next-next gen is terrifying?
polycounter lvl 8
http://www.eurogamer.net/articles/2011-12-19-naughty-dog-move-to-next-gen-is-terrifying
Is this really true that developers are already working ps4 titles? I kinda can understand that transition was hard for them from ps2 to ps3, as it was to many as well. Any ideas what future might bring to modelers or texture artists or level designers when we speak of "next gen?"
I'm sort of happy with the current skills & tech background I have, and I wonder what kind of "completely new stuff" I might have to learn in the near future about game development as a level designer who enjoys making his own environment assets?
Is this really true that developers are already working ps4 titles? I kinda can understand that transition was hard for them from ps2 to ps3, as it was to many as well. Any ideas what future might bring to modelers or texture artists or level designers when we speak of "next gen?"
I'm sort of happy with the current skills & tech background I have, and I wonder what kind of "completely new stuff" I might have to learn in the near future about game development as a level designer who enjoys making his own environment assets?
Replies
The thing with Sony is they don't seem to make things very accessible to people. Their dynamic themes, for instance, has their own friggin coding language. They want everything to be unique and their own. It makes sense to an extent. But why make it harder for developers to create amazing games for them? I dunno. I don't know too much about developing full unique games for ps3, but doing stuff for dynamic themes, and Playstation Home is a pain. Lots of hidden limits, and weird work around to make things work.
For example many "next gen" environments are in current-gen games.
1. The move to high-poly/normal maps was enormous. Even with new console hardware able to display more polys and higher-res textures, the high-poly assets are still going to be the same.
2. It's feasible the biggest improvements will come from utilizing the hardware to make the in-game result closer to the source assets (rather than requiring more detailed assets). If you look at something like RAGE, what would improve graphically if it was built for the PS4? The textures would be higher resolution -- closer to the resolution they were painted, more than likely -- and the lighting would probably be more dynamic. Neither of this things actually require more artist time, just more graphics memory and storage.
I'm not saying things won't change, but it's hard to imagine the kind of shift in asset creation time that happened with the current generation. And obviously the people in the know can't comment... :P
Next next gen is terrifying for the fact that moving consoles is terrifying, will there be enough of a market there? the 360/ps3 market is already in place, it's huge, the new one will be smaller, so do you make that economical jump?
Less people will be buying that new game of yours if you do, and you can't increase the price of the game, since people expect it to cost just as much as the 360/ps3 variant.
That's always the case with a new console generation. If you ship at launch, there's nobody to sell to. You're hoping they buy your game when they buy their console but it's definitely a slow burn as opposed to a huge sale spike on Day 1.
I personally think they need to carefully manage the hardware overlap and make sure that you can get on their network with whatever system you have. What hardware you have determines what games you can play but I don't think they should separate the pools of customers and force people to start over on a new network, that would be suicide.
From an artist/animator standpoint, I say bring it! I welcome new hardware, new specs and new ways to work.
"Next gen" was used when xbox 360 and ps3 were coming out, then people just never stopped saying it once they were out. Its frustrating and kind of annoying when people say next gen now. The moment those consoles were released, next gen became this gen. "next gen" is now for ps4 and xbox "720", I think.
EDIT: 1024th post woot!!!
I mean seriously...
That is silly :P I can give you that if you want to, right now, on currentgen consoles :P
Nextgen will be much more about shaders and rendering than polycounts and texturesizes. Sure, we will upscale the textures a step or two, and that will be enough, then the shaders will handle the rest.
As I said in another thread, if you would take uncharted 3, raise the polycount to 10x what it is now, would be vastly different? No. :P Imagine avatar movie in 3d at the cinema, and imagine that the mechs "only" had 30k tris. You think you would be able to tell that well? I think rendering and pipeline improvements will be the biggest changes.
Hm, I dont think that ever really works well in this industry. Even within one generation I've had to learn and grow a ton.
