[ QUOTE ]
There is no way HL2 even comes close to doom3 tech...
[/ QUOTE ]
Yes, but how many D3 mods can you list? I can only find a few through a google search. Let's see, there is one about duct take so you can see. There's one about a flare so you can see things you throw it at. There's on about hello kitty so you can laugh at what you see. I think there are a couple mods that involve walking down dark hallways. Oh, and someone's remaking Doom1 on the Doom3 engine, even tho Doom3 is a Doom1 remake...sorta.
Ely if your arguement is "doom3 tech is so time consuming no one makes mods for it" I'd like to say that having finished 2 projects with both engines Doom3 has a cleaner, faster and more succinct art pipe than source. Source makes you convert files and download all kinds of shit. Doom3 just 'works' the first time. I only have to write one text file too for doom3. Doom3 also treats every trinagle the same, not so with Source.
Source always breaks on me, everytime it's 'updated' something is fixed at the cost of something else breaking (that was usually working in the last version) the shader language has changed so that old tags no longer work (that's a particularly brilliant idea)
Unreal3 is probably the most user friendly of the bunch, with a gui interface to build shaders. it reminds me of max a lil bit with the material globes and such. Unreal3 is so easy to use, i mean, they even showed ror howto use it *wink*
From what i know about it skanker it's real-time. I thought that paralax mapping you are speaking of is the same thing tho.
The way it works (mind you i havent played with this yet) is you paint a grey scale image and it deforms
-white is extrusion
-black intrusion
-grey (r128, g128, b128) no displacemnet
[ QUOTE ]
From what i know about it skanker it's real-time. I thought that paralax mapping you are speaking of is the same thing tho.
The way it works (mind you i havent played with this yet) is you paint a grey scale image and it deforms
-white is extrusion
-black intrusion
-grey (r128, g128, b128) no displacemnet
I could be wrong...
-R
[/ QUOTE ]
Well, the way we do parrallax mapping is we create a hight map along with the RGB normal map. They take those two and give the illusion of depth. It's not really deforming the mesh in real time. I think UE3 does it that same way. In one demo they showed, you could clearly see that the geometry was simple, and only some bricks were modeled. If it was doing real-time displacement, then the mesh would have been broken up and sub-divided much more than what it was.
We put it on the back burner, because it wasn't worth the cpu time in a 3rd person, console game.
[ QUOTE ]
Ely if your arguement is "doom3 tech is so time consuming no one makes mods for it"
[/ QUOTE ]
No. I'm sure Doom3 is less time consuming. But it appears the tech is so limited to small dark indoor areas. Don't forget the shots we've seen of Quake4 look just like a Doom3 expansion, where as Quake3 held possibilies for many game types. Perhaps ID have something up their sleeves in order to expand on that and provide competition for Unreal3.
I agree with some of the people here about the gameplay not shining yet. However, you do have to realise that this is just a tech demo with excellent artwork. I can already think of 1000 ways to use some of the new physics and volume affectors. I'm actually curious if you can create an animated spacial physics volume. That would create some interesting effects with "jelly" characters when you throw objects into/at them.
Kimset sounds exactly like Quest3d, I had appreciated the interface of flowchart scripting at first, but it became cumbersome to get simple tasks done that would take one sentace of well paced scripting text. I would hope that Kimset generates a text based script, that one can add to or alter at will. Similar to how frontpage has the gui, and the html behind working simultaniously.
The visual tech has been used before, and my team is using it now with the Reality Engine, but I've never seen it so brilliantly used. I'm curious to who the actual artists are working behind the scenes. I'm educating my team on zbrush now using the Gnomon DVD's, and it is amazing what you can accomplish.
Time to update my portfolio. Much respect to the artists and directors involved.
Vermillion: Well your right, I just looked at what you said as another rip on the doom3 storyline. Although "shadowy corridors" can't be applied to any but doom3. Your only talking about two series of games, doom/quake, doom is entirely demons, quake is somewhat demons (moreso just described as monsters) quake2 was aliens and quake3 has none of these (just a blend of themes). Its not as if they don't do anything else but that, the problem is they haven't done a new IP since quake1.
"Unreal 3 can do displacement which no mainstream engine has even announced other than Epic's as far as i'm aware." - While its not exactly mainstream, Darkplaces (a quake1 based engine) can do displacement mapping.
