lastly and maybe the most importantly is the creation of the materials which are broken up into groups. There are currently 4 generic material types (Dielectric, Metal, Plastic, & Diffuse) each have its own properties which make them physically accurate. Not only does it do full global illumination but it also automatically calculates specular effects, sub surface scattering, and caustics.

[candlegirl]

There are several things in place within Maxwell that allows you to create a physically accurate lighting setup. First is the Physical Sky which can only be set by entering Longitude and Latitude coordinates at the moment.

 

Do they allow you to choose between different sky models? (e.g. clear, partly cloudy overcast) Do you know if they provide CIE/IESNA Sky models?

 

Once this is in a Date and Time are chosen as well as the GMT offset. This setup correctly simulates the position and intensity of the sun and sky at any given time and place on earth.

 

I agree

 

The second piece of the Maxwell puzzle is what they call a Maxwell Emitter which gives you the ability to choose the different types of Illuminants that exist like fluorescent, incandescent and so on. After entering an intensity which is measured in Watts this material is applied to an emitter, which can be any object in the scene. In order to be absolutely accurate a model of the emitter can be created by modeling the exact fixture and bulb if necessary. When the final version of Maxwell is released it will also support IES data.

 

I'm glad they will support IES data. Anything else, even a generic fluorescent or incandescant is not accurate or real. There are dozens of fluorescent fixture manufacturers. Which distribution does their typical represent? direct, direct/indirect, semi indirect/direct, batwing, etc...?

 

I'm afraid this still does not answer my main question, do they allow you to specify omnis or volumetric sources with a linear or inverse square law falloff distributions? If so, they still let you "fake it" which means you can create non realistic results if you want to. This functionality, in particular, is what separates lighting design software from architectural renderers.

 

Lastly and maybe the most importantly is the creation of the materials which are broken up into groups. There are currently 4 generic material types (Dielectric, Metal, Plastic, & Diffuse) each have its own properties which make them physically accurate. Not only does it do full global illumination but it also automatically calculates specular effects, sub surface scattering, and caustics.

 

I am familiar with this part of Maxwell too. I think they allow you to specify BRDF/BTDF for the surfaces as well.

 

Thanks for following up. After all, we're all here to learn! I know I am.

 

leora

[Devin Johnston]

Do they allow you to choose between different sky models? (e.g. clear, partly cloudy overcast) Do you know if they provide CIE/IESNA Sky models?

 

leora

 

Absolutely, in Maxwell there are three settings which control the sky they are Turbidity, Ozone, and Water. By adjusting these settings you can control how overcast or clear the sky is. There are also settings to introduce Fog which can absorbed or scatter the sunlight.

 

You asked whether or not you can specify omni or volumetric light sources and if they work with a linear or inverse square law falloff distributions, the answer is yes. In reality if you have a light source the only thing that makes it omni or directional is the fixture it is incased or not incased in. With Maxwell since all the light equations are based on the physical equations of light transport all of the linear and inverse square falloff equations are already taken into account for every light type. At that point it's simply a matter of defining a housing for that light to reside in and Maxwell will then use the spectral and HDR data that each emitter material has in it to calculate the final solution. There have already been numerous tests done by the Alpha and Beta users of the software to test the accuracy of Maxwell, in each case Maxwell simulates the exact same conditions that are found in the real world.

 

I'm not trying to suggest that Maxwell is going to replace AGI32 or any other lighting analysis program. I am saying that it is fully capable of producing physically accurate lighting simulations that are not only beautiful but accurate and unbiased. Right now Maxwell can't be used as a lighting analysis tool because it has no GUI, but once version 1.0 comes out I'm pretty sure there will be some tools available that will allow you to do some limited lighting analysis.

[Jaguar lover]

Some reading about physically accurate lighting simulation software.

 

http://www.integra.jp/eng/techno_main.htm

[Devin Johnston]

Here is an article on Maxwell if your interested, it points out all of it's most important features.

 

http://www.maxwellrender.com/forum/viewtopic.php?t=5422

[STRAT]

You're right; STRAT is entitled to his own opinion. However, I expected STRAT, as a moderator, to moderate this discussion and not just "blow off" a software program he himself claims to know nothing about.

