![]() |
||||||||||||||
|
|
||||||||||||||
|
|||||||
| Notices |
| Work in Progress (WIP) Post your works in progress here to get user feedback and critiques. |
|
|
LinkBack | Thread Tools | Display Modes |
|
|
#1 (permalink) |
|
Member
Join Date: May 2003
Location: Starkville, Mississippi
Posts: 56
Name: Vrushali Palamwar |
Hi
I have been trying to get good results with radiosity for a long time now. I am getting huge light leaks. I have all the edges of wall and ceiling meeting. I have used closed spline, extruded it and converted into mesh for my walls, ceiling and floor. I thought regathering will take care of the leaks but that is not happening. There are other problems, like dark ceiling and noise. The cloth texture has artifacts even with tighter mesh and regathering. Anti-aliasing is not right at ventilator frame and some wall-ceiling edges. I have also noticed that the bump mapping looks different in different views. So I cannot use the same radiosity solution. I have to reset the bump and calculate the solution again. Attached are some views, wireframe view and radiosity settings. I am using IES Sun and IES skylight. Please advice. Thanks Vrushali Last edited by vru; October 25th, 2004 at 12:35 PM. |
|
|
|
|
|
#2 (permalink) | |
|
Senior Member
Join Date: Jul 2004
Location: Dubai
Age: 38
Posts: 132
Name: Dennis Munar |
Quote:
Well, Im having this same problem with radiosity each time i use daylighting which causes light leaks. Im joining you for this ordeal, hope someone with a good heart can help us. Asking for help, Dennis |
|
|
|
|
|
|
#3 (permalink) |
|
Senior Member
Join Date: Aug 2003
Location: Pittsburgh, PA
Age: 31
Posts: 142
Name: Tom Corbett |
Radiosity in MAX/VIZ takes the value of light at each vertex and blends it across the face. In these examples, you probably have your wall extending to the edge or beyond the other wall, causing the light leak.
3 ways to control this: 1. Crank up the subdivision count in Radiosity. Instead of 3 feet, use 1 foot. This will also increase your calculation and render time, but you will get a better spread. You can also apply the "subdivide" modifier only on specific objects to acheive the same effect. 2. Use the "slice" modifier to "refine mesh", and place the slices within that wall intersection. Of course, now no light will be hitting that vertex, and you will see an opposite effect - kind of a "dark leak". 3. Build your walls & ceiling more precisely. Also consider putting large box overtop of your model, to act as kind of a sunshade for everything that is not a window. |
|
|
|
|
|
#4 (permalink) |
|
Member
Join Date: May 2003
Location: Starkville, Mississippi
Posts: 56
Name: Vrushali Palamwar |
Thanks for the suggestions.
I do think there is a problem with the modelling of walls and ceiling and how they meet. I have gone through some other posts here talking about it. But I did not understand what is the right way to model. Could you please see the attached wireframe image of my model and comment on the modelling issue? Thanks a lot. Vrushali |
|
|
|
| Sponsored Links |
|
|
#6 (permalink) |
|
Senior Member
Join Date: Dec 2003
Location: Jacksonville
Age: 29
Posts: 472
Name: Jason Matthews |
My only thought would be to set the filtering up to like 3 instead of 1. Oh, by the way, you don't have to run your solution up so high. You can get good results if it is set to 90 or so.
__________________
Jason Matthews |
|
|
|
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
| Display Modes | |
|
|
|
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| How to render indirect light only | Peppi | 3ds Max | 3 | July 12th, 2004 02:33 PM |
| In Progress, But Light leaks are the death of me! | Jet | Work in Progress (WIP) | 2 | April 27th, 2004 08:43 PM |
| 3dviz volume light beams through glass? | chadsmith | Autodesk VIZ | 2 | January 26th, 2003 07:32 PM |
| Radiosity: Light leeks. | Dibbers | Autodesk VIZ | 5 | January 26th, 2003 06:34 PM |
| LIGHT TRACER &s RADIOSITY | Cesar R | Autodesk VIZ | 8 | December 17th, 2002 10:21 AM |