This is why I don't think this unreal tech is that impressive, because all of these features have been seen before, they just haven't been combined and seen with such amazing art (because of lack of tools most likely). Not like the first doom3 videos, or HL2 lip synch videos, where the new tech was really shown for the first time.
Why is anyone talking about gameplay at all in this thread, this is a TECH VIDEO, nothing to do with gameplay.
I believe you are way out of line with this comment.
No rules were broken, threads offen go off on differnet tangents. I'm a hobbiest, I've spent allot of time & money learning the Unreal engine. This matters to me.
Perhaps my post was too eary in the thread.
[/ QUOTE ]
i'm sorry you feel that way, but as a moderator i'm looking out for the long-term health and maintenance of this community. it matters to everyone here; i just wanted to see a good topic of discussion not be derailed by a related but divisive off-topic issue. this thread has been doing well and i stand by my remarks; if you would like to discuss gameplay implications of the Unreal 3 engine in a civil manner you are welcome to start a new thread on the subject.
as to my own on-topic contribution, i have to agree that it's quite painful to know that you're looking at some really beautiful footage completely washed out by a second-hand video recording. when i visited Epic in 2003 to preview ut2k3 it was clear how rock solid the overall toolset was. i'm excited to see the next generation continue that trend. a really robust toolset is good for everyone involved, especially us artists.
Actually from what I have seen and even Ror mentioned. It wasnt the Gameplay issue that got nasty it was infighting between "vetrans" here about how good the engine was. Im still not sure why you didnt bother to settle that dispute with your admin abilities.
THATS IS SOME NICE EYECANDY!!...can't say ive been following the U3 engine devlopments so all of this is new to me,amazing stuff.
As for the thread,intresting mix of opinions,though I gotta agree with whoever it was that said something like "The engine isn't that surrpsing,but its the quality of the artwork in it that brings out its buety"...well ok not quite like that but you know what/who i mean .
Great tech,does not automatically mean Great game,when making a game theres an element of this so called thing known as TEAM-WORK... You need half desent tech,you need good designers how the tech is used and what it will be used for,you need a good team of coders to make things happen in the virtural world and you need artists to make this world happen and come to life. Without anyone of those things a game cannot happen.....then if you have people who have no idea what they are doing in any of those but have there hands on such tech,your simply not gonna get the quality of work shown in these videos.
Im yapping,but its like my rant in the Requests section of this forum,why is every TC,big mod project(bar afew charcter,imports,reskins etc) on that page now a HL2 mod?
Everyones drawn by amazing tech,most of which will not be used by modders becuase many of them will simply not know HOW to use it.
sorry ive gone rather offtopic,But Ror make sure you and the lads/laddets over there at Epic keep up the good work!!
I know that gui interface is very "artist friendly" way of ushering in shader tech to the slobbering masses that make the 3d game art community, and I'm all for it, but I hope that the gui writes a shader file that can be manipulated as mentioned earlier by one of the TRI guys for kismit.
I know I get frustrated with the shader stsyem in Source alot, and gui would make life easier at the begining when i was learning it all. But after the learning cruve is over it's much faster to copy paste and 'replace all' when creating say.....15 new materials all at one time.
Today I decided to make a sweeping change to all my art assets in a map, it was a shader command that needed to be swapped for another. using the replace all feature in editpad pro it was a very simple task and I didnt have to edit a gui 25-30 times to get all the assets to have the new shader commands. 10 mins later I imported all my new bsp textures and it was just a matter of pasting the default shader values and 'replace all' for the .tga names.
That's what makes text based shader control powerful; the speed of batch processing and editing. I hope this is an option in UE3.
Im curious about the terrain deformation. I hope they apply that to vechiles so you can have dents 'n dings in your tank and stuff as it takes on damage. getting wheels blown off and whatnot would be totally awesome.
Im stoked cuz i wanna destroy buildins and level that pretty city of yours
Grind: check out the latter half of vid 4. Looks to me like the most amazing rendition of an ocean yet seen in a videogame. Just my opinion, I hasten to add
Your totally right DAZ like I said earlier it looks like it could be REAL-TIME from the final fantasy 7 movie ..the style is even similar, that ocean looks amazing I cant wait to see more !!!
