Community
VRED Forum
Welcome to Autodesk’s VRED Forums. Share your knowledge, ask questions, and explore popular VRED topics.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Can VRED illuminate a PMMA Light Pipe properly?

14 REPLIES 14
SOLVED
Reply
Message 1 of 15
Anonymous
1331 Views, 14 Replies

Can VRED illuminate a PMMA Light Pipe properly?

I am evaluating VRED for its' potential for use by my company, an automotive lighting supplier. We are presently using Modo V12 for this purpose, and I have been testing to see if VRED can improve on our usual results.

I am having no luck lighting up the type of light guide known a light pipe. A roughly tubular, faceted item which projects light either forward or aft of the vehicle. These have grown to be quite common in various automotive headlamps and tail lamps, and we need a way to render them with a reasonably good degree of accuracy. We use SPEOS to design our optical prescription (facets) to create the required light coverage and pattern, and our renderings need to reflect what we see in our prototype and production assemblies.

I have what started out as a 30 day trial of VRED Professional and so far, I have not been able to get light pipes lit. I followed the process outlined in this document, "LightSimulationwithVRED_PascalSeifert" and in the Autodesk video on Caustics without success.

I am attaching an example of what we would expect a lit light pipe to look like.

Can VRED do this, or improve upon it?

Thanks!!

 

Pipe_one.jpg

 

14 REPLIES 14
Message 2 of 15
Anonymous
in reply to: Anonymous

I should mention that there are LEDs at each end of this light pipe. They are defined by a photometric IES file, but they could as easily be created as spot lights with the appropriate angle of coverage. I use  the face of the LED in the CAD model to set up the location and vector correctly.

Message 3 of 15
michael_nikelsky
in reply to: Anonymous

Hi,

this should be possible within certain limits. The VRED raytracer is recursive at the moment, so it has a limit in the number of bounces are ray can do, which currently is 1024 bounces. This should be enough for most but not all cases. You will need to manually set this limit in the render settings and you should also switch to Full global illumination mode since there are some quite high limits active in the other rendering modes.

 

A spotlight however will not work since it does not exist in reality. You need to have a light with an area like a disk or rectangular light. You can assign an IES profile to those lights as well but you must be aware that VRED deals with it by placing the profile on every point on the light surface since IES profiles strictly speaking are only valid for point lights.

 

However, if you could share an example scene and a reference image I am willing to try to set this up since I never had the chance to do this kind of visualization myself.

 

Kind regards

Michael

 

 



Michael Nikelsky
Sr. Principal Engineer
Message 4 of 15
seiferp
in reply to: Anonymous

Attached a small example. I´m using Photonmapping in addition for creating caustics on the wall and ground. Most important for the 3D model are the small fillets on the prism´s of the backside of the light guide. This is actually where light starts reflecting/collecting.

 

Hope this helps

Pascal

Message 5 of 15
Anonymous
in reply to: seiferp

Hello,

Thanks for your replies!

Here is the 3d data for the light pipe that is shown in the originally attached rendering. The geometry for the LEDs is included. This file is based on an FBX translation from the original MODO v12 model.

1024 bounces for rays should be adequate for most light pipes we use.

Gordon

Message 6 of 15
Anonymous
in reply to: Anonymous

So, should I gather that VRED Pro is not able to do this?

Message 7 of 15
michael_nikelsky
in reply to: Anonymous

I did a quick test render with the lightpipe you provided. I had to guess all the values so don´t expect it to be accurate.

Lightpipe RendertestLightpipe Rendertest

I also attached the vpb file so you can take a look at it.

 

Kind regards

Michael



Michael Nikelsky
Sr. Principal Engineer
Message 8 of 15
Anonymous
in reply to: michael_nikelsky

Thank you very much! This is much more like what I've been looking for. I am going through the file you attached to see how this was done. I want to apply these techniques to some of our other light pipes to show to my management. I see that caustics are not turned on. Or am I missing something here?

Message 9 of 15
michael_nikelsky
in reply to: Anonymous

Hi,

 

Caustics are not needed for this type of visualization, this is pure path tracing. You will need caustics when you want to visualize the illumination that is caused by the light pipe on other objects.

 

Kind regards

Michael



Michael Nikelsky
Sr. Principal Engineer
Message 10 of 15
Anonymous
in reply to: michael_nikelsky

So, you are using the LED geometry with the "Default" material mapped to them as light sources. And a normal Acrylic material on the light pipe? Put 1024 bounces on in render settings, and it lights the pipe?

Thanks!

Gordon

 

Message 11 of 15
michael_nikelsky
in reply to: Anonymous

Yes, that is the basic setup. Of course, if you want to use an IES profile you will need to create a disk or rectangle light instead and place them accordingly. But the rest of the setup should be identical.

 

Kind regards

Michael



Michael Nikelsky
Sr. Principal Engineer
Message 12 of 15
seiferp
in reply to: michael_nikelsky

A quick one with caustics using the following settings... Would be really interesting to see how the real lightpipe behaves Heart

TestLightpipe.pngPhotonmapping Settings.png

Message 13 of 15
Anonymous
in reply to: seiferp

Thanks! I have been able to more or less replicate what you have done in order to light the light pipe. The thing for me to do now is apply this approach to a couple of our newer light pipe projects for which I can get accurate photometric results and photos of prototypes. That will tell me how close we are getting to reality with the renderings. 

Gordon

 

Message 14 of 15
michael_nikelsky
in reply to: Anonymous

Hi,

I would be interested in your results as well since we never had the chance to do acomparison to a real world light pipe. Also if you come up with specific issues, just contact us again.

 

Kind regards

Michael



Michael Nikelsky
Sr. Principal Engineer
Message 15 of 15
Anonymous
in reply to: michael_nikelsky

Thanks for your help! I've only got a week left on my 30 day trial license, so I will have to fit more testing in along with production renderings that I am on deadline for.

Gordon

 

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report