Next Gen?

The Playstation / Nintendo 64 generation was a landmark breakthrough - suddenly 3D gaming on console was the norm.

The Dreamcast, PS2, Gamecube, Xbox generation was another, smaller, landmark - suddenly 3D looked good.  If you ported a previous generation game to the new generation, you got instant results - your textures were perspective correct, you stopped having a lot of sorting issues, yada.  At first, a developer could impress people without doing much, if any additional work.

You do a straight port of a current gen game to the new next gen - it looks identical.  Unless you're hooked up to an HD TV.  The burden is definitely on the developers to prove that these new systems are worth upgrading for.

I'd almost think that maybe this could be the generation where people wise up and realise - "Maybe I don't need a new game console" - but it won't be.  We'll see massive marketing budgets.  We won't be able to get away from the Xbox 360 and PS3 advertisements.  People will buy these consoles like hotcakes without even knowing why they should.  Or shouldn't.  And someday there'll be a killer app for them, that isn't available on a previous console, and I'll have to buy one.

Comments

[Factory]

"At first, a developer could impress people without doing much, if any additional work."
Hmm this effect will get stronger as tech progresses, slowly it will not be an issue of how well the HW is technically exploited, but of how it is artistically exploited. By this I mean such things as having x many polys onscreen ay 60fps will be less important than the quality of textures, and structuring of those polys in the models.
Of course this applies to graphics and sound, I think code has already passed that corner, and gameplay is now generally not held back by memory and processing limitations.

[Aaron]

"And someday there'll be a killer app for them, that isn't available on a previous console, and I'll have to buy one."

You say that now, but I bet you wind up getting one sooner rather than later. Practically everyone I know says the same thing as you, but almost all of them wind up conceding that they'll probably just get one or both (Xbox 360 or PS3) soon after launch anyway.

[Brian Krueger]

For the first time, I have no plan to pick up a next-gen console at launch, and I'm a gamer. The average consumer honestly cannot see the difference between a high quality Xbox game and GTA, which is crap graphically. I really don't think we're going to bowl people over with visuals this time out, especially if we go for photorealism and approach the uncanny valley.

What I'm hoping for is that we finally break the shackles of photorealism. Just as synthesizers make lousy guitars, they make for awesome new sounds. Where are my awesome new art forms that are enabled by computer graphics? Perhaps we can find a spark there?

I'll wait for the price to come down this time.

[Josh]

Do the next gens really look that identical? I don't think Unreal 3 would ever run off of a PlayStation 2 for instance. Perhaps current gen just made 3D look good, but the next gen might make them look great ... which has always been enough for the consumer. Heck, the entire PC game industry seems based on this premise.

[Jamie Fristrom]

Pretty sure I won't - I didn't with the previous gen. Ico made me get a PS2, Splinter Cell made me get an Xbox, and Metroid Prime made me get a Gamecube. None of those were launch titles.

[JP]

If you can see a "big difference" between current and next gen games (of which there are admittedly few non-total-bullshit examples of the latter), then chances are you are a game developer, graphics programmer, or someone who reads gaming publications religiously and gets taught all the little terms to look for ("normal mapping", etc). Unless I'm totally mistaken such people are very much in the minority.

I said a while ago that the hardware makers are going to have to resort to side-by-side screenshot comparisons to demonstrate the differences between current gen and next gen titles (LOOK, MORE PIXELS!!) and I stand by it.

The stage is set for game marketing's greatest hollow triumph yet.

[Fran]

do consumers buy consoles like they buy dishwashers? why does madden sell a million copies every year despite only incremental updates?

also i have to agree with josh. the returns are diminishing but the graphical arms race probably won't end just yet. my mom can tell the difference between these two:
http://image.com.com/gamespot/images/2004/news/02/18/ut2004/ut2004_screen001.jpg
vs.
http://image.com.com/gamespot/images/2005/152/928234_20050602_screen002.jpg

[JP]

Well, those two shots are a bit apples-oranges... a better comparison would have been a UT2k4 character against a black background with a custom lighting setup (the lighting especially makes a huge difference, outdoor light tends to be more even / washed out and less dramatic).

The big question is, is that a $300-400 difference to her? Especially when the games are going to be selling for $60 each?