I HOPE that we get to see some advancements in animation, specifically the incorporation of vertex animation for faces instead of the clunky bone based systems we use now.
I HOPE that we are allowed to animate the scale on bones and that we can increase bone counts significantly.
I HOPE we can use more dynamic solutions on characters for secondary motion, or sliding skin or muscle bulges.
I agree, the next-next gen will be about shaders and rendering. That's what really needs to change for things to look better. I also am a believer in voxels and cloud computing coming into play at some point, maybe sooner than people would expect.
IMO the next gen isn't until more consideration is taken to the individual reflectance models of materials, semi-accurate radiosity, and until there is displaceable matter within the game world. Until then it's just a polished, high-end pc version of what's going on today.
I think they need to licence an engine like Unity or make their own in house. Then have that tailored for the console with standardised, things, like every game has to have physics, euphoria etc, the games dont have to use them but they are available if they want(for a nominal fee of course). Then they should have something like the Unity Asset store where developers can buy and sell premade code and other things.
Developers already have problems developing for these new controllers. They dont have the time or money to iterate them. Sony/Microsoft has to explore and refine, there own consoles and how they want to be used, if they want good games that is.
Just so I'm clear here, did you really just suggest that Sony/Microsoft should licence/develop a console specific engine and anybody developing for them should use it? Really? You don't think that that would a phenominal amount of work to develop a multi-platform title since the'd have to use different engines. Not to mention the millions upon millions of dollars company's have spent making their own. Why would the have any intrest in throwing that away? I really don't see how that would be in anybody's intereest, and at best you'd end up with Microsoft/Apple circa 2000 situation where all the titles are on one platform.
I'm not so sure, I mean in the switch from PS2 to PS3 etc. Industry standard engines as big as UDK, Unity and Cryengine weren't as readily available. I think it will take a while for these packages to catch up with the next gen, but I don't think its a stretch to imagine being able to download Cryengine and develop a game for the PS4 in two years.
Add in the massive increase in digital sales and indie markets on consoles and I think it will be far more easy for smaller studios to develop smaller titles. As someone else said, new tech may be a bitch to code for in the early days, but artist pipelines probably wont change too dramatically.
I'd also add my own prediction: Steam like Digital distribution services on Xbox and PS4. For example you buy Final Fantasy XVII in a retail store and tie the Key to your PSN account and you can download the same whenever you like ala steam.
:thumbup: Definitely more shader instructions to play with. No clue what a 360 or PS3 can push instruction wise.
This is going to price all but the largest and heaviest studios out of the AAA game arena and potentially spells disaster for a lot of medium sized outfits.
engine wise - I hope it's all going to be about lighting, shading and post processing because those things are all shit on current gen hardware
Next gen will be about processing time spent on AI and physics.
MineCraft and the whole Indie scene. This wasn't as mainstream as it is now when we made the transition from PS2/Xbox/Gamecube to the current gen consoles.
You'll find that there will be casualties where studios will over reach themselves but in theory, with the likes of Unreal Engine, Unity and CryEngine, they're taking a lot of the pain out. Artists may have to adapt their work flows slightly but nothing like going from the previous generation to current. If anything it will be more liberating, as long as the tools are made to help artists make the most out of the tech whilst making it easy.
Plus also because the smartphone market is maturing at a quick pace and can run games which aren't too far off what you are seeing on the consoles, plus the likes of the Playstation Network and XBLA, teams/companies who are worried about making a massive transition from current to next gen can go sideways and keep the pipeline similar to what it is now but still use some of the goodies of the next crop of machines.
You don't need to be pushing billions of polys, have dozens of shaders on a model, and add 4x the noise that appears on a lot of assets in the current generation of games.
I am quite looking forward to seeing what can be done with the new consoles, and how tool makers/plugin writers will try and answer the call for good user interfaces to get the most out of shaders, etc but they're not the be all and end all. They're just there to help us make pretty things whilst not wanking over them for several days/weeks/months.