It does look promising. But given the quality of the video, and the distance away from the camera's position...I'm hoping for a closer angle. Also how objects and vehicles will interact with water. Battlefield2's newest video shows a damn good ocean as well.
awesome stuff, I love the new things they added like that glowing blue thing and the final scene *drool* theres not really much else to say so I'll just add to the bucket of OMFGWTFH4x this ROX!
The ambush scene won't make it out I doubt as Tim and Alan knew there were cameras in the audience when those 4 that leaked did and chose not to show the city adventure scene because of that.
I expect when we have vastly exceeded the quality of the tech shown in the vids and are a little embarassed by them, then the bigwigs will choose to release them heh.
Usually thats the way these things work for some reason, perhaps theres a logic there, but for artists its frustrating and I imagine also for coders and LD's.
Someone asked who was responsible for the art assets and I'd have to list most of the art team and ld's and coders to make sure you knew everyones names.
Those videos are awesome, and as for the people concerned with gameplay, the videos are clearly titles as "unreal ENGINE 3" and not "unreal 3". ID never really showed anygameplay, if any, when doom3's tech was showcased. The only way to demo the game at that time was to get the leaked alpha. and even then, the game wasn't optomized, nor could most video cards handle it.
I really wish the cityscape could be seen more clearly from those vids. it looks very well done. wish i could see more...
Got to see this at the show and it is indeed far more impressive than anything else I've seen. But. . .
I did want to lend some weight to the point that the artwork behind the engine is quite a bit of what you're seeing. I went to a talk with an Art Lead and Modeler from Epic talking about how they make their assets and it's not even in the ballpark of being realistic for most developers out there. Jerry, the Lead, mentioned that with a 7 person art team and release planned for '06, they are already planning to do some outsourcing to get the game done.
Why? They showed an APC vehicle that they had worked on: 3 weeks to concept, a full month to model, and 2 weeks to texture. I'm guessing those were not 40 hour weeks, but who knows. That's probably about triple what a regular old mesh-and-texture version would have taken. It paid off in an incredibly impressive piece of art (especially when you're trying to sell a graphics engine), but anything but the most intimate of games cannot warrant that kind of time spent on an asset. From the demo, this other game Epic is making will take advantage of the tech and will be unbelieveably pretty, but most other games will have no reason to nor have the resources to.
Just to sum up: the tech is great, the tools seem even greater (their particle tool was just sick), yet great artwork and presentation is a lot of what is making these demos so impressive.
Doc: Aye, althought I can't be totally specific for you, its safe to say that some of our stuff takes a long long time to produce.
I'm personally cranking out stuff much much faster than you are indicating and still producing an incredible result thanks purely to the tech.
The vehicle that Pete Hayes did was not a typical creation, it was a special case also, so don't be too sure of using it as a yardstick as its not more useful and estimation than using my workrate purely as a yardstick.
This stuff does take longer, but the tech still does a ton of extra work for you that takes your work beyond the visual level anything else will atm.
That said, its scary sometimes how much time the simplest asset takes now that it requires hi poly as well as the same old low poly plus 3 times as many textures, theres no way around that and its only the ease of the toolset as a whole for the entire game production side of things that helps to balance it out.
Mark Rein has been talking up how much programmer time the new tech saves and how much more power it puts in the hands of LD's and artists and he's completly right.
In a way, this is the most sane choice because as anyone who makes games knows, it all comes down to the coders in the end right?
The coders are the poor buggers that have to do the insane crunch in the end to make it work.
With out tech, a lot more of the work is passed on to the artists, which is great for leveling out the whole game production process but it is still going to make life a lot harder for all artists and also leave a lot of artists behind because with hi poly, you can't get away with being a hack as easily.
When you can create a model to the same level of quality as reality, its more obvious that if you don't truly understand the forms of each part of reality you are trying to create.
[ QUOTE ]
Ror: "Recently on the game illuminati board there was a thread damning and mocking Epic and stating that they could not possibly get these visuals running on a standard PC."
