Lighting, Mood, Story telling, Specific looks driven by specific art direction... I guess these things don't matter when you procedurally generate fields, meadows and mountains.
I think automation like that can be very very helpful, it can lighten the load on a team and allow them to focus on more specific goals. Is it an end to a means probably not, is it a good start for artists to take over and actually create a game or a story? Sure!
I look at this more of a stage in a theater, you still need people to come in design sets, create costumes, write compelling plays and preform it all.
It's not a terribly new concept.
Elite 2 had an entirely procedural galaxy, on a floppy :P
For the right games it can be wonderful, I'm a big fan of procedural landscapes and worlds, the biggest pro is that a massive almost infinite world can be generated from scratch from just a seed and be the same every time.
[ame]www.youtube.com/watch?v=yrcEUO7TC28[/ame]
We'll still be hungry for handcrafted experiences though.
Lighting, Mood, Story telling, Specific looks driven by specific art direction... I guess these things don't matter when you procedurally generate fields, meadows and mountains.
I think automation like that can be very very helpful, it can lighten the load on a team and allow them to focus on more specific goals. Is it an end to a means probably not, is it a good start for artists to take over and actually create a game or a story? Sure!
I look at this more of a stage in a theater, you still need people to come in design sets, create costumes, write compelling plays and preform it all.
pretty cool stuff, but im not worried. Procedural programs are very limited at the moment. And even if they really get good, they will hardly be used for supper complex stuff like asset creation.
The terrain looks unnatural to me. Could be the shit texturing or lighting. Regardless, a human artist's eye will always be needed, as Mark pointed out.
While it can be slightly breathtaking, the lack of control over the look of the terrain compromises it's entire use. I wouldn't mind using this as a rough base though, and then retopo/remesh. Some of those surfaces look very interesting, and difficult to hand-model.
I could see this as a way to create a natural looking terrain with realistic physical shapes and environment artists could come in after and add the "man-made" features like civilization. That Infinity demo is honestly breathtaking. The thought that you could even enter a planet's atmosphere at all, much less expect to see the intricate and nuanced details inherent in that planet is nothing short of amazing. That's just never been done in any space simulation before and I've always been aching to just fly straight into a planet to see what's on the surface.
That Infinity Demo is fantastic. I just hope in the final release, the author 'slows' down the ships, atleast when near a planet, because the speed at which they decent really screws up the mental reading of a size and reach velocity.
That Infinity Demo is fantastic. I just hope in the final release, the author 'slows' down the ships, atleast when near a planet, because the speed at which they decent really screws up the mental reading of a size and reach velocity.
I think it's a tricky problem with speeds considering the time it would take to actually travel to places with realistic velocities, but we'll see.
The thing about this tech is that it's not to replace the process of making environmental art, it has its extremely interesting place to give us dynamic and constantly new environments, it doesn't even have to be realistic or good looking to add to gameplay.
Spore touched on the subject too.
If we're going to talk about procedural tools that have had its big place in the industry then speed-tree is a good example.
^^ exactly. That head scan tech was mediocre at best. Though the stuff in LANoire is pretty damn nice. Even that requires a good amount of human cleanup/processing, I'm sure.
I love procedural environments. Very few games have actually pulled them of well yet, but there's a lot of potential. I love the idea of games being a personal experience, a personal story, and giving the player his own unique environment to explore is an awesome way to help that experience. True interactive storytelling -> give the player his own story, his own world!
I love procedural environments. Very few games have actually pulled them of well yet, but there's a lot of potential. I love the idea of games being a personal experience, a personal story, and giving the player his own unique environment to explore is an awesome way to help that experience. True interactive storytelling -> give the player his own story, his own world!
Could be some kind of fantastic wet dream in voxlap.
Didn't we collectively say the same about characters when that tech to scan heads first came along?
It's also flawed on the basic level of you can't scan stuff that doesn't exist. I'm sure at some point in the future it'll work well for humans. But any kind of creature, or even an exaggerated human, is not possible for obvious reasons.