I still remember when it was all measured in how many bits the consoles were i.e. 8bit, 16bit, etc... actually what are we on now and what's the next lot? 512bit?
Also, i suspect launch titles on Gen4 consoles will probably launch on both Gen3 and Gen4 with the exceptions of platform exclusive titles like Halo, Gears and Uncharted.
At least the ones who want to make any money...
yes i agree for the most part... that will be where the biggest extra spends will be, needs to be as my girlfreinds says " it looks good but he moves like robot" @uncharted 3 which lets face it is deffinatly above the curve in character movement
i think this next-next-gen will be the generation of soft surfaces or non-surfaced objects, nice alpha, sub surface lighting, soft/3d particles really good fog and volumetric techniques which are nearly there anyway
[ame="
edit- also i think its interesting to se what happens in the middle-end of a generation when i think the bigger leaps forward will be made
call of duty 2 (360 launch title)
and later that gen, modern warfare2 (360)
EDIT_due to porr research skillz
2007 PC:
2007 Console:
2011 PC:
2011 Console:
I dont think so. Also tesselation is not THAT big of a deal, requires very little to no change of the pipeline we currently have. Look at nDo2 for example, create normal, AO, cavity, heightmap in one go, slap it on in UDK and you got yourself tesselation. Making masks for tesselation is not that hard either actually, so smart tesselation might be used on some surfaces, like terrain and rocks, but I think it will be a minor change. I do think that we will have alot more fx and shader guys though. People will have to unlearn alot of old haxx though, and think more of getting a good clean rendering.
I also think that artists will have to think more in terms of constructing things of materials etc than just thinking in terms of "this part needs more spec since its metal"
Ryswick wins the award for most retarded comparison pics ever, haha
PS3 2011:
TROLOLOL
Edit: LOL yeah, Snefer pretty much nailed it.
Fun games will always be played more and longer than the games with the best graphics/music.
"Quick, microsofts new console is out, we need a game!"
"Just port call of duty 2 to it!"
Yes you did
edit: no santa smilies polycount?
I could see tesselation displacement offsetting the use of binormal correction on Normal Maps, seeing how it will be a few billion polygons on screen at a time. Basically might be similar to plopping in our Zbrush Models into the game, and it will go down in Subdivision level based on distance to camera (something to that effect).
Texture sizes will go up a bit, but the addition of Displacement maps will be pretty key to cool silhouettes (and possibly cool deformation).
UDK has that new Multi layered SSS which is pretty sweet. Makes for much more believable skin.
Ultimately I can see games becoming a LOT more expensive to make. I can see outsourcing a LOT more common for meanial tasks, and our "1-month per unique character pipeline" will turn into 2-3 months with shitloads of animatable blendshapes.
All you need is one good shader artist that creates specialized functions, and all the artist has to do is plonk in the textures where needed and be done. Plus, you can't simply throw displacement maps at everything, even for interesting shapes when they're of a certain size as it will look weird.
Are you honestly telling me that it's cheaper to recreate an entire function for SSS in UDK on per basis case per model, vs a simple map in the SSS slot of you DX11 capable engine which can be masked by a simple Lerp by your Gloss map?
Honestly, new tech will make stuff easier, not harder, hell, even animators will have an easy time with deformable models since they don't need to do what the guys in the old engines did, which was either make several morph targets, bake in havok physics and/or animate by bones, just look at Prey, Doom 3 and Quake 4, you can't honestly tell me that was cheap for the trouble of animating a static prop vs. what we have today or tomorrow.
This is not how realtime displacement works. Realtime displacement does nothing for the shading (only silhouette and shadows). And the tessellated surfaces aren't currently anything like the sub division surfaces you see in max or zbrush.
The good looking games come from the good art teams, tech dont matter.
Sunuva! You serious?
