What would make Source and the Source SDK better?

I am working on getting a lot of 3D Engines up on this page. 6 pictures and 2 videos if possible :D. Feel free to discuss!

Thread Music

Source Engine: Obviously, click to enlarge.

Caution: Loud on ears[media]http://www.youtube.com/watch?v=-kvxHhKjZ60[/media]

Unreal Engine: Obviously, click to enlarge.

Leadwerks Engine: Obviously, click to enlarge.

Unity Engine: Obviously, click to enlarge.

Cry Engine: Obviously, click to enlarge.

Unigine Engine: Obviously, click to enlarge.

I don’t see why this wouldn’t be good.

Not exactly sure what this thread is about. Both engines could definitely adopt some things from each other. Source level design is easier to learn and less time consuming than Unreal Engine level design. The only thing that bothers me about Source is lack-luster technology and inability to create larger, procedurally generated worlds. But I guess you made your point, it would be nice to retain Source’s ease of use in creation of content while adding some of the brilliant features that the Unreal Engine has.

Unfortunately, Source remains behind a few years deliberately to be compatible as well as easier to render for older hardware, Valve retains a lot of their community and players because of that approach.

By the way, your poll does not make sense, neither does the thread title.

Sorry, I’m horrible at those things, mind recommending a thread title?

Thread title? It’s mostly dream-discussion, “what if -” if you will. I think a good title would be “What would make Source and the Source SDK better?” That would naturally lead to what you want, and would give some broader discussion as well.

Thanks dude. Your very helpful :smiley:

I thought you found a way to port them from the title. :eng99:

Dude, like it won’t let me change thread title. I guess its not supposed to or it just hates me. I’m going to just ask a moderator to do it to save players from thinking about how the thread title sounds.

I actually think Source looks better than Unreal, at least realistically. Unreal may be pretty, but it still feels like a video game.

I’m saying like, if Mirror’s Edge maps were ported, imagine the beauty.

You would not particularly get the same visual results. The models would look nice, but you are still constrained to Source’s lighting.

Damn man, that Water Tunnel video was sexy… I want to know how to make awesome stuff like that.

Even if you did port the maps and the models and the textures with normals, specs, etc, they still wouldn’t look anywhere close to the same. Not only do the lighting systems differ, so do the shaders which effect things like reflections and normal maps.

But, wouldn’t Source try to make the best of the Unreal Lighting?

What do you mean, it isn’t like you can just copy over a few .dll’s and a compiler .exe and get Unreal Lightmass (I assume that’s what you’re talking about) on Source. They would have to code it themselves, and there’s probably going to be an updated vrad with Portal 2 that has dynamic lighting, but NOTHING near Lightmass…

It only takes minimal knowledge of game engines know that it would not yield any beneficial results to port them over.

robmaister, could you explain what Unreal Lightmass is? It sounds interesting.

I don’t actually have too much faith in the next iteration of Source, watch the new HD videos of Portal 2 uploaded by Valve and you tell that there is no dynamic sunlight being shown, it’s just another light entity. It appears that there only seems to be coned shaped dynamic lights available as well. Also, they seems to use the lighting very sparsely(one per room).

Lightmass is just a really nice tool that calculates lighting from the emissive channel of textures and models (unlike lights.rad, it calculates light from the entire surface instead of from a point a few units in front of the object. It also does the whole light bounces and color bounces thing, but it lets you control EVERY aspect of it and fine tune it down to exactly how much color you want to let bounce, how vivid that color should be, how bright light bounces should be, etc.

And the best part is it compiles in roughly the same time as their older lightmap-calculating tool, but generates lightmaps 1000x better

And what is stopping Valve from doing methods such as lightmass? They are alway about performance first and this seems to be an alternative far better than what they have now.

Partially related, stupid question; can I map for Mirror’s Edge, and if so, how?

Probably not.