Unless of course they're going to do a physical sculpt in clay or something and then scan that. But at that point that goes right back to an artist doing it.
Likewise for environments. Maybe at some point we'll get tip-top terrain generated. But I don't see how it'll produce the more fantastic environment elements.
The environments remind me of CGI from back in the early 90's. I think we'll see procedural texturing take off quicker than any of this stuff.
Many developers will incorporate procedural landscape generation their process, but manually tweak it afterwards.
I know for oblivion they would have procedural processes going on their heightmap, and then manually build stuff on it afterwards.
And I'm wildly guessing that something similar is done for the just cause games.
It's fully possible to have a perfectly good looking procedurally generated landscape and have the generation plop down pre-made assets like trees and stuff, it's just that most games aren't made to have procedural content, they have specific custom made content that is the same every time.
The computer doesn't think like a human does, this is why people use these as tools, not a be all end all solution, unless they want a truly procedural game that can be different every time.
Hmm, I've never been too hot for the idea of procedurally generated environments/levels (that are not meant to be adjusted manually). I would rather have an area carefully crafted by a fellow human being than a cold, calculating machine. Might be me, but the idea of a world that can change every time you play it just doesn't appeal to me. I've always been a fan of games that feel personal, tight and linear (I know, crazy:poly122:). For the same reasons, I've never liked the sandbox genre or gameplay that relies too heavily on systems and rulesets.
Well, now to get back on the topic. I could see such procedural environments used in certain types of games, mostly the ones that have a different sense of scale. Flight sims could benefit from something like that, especially if the major features of the terrain could be derived from the satellite data where as the details generated procedurally.
Lighting, Mood, Story telling, Specific looks driven by specific art direction... I guess these things don't matter when you procedurally generate fields, meadows and mountains.
I think automation like that can be very very helpful, it can lighten the load on a team and allow them to focus on more specific goals. Is it an end to a means probably not, is it a good start for artists to take over and actually create a game or a story? Sure!
I look at this more of a stage in a theater, you still need people to come in design sets, create costumes, write compelling plays and preform it all.
i'm fairly certain that this falls into your second paragraph, but i would think that being able to procedurally generate an entire world, and then go in afterwards and adjust lighting/mood, would make things much faster for a team?
looks like terragen and terragen takes a crap load of work for an artist to get a decent render that actually has the right procedural textures in the right places with the right lighting etc so I dont see this making much of a difference to environment art as a trade, more likely it could offer an alternative solution to an age old problem and Im all for that.
That Infinity Demo is fantastic. I just hope in the final release, the author 'slows' down the ships, atleast when near a planet, because the speed at which they decent really screws up the mental reading of a size and reach velocity.
Well, it would be realistic to limit the speed in atmosphere since drag would make it hard to move at incredible speeds, but in space, there is no drag so you can go very fast. So, allow super speeds in space, but put a ramp limit for traveling in atmosphere (the deeper you go, the thicker the air and more drag).
Yeah, that's what I meant, by dilating the time of entry, you can fake the feeling of approaching the planet and breaking orbit.
That, or make it so that the player needs to 'land' the ship correctly...I'm pretty sure keeping a right nose angle, and having a bunch of lights go red and start beeping will help distract.
^^ exactly. That head scan tech was mediocre at best. Though the stuff in LANoire is pretty damn nice. Even that requires a good amount of human cleanup/processing, I'm sure.
While head scanning didn't kill character artists industry wide (and isn't even close), I do know some people who refused work on some of those games saying that they'd rather work for a company where they're not just retopo artists.
But the argument is old. They said this about head scanning, procedural textures and so on. The only real danger is landing a job at such a studio and then doing something different than you thought you would do.
It's not a terribly new concept.
Elite 2 had an entirely procedural galaxy, on a floppy :P
For the right games it can be wonderful, I'm a big fan of procedural landscapes and worlds, the biggest pro is that a massive almost infinite world can be generated from scratch from just a seed and be the same every time.
