lightmap tool
Author
Another question for today, does somebody know a handy free tool to create lightmaps? It would be even nicer if it would support advanced stuff like radiosity. What I would like to have is a little program where you can put in a 3ds (or other format) model, setup some lights, and then let a bitmap(s) roll out and eventually a list of lightmap texture coordinates. Does something like that exists?
Author
Thanks for that link! But that brings me to the next question. This program can import ASE and OCT files. I never heard of OCT. I know ASE a little bit but I don't have software that can read/write those files.
I'm using Lightwave and a friend of my uses 3D Max. So does someone know a lwo/3ds/max to ASE convertor? And is there something that can write this OCT file back to one of these files?
BTW, I wonder where the lights stay. *.LWO doesn't store lights as far as I know, not sure about *.3ds and *.max
Greetings,
Rick
I'm using Lightwave and a friend of my uses 3D Max. So does someone know a lwo/3ds/max to ASE convertor? And is there something that can write this OCT file back to one of these files?
BTW, I wonder where the lights stay. *.LWO doesn't store lights as far as I know, not sure about *.3ds and *.max
Greetings,
Rick
FSRad is open-source, so if you need a specific import format, I'm assuming one could write their own.
I haven't used FSRad, but for people who have, do you need to load a high-poly model to get
a nice anti-aliased lightmap, or will you get a jagged lightmap if your model is low-poly?
That is to say, a model with more vertices gives you a better lightmap in FSRad?
I haven't used FSRad, but for people who have, do you need to load a high-poly model to get
a nice anti-aliased lightmap, or will you get a jagged lightmap if your model is low-poly?
That is to say, a model with more vertices gives you a better lightmap in FSRad?
So, when using radiosity, lightmap quality is independent of vertex or polygon count?
I guess I can believe that. :)
I guess I can believe that. :)
Quote:Original post by JakeMExactly. Lightmap quality (artefacts like Mach Bands and at triangle edges aside) is directly proportional to tesselation detail of the scene. How much is the scene tesselated depends on particular implementation of radiosity and its parameters, mainly whether there is Uniform Subdivision or Adaptive subdivision of the patches. The first one is worse in quality, but faster to implement. Quality is lowe especially on shadow boundaries. adaptive subdivision compares differences in Energy between neighbouring patches and if it`s over some threshold, subdivides respective patches until the difference is lower than some threshold value.
So, when using radiosity, lightmap quality is independent of vertex or polygon count?
I guess I can believe that. :)
Of course, there are more issues affecting the lightmap quality, especially how much energy has been redistributed throughout the scene.
But generally, you can`t affect the lightmap quality by tesselation of original mesh, provided we`re talking about some full-blown radiosity implementation.
I see nothing about radiosity or global illumination on that site.
They don't even describe what their lightmapper does, technically.
I would guess it's doing direct illumination only.
Aren't direct illumination lightmappers generally slow, and do not give very good results when you work with low poly models? And isn't it nail-bitingly slow when trying to render soft shadows using direct illumination? Even on low resolution lightmaps, quality is generally not very good either.
They don't even describe what their lightmapper does, technically.
I would guess it's doing direct illumination only.
Aren't direct illumination lightmappers generally slow, and do not give very good results when you work with low poly models? And isn't it nail-bitingly slow when trying to render soft shadows using direct illumination? Even on low resolution lightmaps, quality is generally not very good either.
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