[/ QUOTE ]
Slayerjayman, you have to remember that tha highest available commercial tech, will become Middle of the road in 18 months. Sure they are probably using Dual Procesors with twin $400.00 videocards crosslinksed for more power, but it probably DOES run on a Windows box that fits in a standard tower case. All of that power will be at the upper end of average byt Christmas 2006, which is around their release date. We're doing the same thing at my company, our machines all have blisteringly expensive Videocards, and tons of RAM< but they are still Windows Boxes. Some of our employees, recently hired from other companies got the hands on demo of the Unreal 3 Editor, and it made them feel godlike, and they love thecontrol and the material editing capabilities, and the editor was running on a Windows Box.
clearly some of the feature set has been driven by the various potential clientele--like the seamless loading between maps business being targeted at MMOs--but instead of spreading everything too thin it really just seems to make the whole package all that much more robust and appealing. when i think of the current-generation unreal engine-licensed game and how diverse they are (Lineage II as compared to Splinter Cell), it's really exciting to think of what will be produced/be possible with this next gen of tech. and yes while there is an emphasis on graphics, the overall stability and consistency of the toolset is the real draw if you ask me. while there will be a whole lot of companies struggling to adapt to the nuances of next gen art asset standards, previous licensees will be able to get going faster in other areas. i think we'll be seeing greater overall complexity and maturation of other aspects of the development process... toolset stabilization is really key. visiting Epic just prior to the release of UT2k3 i was very impressed with the integration and overall suite of tools the unreal 2 engine offered and i know that's only going to get better with this upcoming iteration. exciting stuff
So Ror, something you touched on got me thinking. The videos and the the guys dialogue ( when you can make it out ) seem to emphasize the power of the toolset available to the creatives. So in essence, many things that previously would have had to have been hardcoded special events, that would have required engineering time and other resources, can now be done by level designers and artists alone. Which effectively gives them more power, and ultimately means that the creation of level content and actual gameplay has the potential to be much quicker than we are traditionally used to.
Am I right in that assessment? And If so would you go so far as to say that If there is increased efficiency in that area, that in the overall production process it potentially offsets the increased time needed to make actual Art assets? Does that make sense? Understood If you cant really discuss.
Daz: yeah thats exactly the point. Art takes longer but everything else that makes a game is vastly easier and quicker to produce.
The real problem imo now is choosing wisely at the start of production what you want and need to accomplish because theres no limits to what you can do anymore , not in the way you could before.
For instance, creating terrain is like playing populous, its just the hand of god , raising and lower, realtime, like a game and bang, you suddenly have an amazingly convincing world.
But then you will want to tweak that world and because the terrain looks so realistic for free as does the sky and the fog and the water, you are then compelled to continue with that trend and make a lot of foliage assets that are as real.
This is a real problem imo, at least the UE2 tech was still limiting in terms of lighting, shadows and polycount so things didnt look so real that you were always compelled to take so much time. Likewise, the toolset still requires coders to make the gameplay so you were limited by how many of the little ideas you could put in, how many environment events you could script and so on.
That limitation is gone too.
I'm not sure if I'm making it clear enough, but personally i think theres a darkside to our tech in that there's so much you can do now , but those things take time, its going to be really easy to get caught up in that and want to do everything you could not do with previous tech.
I really am very excited to see what other developers do with the tech because we can't even touch on a fraction of its power and still maintain a realistic continuity, we can only showcase some of its potential with some nice art and design.
So I guess in short, I think the problem with next gen production is going to be more to do with our having too much power and learning to be more mature and conservative in its use in order to continue to finish games in a timely fashion rather than indulge our lust to do everything!
The last generation of technology was much more about being creative and sneaky
[ QUOTE ]
great artwork and presentation is a lot of what is making these demos so impressive.
[/ QUOTE ]
But isnt that the point? Why even buy such a robust toolset if game devs wont have the time to use it its features, or use them well for that matter? Bloodlines compared to HL2 is a perfect example of buying a really top notch toolset expecting that it can do wonders for the game, but the team just did not have the time or understanding and in some cases, the capacity to make the assets in order to showcase the features that UE3 is certainly more than capable of doing. It would be no different than buying a SUV just because it looks cool, people can easily recognize the name, lots of seating room, it has lots of power under the hood and it rides well. Thats all good and dandy except that you bought it by putting a second mortgage on your house and the only time you use it is when you drive 3 miles to work everyday. Point is, if you are not going to use the resources that the engine has to offer then theres no real point or reason to get it.