The only thing they could add would be hotzone shaders. Specific areas will be crafted on the procedural...it would be a great way to make a ME1 type of planet interaction, but minus the gravity defying Mako.
Replies
I think automation like that can be very very helpful, it can lighten the load on a team and allow them to focus on more specific goals. Is it an end to a means probably not, is it a good start for artists to take over and actually create a game or a story? Sure!
I look at this more of a stage in a theater, you still need people to come in design sets, create costumes, write compelling plays and preform it all.
Elite 2 had an entirely procedural galaxy, on a floppy :P
For the right games it can be wonderful, I'm a big fan of procedural landscapes and worlds, the biggest pro is that a massive almost infinite world can be generated from scratch from just a seed and be the same every time.
[ame]www.youtube.com/watch?v=yrcEUO7TC28[/ame]
We'll still be hungry for handcrafted experiences though.
+1
I think it's a tricky problem with speeds considering the time it would take to actually travel to places with realistic velocities, but we'll see.
The thing about this tech is that it's not to replace the process of making environmental art, it has its extremely interesting place to give us dynamic and constantly new environments, it doesn't even have to be realistic or good looking to add to gameplay.
Spore touched on the subject too.
If we're going to talk about procedural tools that have had its big place in the industry then speed-tree is a good example.
Could be some kind of fantastic wet dream in voxlap.
It's also flawed on the basic level of you can't scan stuff that doesn't exist. I'm sure at some point in the future it'll work well for humans. But any kind of creature, or even an exaggerated human, is not possible for obvious reasons.
Unless of course they're going to do a physical sculpt in clay or something and then scan that. But at that point that goes right back to an artist doing it.
Likewise for environments. Maybe at some point we'll get tip-top terrain generated. But I don't see how it'll produce the more fantastic environment elements.
this man understands
It's not so much the usage of voxels which pretty much came and went, but the combination of voxel+destructability and procedurally generated worlds
Many developers will incorporate procedural landscape generation their process, but manually tweak it afterwards.
I know for oblivion they would have procedural processes going on their heightmap, and then manually build stuff on it afterwards.
And I'm wildly guessing that something similar is done for the just cause games.
It's fully possible to have a perfectly good looking procedurally generated landscape and have the generation plop down pre-made assets like trees and stuff, it's just that most games aren't made to have procedural content, they have specific custom made content that is the same every time.
The computer doesn't think like a human does, this is why people use these as tools, not a be all end all solution, unless they want a truly procedural game that can be different every time.
Well, now to get back on the topic. I could see such procedural environments used in certain types of games, mostly the ones that have a different sense of scale. Flight sims could benefit from something like that, especially if the major features of the terrain could be derived from the satellite data where as the details generated procedurally.
i'm fairly certain that this falls into your second paragraph, but i would think that being able to procedurally generate an entire world, and then go in afterwards and adjust lighting/mood, would make things much faster for a team?
[ame]http://www.youtube.com/watch?v=_CCZIBDt1uM&feature=channel_video_title[/ame]
Alot of procedural stuff, it looks fantastic, and offers great possibilities for gameplay .
Well, it would be realistic to limit the speed in atmosphere since drag would make it hard to move at incredible speeds, but in space, there is no drag so you can go very fast. So, allow super speeds in space, but put a ramp limit for traveling in atmosphere (the deeper you go, the thicker the air and more drag).
That, or make it so that the player needs to 'land' the ship correctly...I'm pretty sure keeping a right nose angle, and having a bunch of lights go red and start beeping will help distract.
While head scanning didn't kill character artists industry wide (and isn't even close), I do know some people who refused work on some of those games saying that they'd rather work for a company where they're not just retopo artists.
But the argument is old. They said this about head scanning, procedural textures and so on. The only real danger is landing a job at such a studio and then doing something different than you thought you would do.
That.... was fucking beautiful.
And the thing is, everything star you see in that game will be an actual sun that you can go to, and every sun might have planets.
There just isn't any way of NOT having it procedural, that's the beauty of procedural stuff.
It could be pretty powerful in the right hands and used in right way.