Make no mistake, Microsoft is pushing the "HD Era" so hard because they took one look at how the Xbox2 stuff was looking on a normal 480i display, and realized hardly anyone would appreciate the difference. Cart drives horse, as it so often does in this industry.

[Aaron]

Jamie - you are much stronger willed than me. I wish I had the wherewithal to hold off on new console purchases. I think I'm just going to lay into the 360 at launch. (but I do get monetary incentive to do so as well)

[Josh]

"Well, those two shots are a bit apples-oranges... a better comparison would have been a UT2k4 character against a black background with a custom lighting setup"

OK. Here ya go: http://www.nextgen-engine.net/img/screens/malcolm.jpg

I definately see a difference, and it does indeed go beyond polygons. With these new engines there are new rendering and lighting tricks that a console like the PS2 simply isn't capable of. I don't own a HDTV and don't plan on buying one anytime soon, but I probably will get a nextgen sometime in the coming year.

[DrBitchy]

I do believe that we will see benefits right away even if we are running the next gen on our 10 year old tv.
I don't necessarily see the major benefit of the next gen being in poly count or in fps but just as much in the story itself, we will be able to pack way more hooks and supports into a scene that helps to increase the drams/action/horror of our games..

[JP]

we will be able to pack way more hooks and supports into a scene that helps to increase the drams/action/horror of our games..

Please explain how more powerful hardware will accomplish this.

http://www.nextgen-engine.net/img/screens/malcolm.jpg I definately see a difference, and it does indeed go beyond polygons.

Sure. I do think the lighting makes a lot more difference than the fidelity increases (higher poly, higher res textures). But the difference between the UT99 char and the UT2004 char is probably much bigger, to most people and especially to people like that guy's mom. So I think it's still accurate to say that we are at the point of diminishing returns.

Gamer d00ds have been trained to look for stuff like per-pixel lighting, just like they've been trained to ignore all the huge visual problems that games exhibit all the time - characters moving robotically, having weird or blank facial expressions, and inconsistent fidelity levels across the same space. Until we devote just as much attention to those as we do to squeezing in MOER PIXELS, MS et al's talk of making gaming into a wider cultural pastime is totally deluded.

[Josh]

"Please explain how more powerful hardware will accomplish this"

Because there's only so much you can shove to a screen with the current hardware at any resolution. Once again, not just polygons in general ... but objects as well. There's so many other dimension than just screen resolution that I don't even get what all this HDTV hubbub is about. I'd much rather see more realistically lit worlds, more interactive worlds, worlds with realistic clutter and crowds the just pumped up clarity.

What about the recent fluff comparing racing pics with real world pics? That's the kind of hype which will result in gamer demand asking their moms for these things. As for that guy's mom ... she ain't even gonna look at the back of the box.

[JP]

I was quoting, specifically:

"but just as much in the story itself, we will be able to pack way more hooks and supports into a scene that helps to increase the drams/action/horror of our games.."

What do you mean by "hooks and supports", and how exactly will newer, more powerful hardware lead to better stories/drama/horror etc?

[Josh]

Well I shouldn't have spoken for the doc, but my assumption was that scenes could be more interactive, more full, more anything, simply because you can do more scripting, have more objects, etc. But I could be off and maybe he meant something else?

[JP]

You don't need powerful hardware for any of those things. Katamari Damacy presented a completely new and different kind of interactivity, and it had a very simple art style, and used a very simple graphics engine. The physics-based gameplay was the innovative part, and that was pulled off very capably by the "aging" PS2.

[DrBitchy]

I understand fully well that we do not 'need' more powerful hardware to do these things.
Josh you spoke for me well. "scenes could be more interactive, more full, more anything, simply because you can do more scripting, have more objects, etc."
Sure many games at this time can fit in this explanation, but let your minds eye see how much more could be added into any one scene, more detail all around.

[Josh]

Katamari had a simple art style in part because of it's level of interactivity. Nor was it a true physics engine, nor was it even the level of collision the designer originally had in mind and several things which were cut because of the PS2 engine.

Now, you compare that to an online game with true a physics engine being used online with a fancy renderer and high polygon models ... it's a completely different beast. There's no way you could do that with the current gen hardware, you can barely do it with current gen PC hardware. It's going to take the next gens to deliver it. And having an immersive, interactive, online arena ... that's a pretty compelling reason to plunk down some cash.

