How accurate is Viz/Max IES Lighting?

[Bishop]

Now that Lightscape is no more in development, I heard the engine was incorporated into Viz4 Radiosity, subsequently Max5, then now Max6. I was wondering just how accurate is the lighting engine in Viz4 and Max6??

 

If I use photometrics lights with their corresponding IES web files, just how accurate is the result? The Psuedo Exposure allows for showing Lux levels in the space, just how accurate is this?

 

Also is there a way that I can get Lux contours from either Viz or Max?

 

Any help here would be great cause I've been asked to do a lighting study of a convension centre and I need to tell the client just how accurate are our renders.

[William Alexander]

Bishop,

 

It's considered accurate. Although having spoken with Engineers about this, they tend to stray away from it. Both I spoke with had bad experiences with the way they understood the softwares read out. They both do it manually, actual calcualtions and experience.

 

You may consider for the sake of knowing for sure, having a Engineer do some calcualtions for your convention center project and check them agianst the psuedo calcs in viz/max/LS. And of course post the results!!!! Or you could of course find the measuring equipment and do it yourself! lol

 

Not much help, I know, but this is the obstacle to verifing if it's actually accruate and accepted. ihmo

 

rgrds

WDA

[CAD]

Also is there a way that I can get Lux contours from either Viz or Max?

I think Viz and Max were only designed for visual results and not for lighting calculation results.

 

I know of a lighting software that provides "accurate" lighting calculations from a given space which is also created in 3D. The whole software can be downloaded and its free!

 

You'll just have to work your way out in getting good visual results but it is, however, very accurate when it comes to lighting calculation results.

 

Just my two cents,

Robin

[Bishop]

I have spoken to some Lighting Consultants, they seem to accept that lightscape get pretty accurate results. I know that the lightscape engine was not completely tranfered to Viz, it was only partially adapted. So is the new engine accurate or not?

 

I understand that 3D Studio is a visual software and most lighting software is accurate but it is difficult to generate visually stunning images.

 

So where is the middle ground? Lightscape was one, heard some good things about Radiance. But just how far can we push Viz/Max in terms of lighting accuracy?

[vru]

I think Viz and Max were only designed for visual results and not for lighting calculation results.

I dont know how lighting calculations are done ...but i would like to know if the visual results are accurate??? - this could be a stupid question as I have no knowledge about it. But I would like to know if viz/max photometric lights with IES web values are used to create lighting studies - atleast visual effects?

Thanks

Vru

[John Dollus]

The radiative transfer engine from Lightscape was NOT transferrred to Max/Viz. About the only thing they have in common as far as quantitative lighting analysis in concerned is the fact that they both use photometric files.

 

There is not a way to display iso contours in Max/Viz (nor was there in Lightscape). A point analysis with false color is the closest thing.

 

While Max may be 'close enough', it is not an analytical tool such as Lumen Micro, AGI, Radiance or any of the others.

 

If you are expected to produce engineering data from the results, I would recommend a tool other than Max. In court, there are too many holes in the Max method to cover your butt if the client is not satisfied with the levels after the project is constructed.

[Ernest Burden III]

I've been asked to do a lighting study of a convension centre and I need to tell the client just how accurate are our renders.

I suggest you pick one--lighting analysis or artist's rendering. Trying to do both will likely lead to doing both poorly.

 

I provide renderings. I pride myself on accuracy, but it is never a guarentee (especially since clients often want The Big Lie) and no-one is expecting to build from my product. For me, accuracy is more about dimensional than lighting.

 

I play loose with lighting values. I use Lightscape but will put in a 10K watt bulb if it helps light a picture the way I want it to look. Remember, most lighting designs aim for fairly even illumination, which, when done correctly, makes for a boring rendering.

 

So if you are doing renderings, do renderings. Do not promise scientific documents if you are not a specialist in that field, you are simply getting in over your head. At least promise nothing beyond a reasonable rendition of the lighting scheme.

[Bishop]

What then would the "Lighting Analysis" be for?

This is an exerpt of the Help file on Lighting Analysis:

 

Lighting Analysis

To query light levels, analyze the data, and produce reports, use the Lighting Analysis dialog. This dialog provides rendering data on material reflectance, transmittance, and luminance.

 

For example, a lighting engineer might need to know if light fixtures in a scene provide an even level of illumination on the walls of a building. The engineer uses the Lighting Analysis dialog after placing the lights in the ceiling and processing radiosity. The engineer inspects the light levels and material reflectance in the scene and then adjusts the brightness of lights, changes units, or reduces material reflectance.

Now I am a little confused. It 'claims' in the Help file that we can get the luminance value from the solution. Just how accurate is it? Since an engineer would use it....

[Bishop]

Anyone with comments on the Help file above?

 

If Autodesk/Discreet is not continuing development on Lightscape and claims that the technology is incorporated into Viz then Max, buy since the incoporation was not complete, as John outlined, are they abandoning completely the accurate lighting simulations?

 

The lighting engineer that I was talking to was confident with Lightscape results, which gives it some creditability. But was not convinced with Viz or Max.