As Ror mentioned earlier, its going to take an even more incredible amount of planning and budgeting and reasoning to make games that the UE3 can muster in a reasonable time period. I can really only think of one phrase that can sum up the kind of production path that these more recent game engines:
I sat in on this actual presentation. It was running on a computer that was only slightly faster than my pc at home. I will have that level pc in the next year easily. I was literally blown away. I was hooting and hollering (as Doc_Rob can attest). No other game out there has been as visually delicious as this one. I've played doom 3 and loved it, Played halflife 2 and I can't get enough of that, but watching this demo (and I did get to see the ambush scene) was the single best visual experience I've ever had in a game. I hope this serves as a lesson to developers, than when you hire top notch tallent and give them ample time, you can make incredible artwork that pushes the bar. I was very *very* inspired.
I also got to talk to Pete Hayes the vehicle modeler for about an hour afterwards, and he is a seriously cool dude. He explained to me a lot about the asset creation process that I enjoyed hearing.
hehe, I was razzing Pete about becoming a mini-celebrity after his presentation. The fact that you dropped his name gives me fuel to keep up the pressure!
yah this stuff is no doubt amazing. i'm quite excited to start working with it too. i haven't felt this way about an engine for a lonnngg time. great thred, too. thanks for the info, Ror!
It's really great to here from people who actually got to see the presentation are have been kind enough to talk about it openly enough and in a balanced enough way to make it clear to the naysayers in the thread that I have no been spinning hyperbole and trying to hide some deceptive ruse we employed to make it run like it did.
I really appreciate that calm and frank attitude as the routine denial of things being possible each iteration gets more silly each iteration as we have more and more history in our industry that proves the opposite is true.
I'm happy to hear of Pete getting some good press too, he works like a madman to make his stuff as impressive as it is. Theres a lot of talented people at Epic, but I like Pete's stuff a lot because it always looks like it would function in a way that I don't think mine would.
I would think that with art assets taking as long as they take to make that there would be a strong need for modular design and level assets that can work in multiple ways or with multiple texture sets. With each art asset taking a valuable amount of time to come to fruition, it would be paramount to design assets that can work together very well in order to get the best value for the time spent on each. You've already pointed out that when planning out something on the scale of an entire level or world you would be constantly walking a fine line between what is creatively possibly and what is reasonably feasable. I would think in single and multiplayer experiences that you would want to extend this type of design to the characters as well so that the game isn't stagnated by the same 4 or 5 enemies the whole game. Is this type of thing something that would be good to put in a portfolio and is it something that will be sought after for engine 3 games?
I got to sit next to poop during the class and my mouth dropped a few times at the things they were showing. The work time framed seemed on par with the visual quality in my opinion. Although Pete said people really speed things up once they've done it once. First class I've ever been to where the speaker said "You need to pimp your shit", thought that was great. Can't wait to get whatever it's going to be. I remember walking out saying "damn, I need to upgrade again" just like the HL2 class from last year.
So, since Legend is out, whos going to make the single player Unreal 3? I refuse to think that all of the new titles are going to be online dork fests. Even though every now and then I appreciate online play, it's the industries weakness as developers. Though not entirely developers faults, mostly due to the fact that online play sells. Bah..Pray that there will be a single player Unreal 3.
Wow those trees look really nice Daz, but then again Speedtree had these great looking trees for a few years already, and they never appeared in any game so far.
Well... maybe that is about to change.
I couldnt say for sure Daz, I think we are using it partly and I KNOW for sure we are also building stuff ourself , or rather I have seen other people doing and thats why im having to fumble toward and educated guess here!
Ah. Btw, I hope that didn't come across as me taking away from your work by suggesting that not everything was hand built. I really just wanted to wow with more shots Someone sent these out at work today.
It seem's you guys and the guys at bethesda think the same as your both using speed tree insted of making your own. I wonder why at GDC they were pimping that speed tree was being used by bethesda and not epic as well?
Replies
There is no way HL2 even comes close to doom3 tech...
[/ QUOTE ]
Yes, but how many D3 mods can you list? I can only find a few through a google search. Let's see, there is one about duct take so you can see. There's one about a flare so you can see things you throw it at. There's on about hello kitty so you can laugh at what you see. I think there are a couple mods that involve walking down dark hallways. Oh, and someone's remaking Doom1 on the Doom3 engine, even tho Doom3 is a Doom1 remake...sorta.
Unreal 3 can do displacement which no mainstream engine has even announced other than Epic's as far as i'm aware.
[/ QUOTE ]
Was it real time displacement or the parallax mapping they showed a while back?