[JP]

"you can barely do it with current gen PC hardware. It's going to take the next gens to deliver it."

You seem to have some pretty wild expectations built up around the next generation of consoles. They are no more powerful than any top-of-the-line PC I could put together, TODAY. I honestly don't understand why people think the new systems will be capable of anything more than that. Please stop buying the hype. We already have "immersive, interactive, online arenas", and if we make better ones in the future they are going to be driven primarily by innovations in design and craft rather than more megahertz.

"Nor was it a true physics engine, nor was it even the level of collision the designer originally had in mind and several things which were cut because of the PS2 engine."

1) The term "true physics engine" seems bankrupt of any meaning in this context.
2) Katamari Damacy was a very, VERY impressive technical achievement, all the moreso because of the hardware it was running on.
3) Would it really have been a *better* game if it had had that extra stuff? Design is not about adding stuff until it's great.

[JP]

"you can barely do it with current gen PC hardware. It's going to take the next gens to deliver it."

You seem to have some pretty wild expectations built up around the next generation of consoles. They are no more powerful than any top-of-the-line PC I could put together, TODAY. I honestly don't understand why people think the new systems will be capable of anything more than that. Please stop buying the hype. We already have "immersive, interactive, online arenas", and if we make better ones in the future they are going to be driven primarily by innovations in design and craft rather than more megahertz.

"Nor was it a true physics engine, nor was it even the level of collision the designer originally had in mind and several things which were cut because of the PS2 engine."

1) The term "true physics engine" seems bankrupt of any meaning in this context.
2) Katamari Damacy was a very, VERY impressive technical achievement, all the moreso because of the hardware it was running on.
3) Would it really have been a *better* game if it had had that extra stuff? Design is not about adding stuff until it's great.

[JP]

"you can barely do it with current gen PC hardware. It's going to take the next gens to deliver it."

You seem to have some pretty wild expectations built up around the next generation of consoles. They are no more powerful than any top-of-the-line PC I could put together, TODAY. I honestly don't understand why people think the new systems will be capable of anything more than that. Please stop buying the hype. We already have "immersive, interactive, online arenas", and if we make better ones in the future they are going to be driven primarily by innovations in design and craft rather than more megahertz.

"Nor was it a true physics engine, nor was it even the level of collision the designer originally had in mind and several things which were cut because of the PS2 engine."

1) The term "true physics engine" seems bankrupt of any meaning in this context.
2) Katamari Damacy was a very, VERY impressive technical achievement, all the moreso because of the hardware it was running on.
3) Would it really have been a *better* game if it had had that extra stuff? Design is not about adding stuff until it's great.

[JP]

ecch. sorry about the triple post.

[Josh]

"They are no more powerful than any top-of-the-line PC I could put together, TODAY."

Right. Hardware that costs about $2,000 - 3,000, or even more. You're comparing apples and oranges at this point. Most games don't shoot for that, because the vast majority of people don't have top-of-the-line PCs, whereas on a console they all get to shoot for that and most of the limitations are based on software techniques like compression, etc.

"We already have "immersive, interactive, online arenas", and if we make better ones in the future they are going to be driven primarily by innovations in design and craft rather than more megahertz"

No, we really don't the kind of immersion online that we could ... having worked with Karma and the current Unreal Engine and read what Epic has planned with UE3, it's just not the same playing field.

"The term "true physics engine" seems bankrupt of any meaning in this context."

There is an ocean of a difference between what Katamari considers physics and engines like Havok, Karma, etc. Yes, Katamari was an excellent game. I love it to death. When did this become an argument of design over technology? I'm simply saying that if I sit a PlayStation 2 game next to a PlayStation 3 game, there's going to be some pretty noticeable differences - and not just for the serious geeks out there.

[JP]

"Most games don't shoot for that"

But some do. Unreal3 does, and while it's very nice it's not going to *fundamentally* change anything other than Cliff's spring 2006 wardrobe.

"No, we really don't the kind of immersion online that we could ... having worked with Karma and the current Unreal Engine and read what Epic has planned with UE3, it's just not the same playing field."

You seem to be conflating immersion with graphical fidelity, and while the former can certainly contribute to the latter, they are simply not the same.

