I've been looking at the Irrlicht Demo and Q3. And it seems to me that Irrlicht uses a different way to combine lightmap and texture.
As I checked in Irrlicht code, Irrlicht modulates texture and lightmap and multiplies by 4 after so general lighting is about the same level as Q3.
However in Q3 with some tests we found out that Q3 does not actually the same:
A textured wall with a 0.5 intensity lightmap looks exactly the same in Q3 as the wall with a 1 intensity lightmap.
A white textured wall with a lightmap that linear goes from 0 to 1 actually reaches full white in Q3 at a lightmap intensity of about 0.35.
So it seems that Q3 preprocesses the lightmaps (multiply by 2 or some Gamma ramp ??) and applies some Gamma ramp at the rendered screen.
Maybe Irrlicht can apply also a likewise two fased lightmodel. Like multiplying the lightmaps internal with 2, and use MODULATE2X instead of MODULATE4X. (The later also fixes problems with some older videocards which don't support MODULATE4X, like I have with my laptop: the lightmaps aren't used at all)[/img]
Lightmap problem
Re: Lightmap problem
That's right. I didn't want to use a Gamma ramp and because Modulating the textures 2X was too dark, I simply chose modulate4X. But I could add modulate2X too, as new material type, if that's better for you. You also could try to increase the brighness of lightmap pixels manually then and try some combination with the new materials when they are available.Mars wrote: So it seems that Q3 preprocesses the lightmaps (multiply by 2 or some Gamma ramp ??) and applies some Gamma ramp at the rendered screen.