We have that uniengine demo up n running, which lets you toggle between wireframe tessellation on and off. It seems to do a decent job of actually adding realtime displacement... or am I misunderstanding this? It does seem to do a good job of augmenting lighting.
[ame="
Tech director here informed me (in lamens terms) it's similar to using vertex offsets generated by a texture (sorry I guess this is not necessarily a 'displacement map').
Ace-Angel:
I don't think it works quite that linearly, mind you I don't actually have actual EXPERIENCE with it (going what I know of Mental Ray). But a multi-layered SSS would require:
A unique:
Subcutaneous skin Texture
Epidermal Skin Texture
Diffuse Skin Texture
A series of masks that let's you blend between them.
I could definitely see it STARTING as a simple 1-shader fits all type thing. Until a game like a theoretical 'Uncharted 5' comes out that shows the huge benefits reaped from unique textures per each head. An old Caucasian dude would likely have more scatter radius than a young Black guy, or have a lot more varicose veins than an asian baby. Or if you wanted them to fight some kinda jelly monster.
Then have the modellers create blend shapes for each expression on top of everything else? (And also going into a LOT more detail per model).
Maybe things WILL get faster I've stopped predicting where technology goes.
But truth is, graphical technological developments have NEVER really been about making things cheaper and easier to do. They've always been about making things look better, and so far, that always seemed to make things take longer to make.
Who knows, this might be the generation where the games industry has to branch off, between 'surfacing artists', and speed modellers. Like film.
More than anything, I think we would get the most visual bang-for-buck if we could get per-poly collision, real-time cloth & hair simulation. But I have no illusions that setting those up would be faster than zbrushing folds n wrinkles, and painting/arranging hair cards.
But us game-art guys are a pretty resourceful bunch. I'm sure we'll find ways to make things a lot faster, and much more streamlined.
You'd be surprised at how small the gap is between the two (at least in terms of modeling, surfacing and rigging) It's a pretty linear progression, film stuff just has a little bit more of everything and the work that's being done on current characters is already pretty close. In fact, in a lot of ways it's even easier..because you don't have to go to crazy with optimizations.
Blasphemy! You could argue that it has a better art direction (the same way I could say FF13 had better looking characters) but you can't deny that The Witcher 2 is the technically superior game.
Not only that, but lords of shadows has a locked camera and thus a focus of detail (and lack of it elsewhere) due to that, if you would be able to turn around in that game it would break completely, much like god of war series.
The witcher 2 has a free camera.
But yes, one can argue art styles.
On the topic, I personally think that we will see some CG movie tech come to the game industry because we'll be able to handle it realtime.
but then you will see a title that uses both dynamicly animated characters for the foreground (lots of bug testing and fixing of those setups) LODDing them to crowds of 1000s of morph targetted characters for the background...(this is pretty much what they do anyway for alot of massive crowd games but i think this will become more common and not just for epic battles)
there youve got MORE work
even in this gen were doing PS2 level work for the LODs ontop of the current gen work for the high... add another level.
Yes, why should they care about cross platform. They should be making development on their console as simply and as easy as possible to attract developers.
Cross platform development hasnt gone that great so far, especially for the PS3. But what difference does it make if your using a different engine if half the work is already done. They normally have another developer do the port anyway.
I'm not saying they should force developers to use it though, they should create incentives to use it, like cheaper than the competition, pretailored and opmitised, with lots of addons and a store similar to the Unity Assets store.
You couldnt get a game like Skyrim from a new developer for instance. A game like that needs over ten years of development of building up over past games.
So you want to end up with only the few developers that have already spent billions and years developing to be able to make games?
Even then Bethesda, will have to create a new engine, or heavily modify, but then depending on how much modifying is required it may cost more to do that than it would be to start again.
The things audience want from games are becoming more refined. Like good pathfinding, and AI that responds to his mate getting blown away. These things are hard to do(they should be the standard)why not have them programmed already, so they developers can get to the focus of what their game is about and making it properly unique.