I'm sorry, but an online game with higher fidelity and slightly more physics feedback (most of which will be non-core-gameplay relevant like wind, and more decoration-ish objects in the simulation) is ONLY going to be significantly more "immersive" if you are a graphics whore. Normal people are honestly not going to care that much. They have not been trained by marketing and press to notice bullshit like pixel shaders and light blooms.

Normal people expect the game worlds they see to look and behave like the real one they inhabit. They get immersed by things like good artwork, good character animation, good lighting, good audio, and an interface that doesn't get in their way. I've already seen several "next-gen" games get these basic things wrong, and honestly all the polygons in the world won't save them.

We need to stop expecting technology to make things magically, inherently better. For the vast majority of game concepts out there, we are past the point where tech was holding us back from creating them. I brought up Katamari Damacy because it's a wacked-out, technology-driven concept that would have seemed utterly impossible 10 years ago, and it was accomplished with ease on fairly primitive hardware.

The continual, *increasingly desperate* fascination with the new hot shit is distracting and harmful.

[JP]

Also, I can't speak for Havoc, but Karma would crack up and die if you tried to do with it the things Katamari does. "Better", more generalized tech doesn't even necessarily create the most impressive results. Katamari's physics were simple and goofy (objects wobble like toys when you smash into but fail to roll them up) on purpose.

Don't even get me started on how underwhelming multi-core performance is going to be for early Xenon/PS3 titles.

[mjh]

Dunno 'bout you all, but I'm awfully tired of 32 megs of RAM. The next generation can't come soon enough.

[Josh]

"Unreal3 does, and while it's very nice it's not going to *fundamentally* change anything other than Cliff's spring 2006 wardrobe"

And current no PC game uses the UE3 engine. So, you're basically comparing next gen PC hardware with next gen console hardware. And no, it won't *fundamentally* change anything, but it will make a noticeable difference, which was the original point here.

"You seem to be conflating immersion with graphical fidelity, and while the former can certainly contribute to the latter, they are simply not the same"

Actually, I've pointed out several times that I'm talking about more than graphical fidelity. UE3 is capable of more interaction than previous engines as well, which is also a contributing factor to immersion and also gameplay as well.

"I'm sorry, but an online game with higher fidelity and slightly more physics feedback (most of which will be non-core-gameplay relevant like wind, and more decoration-ish objects in the simulation) is ONLY going to be significantly more "immersive" if you are a graphics whore. Normal people are honestly not going to care that much. They have not been trained by marketing and press to notice bullshit like pixel shaders and light blooms."

Well, first of all I don't belive you're correct in your assumption having seen what some mods have tried to do using the even the current technology.

Second, I think you're conflating how the industry works and how you would like for it to work. The games industry is largely comprised of graphic whores. Normal people do notice things like bloom, shaders, texture mapping, etc., it's just that geeks actually know what to call it. However, people can still see the difference and prefer the higher quality graphics.

"Also, I can't speak for Havoc, but Karma would crack up and die if you tried to do with it the things Katamari does."

I don't think that's true either. In fact, I'm not entirely sure you would even require Karma to do what Katamari does. You can thump the table with Katamari all you like. It's a great game, but it's certainly not indicative of the way most games are designed or certainly not even perceived by most consumers.

"We need to stop expecting technology to make things magically, inherently better"

Once again, you're confusing your arguments. I'm not saying which is better, I'm saying which is noticeable. What I questioned is that the difference between the console gens are visually imperceptible, which I don't think is true and honestly find very difficult to buy into considering the leap of engine technology which will be available to this new hardware.

But I'm willing to agree to disagree here as well.

[JP]

"UE3 is capable of more interaction than previous engines as well"

Please name an example. UE2 had physics. UE3 has physics. Out-of-the-box, the engine does not do anything significant or new in the interaction department. It's got better tools (which is absolutely very good for developers) and an all-new renderer.

If there is "more interaction" of any sort in a UE3 game, it's because a developer added it for their specific game, and tools being equal they could just as easily have added it in UE2 (the gravity gun ripoff for example).