Source always breaks on me, everytime it's 'updated' something is fixed at the cost of something else breaking (that was usually working in the last version) the shader language has changed so that old tags no longer work (that's a particularly brilliant idea)
Unreal3 is probably the most user friendly of the bunch, with a gui interface to build shaders. it reminds me of max a lil bit with the material globes and such. Unreal3 is so easy to use, i mean, they even showed ror howto use it *wink*
-R
The way it works (mind you i havent played with this yet) is you paint a grey scale image and it deforms
-white is extrusion
-black intrusion
-grey (r128, g128, b128) no displacemnet
I could be wrong...
-R
From what i know about it skanker it's real-time. I thought that paralax mapping you are speaking of is the same thing tho.
The way it works (mind you i havent played with this yet) is you paint a grey scale image and it deforms
-white is extrusion
-black intrusion
-grey (r128, g128, b128) no displacemnet
I could be wrong...
-R
[/ QUOTE ]
Well, the way we do parrallax mapping is we create a hight map along with the RGB normal map. They take those two and give the illusion of depth. It's not really deforming the mesh in real time. I think UE3 does it that same way. In one demo they showed, you could clearly see that the geometry was simple, and only some bricks were modeled. If it was doing real-time displacement, then the mesh would have been broken up and sub-divided much more than what it was.
We put it on the back burner, because it wasn't worth the cpu time in a 3rd person, console game.
Ely if your arguement is "doom3 tech is so time consuming no one makes mods for it"
[/ QUOTE ]
No. I'm sure Doom3 is less time consuming. But it appears the tech is so limited to small dark indoor areas. Don't forget the shots we've seen of Quake4 look just like a Doom3 expansion, where as Quake3 held possibilies for many game types. Perhaps ID have something up their sleeves in order to expand on that and provide competition for Unreal3.
Kimset sounds exactly like Quest3d, I had appreciated the interface of flowchart scripting at first, but it became cumbersome to get simple tasks done that would take one sentace of well paced scripting text. I would hope that Kimset generates a text based script, that one can add to or alter at will. Similar to how frontpage has the gui, and the html behind working simultaniously.
The visual tech has been used before, and my team is using it now with the Reality Engine, but I've never seen it so brilliantly used. I'm curious to who the actual artists are working behind the scenes. I'm educating my team on zbrush now using the Gnomon DVD's, and it is amazing what you can accomplish.
Time to update my portfolio. Much respect to the artists and directors involved.
"Unreal 3 can do displacement which no mainstream engine has even announced other than Epic's as far as i'm aware." - While its not exactly mainstream, Darkplaces (a quake1 based engine) can do displacement mapping.
This is why I don't think this unreal tech is that impressive, because all of these features have been seen before, they just haven't been combined and seen with such amazing art (because of lack of tools most likely). Not like the first doom3 videos, or HL2 lip synch videos, where the new tech was really shown for the first time.
Why is anyone talking about gameplay at all in this thread, this is a TECH VIDEO, nothing to do with gameplay.
I believe you are way out of line with this comment.
No rules were broken, threads offen go off on differnet tangents. I'm a hobbiest, I've spent allot of time & money learning the Unreal engine. This matters to me.
Perhaps my post was too eary in the thread.
[/ QUOTE ]
i'm sorry you feel that way, but as a moderator i'm looking out for the long-term health and maintenance of this community. it matters to everyone here; i just wanted to see a good topic of discussion not be derailed by a related but divisive off-topic issue. this thread has been doing well and i stand by my remarks; if you would like to discuss gameplay implications of the Unreal 3 engine in a civil manner you are welcome to start a new thread on the subject.
as to my own on-topic contribution, i have to agree that it's quite painful to know that you're looking at some really beautiful footage completely washed out by a second-hand video recording.
THATS IS SOME NICE EYECANDY!!...can't say ive been following the U3 engine devlopments so all of this is new to me,amazing stuff.
As for the thread,intresting mix of opinions,though I gotta agree with whoever it was that said something like "The engine isn't that surrpsing,but its the quality of the artwork in it that brings out its buety"...well ok not quite like that but you know what/who i mean
Great tech,does not automatically mean Great game,when making a game theres an element of this so called thing known as TEAM-WORK... You need half desent tech,you need good designers how the tech is used and what it will be used for,you need a good team of coders to make things happen in the virtural world and you need artists to make this world happen and come to life. Without anyone of those things a game cannot happen.....then if you have people who have no idea what they are doing in any of those but have there hands on such tech,your simply not gonna get the quality of work shown in these videos.
