Feature Request?

#1 Hey FruitBat,

I've been working on a number of lighting concepts that achieve the same results as radiosity, but are significantly faster. These include many new lighting options, lighting models, and performance options, that generally can only be found in high end modeling packages. The new options allow the same design and lighting theories used in those packages to be applied to the Lighting Pack.

Most of this "radiosity" work I've been doing is on indirect lighting, or light that appears to reflect off of one surface to illuminate another. I haven't worked on emissive surfaces like the old radiosity map2dif used, but I'll take a look at it. Just out of curiosity why is the rad map2dif dead, can't it be merged with the new map2dif?

-John

#2 This is purely from memory, but I seem to remember Matt saying (or at least implying) that radiosity is a feature to be implemented in the future of the new map2dif?

#3 Matt said:

Quote:Josh did an awesome job of providing a radiosity solution but it just isn't finished. There will be a new lightmapper at some point in the not too distant future but it is unlikely to include the code from TGERad

#4 @John: Great news, I've come to rely on the radiosity version. This is me being ignorant, but isn;t it possible to create a rectangular light source object and just have it throw light in an arc along the area? Emmissive textures would be great too.

Also john, while matts rewriting map2dif would it be possible to ask him if he can expose the objects to the mappers as entities, like the normal ones?

#5 It would be difficult for Matt to expose new entities without writing the underlying code itself, because he would need to know all of the properties to add. Map2dif's code is rather straight forward, making the addition of point and structural entities a snap (like the Lighting Pack's lighting scale entity), so I don't think adding the entities without the code would help much.

Multi-dimensional lighting sources need extremely realistic and accurate lighting models, otherwise artifacts become apparent as multiple sources are used together. For instance if your map contained one 16x4x4 emissive brush you would expect the same lighting as when using two 8x4x4 brushes or four 4x4x4 brushes end to end. Any inaccuracies in the lighting by faking the process will become very apparent very fast.

So technically when using a single block the process can easily be approximated, but realistically I think people would find it difficult to use in a real interior.

I'll try to get a handle on where map2dif is heading, also I'll check out various methods for emissive surfaces in case they don't make it into map2dif, and we'll see where things go from here.

In the meantime the difs generated with the old map2dif should be compatible with the current TGE, right? Why not continue using the old radiosity map2dif?

#6 I will use radiosity for a while, but Matts new one is a lot more reliable in terms of architecture and light leaks. Its a shame radiosity won't be pulled into it.