Better Emission Model

[WAS]

Is this even possible? Terragen's emission is just plainly awful. Even in Path Tracing where it doesn't even make sense compared to any other path tracing engine and emission. And whats worse here is we get the same results of caching in standard renderer, except no ability to improve the cache for slightly better results, as PT doesn't use it.

For example, take this floating ball scene I was going to use to test emission. The scene is too stupid to even realize the ball of light is floating above two cubes, and acts like it's hugging the ground in the background with a huge lit circle on the ground, far away from the scene. Lol What the heck.

Like literally, the 3D Preview, and Render think the glowing cube is far away. Lol the cubes are the ones that are far away from the sphere. xD

Additionally, it's weird i need to get to 500 emission just for nothing to work, yet in blender emission of 5 is enough to light the whole scene, accurately, in path tracing. Third screenshot is blender with same sort of setup, emission of only 5 vs TG at 500.

I feel the difference between other DCC, or even a game engine today in 2022 really calls for a rewrite. Please please please. :O

[Matt]

In the TGD, your sphere isn't hovering, it's sitting on the ground (and that's why the render looks like that). Have you tried doing a 1:1 comparison with the Blender scene? I'm pretty sure TG's path tracer will handle this emission reasonably correctly if you use the same geometry and similar materials.

[WAS]

I could export it. But I set the sphere to Y 5, so not sure how it would be on the ground let alone way off in the distance. Did t copy any coordinates or physically move the sphere.

There is a environment light in blenders cycles I can't figure out. Where you can see everything not lit by the sphere. Not sure what that's about. No lights and no environment HDRI.

[Matt]

But I set the sphere to Y 5, so not sure how it would be on the ground let alone way off in the distance.

Just look at the file you uploaded. It's at 0,2,0 with a scale of 2. This puts it on the ground, away from the cubes.

[Matt]

Thanks for posting the TGD, by the way. It helped to see where the problem was.

Raising the sphere up and moving it closer to the cubes (which are at about -20 in Z), I was able to get some light on the cubes, and luminosity of 5 was enough to see the effect.

[WAS]

Ok so I exported the scene from blender.

Blender gets a better result than TG (at max samples and default MPD AA) in only 2 seconds (at 1080p vs 800x400) vs 1:04 in TG. That's pretty shocking. But the results are definitely similar. My error. I wish this was faster.

Strangely the pyramid looks transparent though.

Attached gathered TG project.

[WAS]

PS Blender IS rendering in CPU only. I have no compatible video card.