 

 

sorry guys, didnt mean to offend anyone here, my comments were aimed at the render engine after looking at the galleries, NOT at the analasys the s/w actually performs. i should have worded my statement different.

[candlegirl]

Hi Kent

I've seen these articles and comparisons. They are very interesting. I don't personally know anyone who is using Inspirer for lighting design simulations, I would be happy to hear additional comments.

 

Thanks - love your tagline

 

Leora

[candlegirl]

Thank you STRAT - that is much appreciated.

 

I had to stop lurking one day - guess now is as good a time as any.

 

I guess the real question is what do people need/expect their software programs to do? It seems to depend on which part of the market you are in (visualization/lighting) and there are a lot of crossover requirements and requests.

[candlegirl]

Here is an article on Maxwell if your interested, it points out all of it's most important features.

 

http://www.maxwellrender.com/forum/viewtopic.php?t=5422

Hi Maxer

I need to register to check it out - I appreciate the reference. I'll try to look at it this week.

[STRAT]

in my opinion LS is, was and still is the most photo realistic looking true radiosity renderer that has existed to date.

 

if AG132 could be integrated with the quality of renderer that LS used i'd buy a copy tomorrow!

[candlegirl]

There are several things in place within Maxwell that allows you to create a physically accurate lighting setup. First is the Physical Sky which can only be set by entering Longitude and Latitude coordinates at the moment. Once this is in a Date and Time are chosen as well as the GMT offset. This setup correctly simulates the position and intensity of the sun and sky at any given time and place on earth.

 

The second piece of the Maxwell puzzle is what they call a Maxwell Emitter which gives you the ability to choose the different types of Illuminants that exist like fluorescent, incandescent and so on. After entering an intensity which is measured in Watts this material is applied to an emitter, which can be any object in the scene. In order to be absolutely accurate a model of the emitter can be created by modeling the exact fixture and bulb if necessary. When the final version of Maxwell is released it will also support IES data.

 

Lastly and maybe the most importantly is the creation of the materials which are broken up into groups. There are currently 4 generic material types (Dielectric, Metal, Plastic, & Diffuse) each have its own properties which make them physically accurate. Not only does it do full global illumination but it also automatically calculates specular effects, sub surface scattering, and caustics.

Hi Maxer

I'm looking forward to checking out the final version of Maxwell once it is released. I've been following its progress for awhile.

 

Do you know how the sliders for turbidity/ozone/water correlate to the CIE/IES sky models. In other words is there a set point where the settings match an accepted (peer reviewed) mathematical model of sky types?

 

For example, a couple of years ago the CIE released the General Skies models which provide mathematical models of 15 different sky types based on luminance distributions of real skies studied over a period of two years - 5 clear, 5 partly cloudy and 5 overcast with different levels of turbidity, pollution, moisture, etc.

[Devin Johnston]

That's a great question, and no I have no idea but I'll try and find out. Since Next Limit is so focused on making Maxwell an unbiased render engine I would assume they have either used this information in some way or found a way to accurately reproduce these different conditions. I'll check back with you when I get an answer.

[candlegirl]

in my opinion LS is, was and still is the most photo realistic looking true radiosity rendererd that has existed to date.

 

if AG132 could be integrated with the quality of renderer that LS used i'd buy a copy tomorrow!

 

AGI32 already has many of the radiosity and raytrace settings LS had: adaptive subdivision, individual surface mesh control (patches and elements), color bleed control, forward and backwards ray tracing in addition to radiosity, spectral radiant emittance (luminaire gel/filter specification), tone mapping, exposure control

 

plus some improved ones: daylighting, diffuse transmission through surfaces, many more lighting analysis metrics

 

There are some nifty visualization options LS and other renderers have, which we do not. I can't argue with you.

 

You can recreate many of the "high-end" images found in Lightscape with AGI32, with some exceptions (caustics, more procedural textures, specular falloff and perhaps more that I'm missing right now).

 

It may not be an apples to apples equivalent for everyone, but for many in the lighting industry - who specify/sell/design/market electric lighting and daylighting and need software to predict/prove their lighting levels from commercial lighting equipment and/or daylighting contribution for LEED compliance or other commercial codes/standards - it is.

 

Yikes - sorry for the run on sentence.

 

Leora - who needs a spell checker desperately