Im yapping,but its like my rant in the Requests section of this forum,why is every TC,big mod project(bar afew charcter,imports,reskins etc) on that page now a HL2 mod?
Everyones drawn by amazing tech,most of which will not be used by modders becuase many of them will simply not know HOW to use it.
sorry ive gone rather offtopic,But Ror make sure you and the lads/laddets over there at Epic keep up the good work!!
john
I know that gui interface is very "artist friendly" way of ushering in shader tech to the slobbering masses that make the 3d game art community, and I'm all for it, but I hope that the gui writes a shader file that can be manipulated as mentioned earlier by one of the TRI guys for kismit.
I know I get frustrated with the shader stsyem in Source alot, and gui would make life easier at the begining when i was learning it all. But after the learning cruve is over it's much faster to copy paste and 'replace all' when creating say.....15 new materials all at one time.
Today I decided to make a sweeping change to all my art assets in a map, it was a shader command that needed to be swapped for another. using the replace all feature in editpad pro it was a very simple task and I didnt have to edit a gui 25-30 times to get all the assets to have the new shader commands. 10 mins later I imported all my new bsp textures and it was just a matter of pasting the default shader values and 'replace all' for the .tga names.
That's what makes text based shader control powerful; the speed of batch processing and editing. I hope this is an option in UE3.
-R
Im stoked cuz i wanna destroy buildins and level that pretty city of yours
(totally amazing everything, they can't do water though; man that looked crappy. like some candy floss smoke stuff falling out of a pipe)
the last bit over the city was like something out of dream! impressive!
(enjoyed reading all this bitching, I think it's still quite healthy. New games need a bit of a punch in the face)
Any news on the ambush video?
I expect when we have vastly exceeded the quality of the tech shown in the vids and are a little embarassed by them, then the bigwigs will choose to release them heh.
Usually thats the way these things work for some reason, perhaps theres a logic there, but for artists its frustrating and I imagine also for coders and LD's.
Someone asked who was responsible for the art assets and I'd have to list most of the art team and ld's and coders to make sure you knew everyones names.
I really wish the cityscape could be seen more clearly from those vids. it looks very well done. wish i could see more...
I did want to lend some weight to the point that the artwork behind the engine is quite a bit of what you're seeing. I went to a talk with an Art Lead and Modeler from Epic talking about how they make their assets and it's not even in the ballpark of being realistic for most developers out there. Jerry, the Lead, mentioned that with a 7 person art team and release planned for '06, they are already planning to do some outsourcing to get the game done.
Why? They showed an APC vehicle that they had worked on: 3 weeks to concept, a full month to model, and 2 weeks to texture. I'm guessing those were not 40 hour weeks, but who knows. That's probably about triple what a regular old mesh-and-texture version would have taken. It paid off in an incredibly impressive piece of art (especially when you're trying to sell a graphics engine), but anything but the most intimate of games cannot warrant that kind of time spent on an asset. From the demo, this other game Epic is making will take advantage of the tech and will be unbelieveably pretty, but most other games will have no reason to nor have the resources to.
Just to sum up: the tech is great, the tools seem even greater (their particle tool was just sick), yet great artwork and presentation is a lot of what is making these demos so impressive.
I'm personally cranking out stuff much much faster than you are indicating and still producing an incredible result thanks purely to the tech.
The vehicle that Pete Hayes did was not a typical creation, it was a special case also, so don't be too sure of using it as a yardstick as its not more useful and estimation than using my workrate purely as a yardstick.
This stuff does take longer, but the tech still does a ton of extra work for you that takes your work beyond the visual level anything else will atm.
That said, its scary sometimes how much time the simplest asset takes now that it requires hi poly as well as the same old low poly plus 3 times as many textures, theres no way around that and its only the ease of the toolset as a whole for the entire game production side of things that helps to balance it out.
Mark Rein has been talking up how much programmer time the new tech saves and how much more power it puts in the hands of LD's and artists and he's completly right.
In a way, this is the most sane choice because as anyone who makes games knows, it all comes down to the coders in the end right?
The coders are the poor buggers that have to do the insane crunch in the end to make it work.
With out tech, a lot more of the work is passed on to the artists, which is great for leveling out the whole game production process but it is still going to make life a lot harder for all artists and also leave a lot of artists behind because with hi poly, you can't get away with being a hack as easily.
