In Unreal3 feature list i found next item:

Ultra high quality and high performance pre-computed shadow masks allow offline processing of static light interactions, while retaining fully dynamic specular lighting and reflections.

What do you think about pre-computed shadow masks ?
Is this some kind of lightmaps ? Or this is a pre-computed shadow cubemap from light source ?

Thanks for your answers.

It’s just a fancy way of saying “we have a black texture that we project onto objects as they move by”. They have this in UT2004.

I’m getting pretty sick of all this overrated “Next-Generation” BS the game companies are pushing, and then they put out a totally unremarkable engine.

Can you explain to me what , do you mean when you saying: “we have a black texture that we project onto objects as they move by” ?

That means it’s just a shadow map that was pre-computed. So what’s happening is for static lights they will precompute the shadows so that they can have high quality soft shadows, and when rendering, these shadow maps will be textured into the rest of the scene (like you would do with lightmapping) while they do dynamic specular and reflection calculations for a convincing light.


You just UV map the object using each vertex position times a multiplier, plus an offset, so that when they move under a tree or something, a “shadow” of the tree will appear on them.

If I got it it was called “shadow cookie approach” some years ago.

If you’ve seen the Unreal 3 “gothic” level fly-through, I think you’d have a little more lassitude for their marketing. It looks remarkably good.

My question is: “where is the army of artists going to come from?”

Current games have already peaked over 200 artists in production. I live in fear of the next generation.

Yeah, it’s like Half-Life2: the game is incredible, it looks really good, it feels really good but it is not so advanced from the technical point of view(doom3 is actually more advanced) - thought HL still has some of most incredible graphics and animations arts i’ve ever seen.
It’s the way of life: it is not so hard to CODE an engine but it’s much harder to USE it :slight_smile:
A good artist is as valuable as a good programmer now - but you need much more artists then programmers(number of first ones tend to rise as number of second is constant) sign
hey, i’m happy that programming is just my hobby :wink:

“where is the army of artists going to come from?”

Yes, larger levels and many more models will call for more artists, but not more graphics programmers.

C’est la vie.
Also, it’s good news to students.

“where is the army of artists going to come from?”

better tools, eg a gothic level i imagine will have all those fiddly bits of architechure on the walls and things, now normally they would be handmade by the artist, but with a lot of good/difficult programming u could allow the artist to ‘grow’ these procedurally. sure some objects dont lend themselves to well to this sort of thing eg humans (but then again we all have 2 eyes etc, ie the same basic template, these could be altered slightly in an algotithm to create various looking ppl) compared to others planst/rocks. ppl in the future arent gonna wanna see the same object or texture tiled in their game world, theyre gonna want unique stuff, which of course this all raises some lovely programming issues of where to find the memory to store all these things (what im grappling with at the moment)