 

Any comments on this?

[abicalho]

I posted this in the MAX webboard last week.

 

I've done all the Photometric Computations by hand to make sure that the Lighting Analysis is using the correct formula to compute values:

 

Intensity at Point = (Original Light Intensity)/Distance^2 * Cos (incidence angle)

 

Do a simple test:

 

Create a 200x200 plane with 80 segments

Place a 1500 cd light at its center (use Align if needed)

Move it 100 units in Z

Set System Units = Centimeters

Solve Radiosity to 97% quality

 

Now let's do some Math. Put everything in Meters (if you're using CD, you need to use Meters)

 

Distance between the Light and the corner of the plane = SQRT(3)

Cos Ang Incidence = SQRT(1/3)

 

Intensity at the corner of the Plane is:

 

1500 / SQRT(3)^2 * SQRT(1/3)

Result = 500*SQRT(1/2) = 288

 

Now compare it with the Lighting Analysis.

 

In MAX I am getting around 289 lx in the corner. And the Lighting Analysis tool tells me the lowest point in my object has an intensity of 276 lx, which is pretty close to 288.

 

Do the same math for other points, if you feel like it. I did it here for various points, objects, light intensities and I got precise results. Compare it with Lightscape and you'll see that both results look similar.

 

Why do they not look the same? Because Lightscape and MAX have a different way to compute Radiosity. MAX stores the Lighting Information per face, whereas Lighscape stores it per vertex. MAX is also statistical, so you'll need to solve for a higher quality if you want more precision.

 

Alexander

[Bishop]

Alex, so what you are saying is the analysis is correct but the renders are not. On a lighting design stand point, if all I want is the figures, would that be accurate? Even if the renders are not true representations of the lighting?

 

I posted this in the MAX webboard last week.

Can you post the link for this?

 

Thanks

[Bishop]

Opps, it's ok. Found it. :angerazz: :angerazz:

[abicalho]

Originally posted by Bishop:

Alex, so what you are saying is the analysis is correct but the renders are not. On a lighting design stand point, if all I want is the figures, would that be accurate? Even if the renders are not true representations of the lighting?

I never said that. I do not know where you got that from my post. The differences between LS and MAX are minimal, and usually happen when the user does not mesh the scene in MAX or sets a very fine mesh size in LS that he cannot duplicate in MAX (because MAX does not have Progressive Refinement).

 

Alexander

[Bishop]

I did it here for various points, objects, light intensities and I got precise results.

I assumed that you meant that the lighting analysis are fairly accurate.

 

Compare it with Lightscape and you'll see that both results look similar.

 

Why do they not look the same? Because Lightscape and MAX have a different way to compute Radiosity. MAX stores the Lighting Information per face, whereas Lighscape stores it per vertex.

And that they looked different because it is computed(rendered?) differently.

 

If they are different then one of it would not be a correct representation then.

Even if the renders are not true representations of the lighting?

What i was trying to establish with this post is if the Analysis were accurate, not the renders. The renders are very visual and with exposure control you can get a bright image from a dark scene at the same lighting levels.

 

Bottom line is, Is the lighting analysis of Max/Viz comparable to Lightscape? Can we use the figures computed to be close to real world?

[abicalho]

And that they looked different because it is computed(rendered?) differently.

If they are different then one of it would not be a correct representation then.

The difference is less than 5% between both... see my math results above.

 

And both have a difference of around 5% when compared with the real math formulas that should serve as the base for the calculation.

 

If 5% is not acceptable for you, then neither software is precise. If 5% is enough for you, then you can arrive to the conclusion that both are similar and produce the same results.

 

There were some problems in Lighting Analysis in previous versions of MAX and VIZ, and discreet worked in solving them to make the software more precise and acurate. Also, MAX 6 features a legend for the Pseudo Color render, which neither MAX 5 or VIZ 4 had.

 

MAX 6 also gives you MAXScript access to querying Lighting information for any point in the scene, allowing you to export this data to text files, etc.

 

MAX 6 trial can be downloaded, so if you're reluctant, give it a try and see for yourself.

 

Alexander

[Richie]

Hello

 

The other point is HDRI.

 

The Luminance Range is so big, that you can NOT

display this luminance data on a computer monitor

or prints.

 

IMHO you can use regular LS renderings with corrected luminance and contrast.

 

I checked the results from LS, and they are really really good (I mean accurate).

 

But as somebody said this before, much customers want the big lie.

 

Richie

[Bishop]

Thanks guys, especially Alex, you really cleared up some doubts I had on the matter.

 

Chan

[pburgoyne]

Alex,

 

which versions of Max and Viz were less accurate. I know that some of the errors were corrected between Viz4 and Viz4 SP2. Is Max6 more accurate than Max5? how accurate is Viz4 SP2 compared to Max5 or Max6?

 

I have used Viz4 quite succesfully to test lighting layouts in our stores, I realise it is not 100% accurate but for quick calculations and as a visual reference it works. No maths to back it up, just final constructed space.

 

thanks

 

Peter