When you can create a model to the same level of quality as reality, its more obvious that if you don't truly understand the forms of each part of reality you are trying to create.
Ror: "Recently on the game illuminati board there was a thread damning and mocking Epic and stating that they could not possibly get these visuals running on a standard PC."
[/ QUOTE ]
Slayerjayman, you have to remember that tha highest available commercial tech, will become Middle of the road in 18 months. Sure they are probably using Dual Procesors with twin $400.00 videocards crosslinksed for more power, but it probably DOES run on a Windows box that fits in a standard tower case. All of that power will be at the upper end of average byt Christmas 2006, which is around their release date. We're doing the same thing at my company, our machines all have blisteringly expensive Videocards, and tons of RAM< but they are still Windows Boxes. Some of our employees, recently hired from other companies got the hands on demo of the Unreal 3 Editor, and it made them feel godlike, and they love thecontrol and the material editing capabilities, and the editor was running on a Windows Box.
Scott
Scott
Am I right in that assessment? And If so would you go so far as to say that If there is increased efficiency in that area, that in the overall production process it potentially offsets the increased time needed to make actual Art assets? Does that make sense? Understood If you cant really discuss.
The real problem imo now is choosing wisely at the start of production what you want and need to accomplish because theres no limits to what you can do anymore , not in the way you could before.
For instance, creating terrain is like playing populous, its just the hand of god , raising and lower, realtime, like a game and bang, you suddenly have an amazingly convincing world.
But then you will want to tweak that world and because the terrain looks so realistic for free as does the sky and the fog and the water, you are then compelled to continue with that trend and make a lot of foliage assets that are as real.
This is a real problem imo, at least the UE2 tech was still limiting in terms of lighting, shadows and polycount so things didnt look so real that you were always compelled to take so much time. Likewise, the toolset still requires coders to make the gameplay so you were limited by how many of the little ideas you could put in, how many environment events you could script and so on.
That limitation is gone too.
I'm not sure if I'm making it clear enough, but personally i think theres a darkside to our tech in that there's so much you can do now , but those things take time, its going to be really easy to get caught up in that and want to do everything you could not do with previous tech.
I really am very excited to see what other developers do with the tech because we can't even touch on a fraction of its power and still maintain a realistic continuity, we can only showcase some of its potential with some nice art and design.
So I guess in short, I think the problem with next gen production is going to be more to do with our having too much power and learning to be more mature and conservative in its use in order to continue to finish games in a timely fashion rather than indulge our lust to do everything!
The last generation of technology was much more about being creative and sneaky
great artwork and presentation is a lot of what is making these demos so impressive.
[/ QUOTE ]
But isnt that the point? Why even buy such a robust toolset if game devs wont have the time to use it its features, or use them well for that matter? Bloodlines compared to HL2 is a perfect example of buying a really top notch toolset expecting that it can do wonders for the game, but the team just did not have the time or understanding and in some cases, the capacity to make the assets in order to showcase the features that UE3 is certainly more than capable of doing. It would be no different than buying a SUV just because it looks cool, people can easily recognize the name, lots of seating room, it has lots of power under the hood and it rides well. Thats all good and dandy except that you bought it by putting a second mortgage on your house and the only time you use it is when you drive 3 miles to work everyday. Point is, if you are not going to use the resources that the engine has to offer then theres no real point or reason to get it.
As Ror mentioned earlier, its going to take an even more incredible amount of planning and budgeting and reasoning to make games that the UE3 can muster in a reasonable time period. I can really only think of one phrase that can sum up the kind of production path that these more recent game engines:
With great power comes great responsibility
(it's not that cheesy ;-) )
I also got to talk to Pete Hayes the vehicle modeler for about an hour afterwards, and he is a seriously cool dude. He explained to me a lot about the asset creation process that I enjoyed hearing.
I really appreciate that calm and frank attitude as the routine denial of things being possible each iteration gets more silly each iteration as we have more and more history in our industry that proves the opposite is true.
I'm happy to hear of Pete getting some good press too, he works like a madman to make his stuff as impressive as it is. Theres a lot of talented people at Epic, but I like Pete's stuff a lot because it always looks like it would function in a way that I don't think mine would.
Some amazing shots here. Click on the U3 tab on the left:
http://www.idvinc.com/html/downloads_images.htm
Well... maybe that is about to change.