Half-Life 2, for instance, is looked down upon by many tech-heads because its lighting model isn't as sophisticated as Doom3's, and it's lower-fi than something like Far Cry in many respects. It's less impressive as a tech demo, yet it is much more likely to appeal to a wider audience because it has good art direction, excellent characters, and lots of good design ingredients. The developers could have spent their time on other things, like Doom3 lighting, but they didn't. And the game is 1000% better because of it.

"Well, first of all I don't belive you're correct in your assumption having seen what some mods have tried to do using the even the current technology."

Well, I've been making mods for many years, and I work with UE3. So perhaps we'll have to agree to disagree. Also, doesn't the fact that they've done such things using "even the current technology" prove my point?

"However, people can still see the difference and prefer the higher quality graphics."

So if this really were true, Doom3 would be outselling HL2.

"In fact, I'm not entirely sure you would even require Karma to do what Katamari does."

That's exactly what I'm saying. The technology was driven by the design rather than the other way around. As a result, it performs much better for its intended purpose. (fyi, Karma uses unreasonable amounts of CPU when a large number of objects are involved in a single collision, so it would be totally unsuitable for KD where such collisions are the norm)

"The games industry is largely comprised of graphic whores."

So you are agreeing, then, that if the present emphasis on fidelity continues, the market will not grow, and under those conditions profits will dry up for all but the largest companies. If I'm being a bit idealistic, then I would say that you're assuming that things are the way they are now for good reasons. I'll trade you my utopian fallacy for your naturalistic fallacy.

Gamers very easily seem to lose sight of the fact that they are a minority. So when a niche game like KD manages to hit the overlap between gamer-dom and the rest of society, they perceive it as enjoyed by a minority. When in fact they are the minority. It's time we set our sights on places outside our tiny (though profitable *at present*) little bubble.

Notice I haven't said once that the extra fidelity the next generation will bring isn't "nice to have". But you have to look at the real costs of those increases. The next generation is going to have a profound effect on the development landscape because of the increased demands of creating content for the new engines. Budgets will go up, team sizes will increase, yadda yadda yadda same old story. The existing consumers will come to demand the new level of fidelity and production values. Many developers are going to die off because of this. The industry will become even more conservative and hit-driven, and will be comprised of fewer, even larger companies. This is bad for almost everyone in the long run.

So please understand where I'm coming from when I'm leery of the better-faster-more hardware arms race. We are allowing the hardware companies to drive the industry, and the artistic medium of games, off a cliff. There are alternatives. I would rather fight for them than drink in the eye candy and the hype, and line up for a Xenon on launch day like a good little robot.

[Josh]

"Out-of-the-box, the engine does not do anything significant or new in the interaction department. It's got better tools (which is absolutely very good for developers) and an all-new renderer."

Perhaps I'm giving Kismet too much credit, but it definately seems a significant tool enhancment. Plus my understanding was that Karma was getting largely ripped out and that the physics are were getting an overall with some emphasis on networking. Plus, I thought that UE3 was going to add new loading routines to allow for larger maps, more MMO-style worlds, etc.

But for instance, I had thought that some of the physics weren't even being designed for current gen hardware, but for things like dedicated physics chips. Least, that's how GameSpot reports it:
http://www.gamespot.com/news/2005/03/09/news_6120126.html

Course, there Rein also says a 6600GT will run it ... but we've heard that rhetoric before.

But I would say better physics engine + better tools to create scriptable objects = more interactivity than before. Perhaps I've read too much into Epic's hype machine however. Unlike you, I haven't had access to the engine.

"So if this really were true, Doom3 would be outselling HL2."

No, not necessarily. Once again, you're adding to my point. All I'm saying that someone could tell the difference despite having a technological background. Sales may be determined by many factors.

"fyi, Karma uses unreasonable amounts of CPU when a large number of objects are involved in a single collision, so it would be totally unsuitable for KD where such collisions are the norm"

I am aware of that, which is why I always ended up abandoning Karma in the long run. However, that was largely for MP play. For SP play you could easily do the same trick that KD does - which is to group objects into one object for the purposes of collisions/calculations/etc. I'm not saying KD isn't clever, I'm just saying it's not very taxing on hardware.

"So you are agreeing, then, that if the present emphasis on fidelity continues, the market will not grow, and under those conditions profits will dry up for all but the largest companies."

I don't see any reason why the current trend can't continue to grow. One can argue whether or not the resulting games will be actually better than the previous generation, sure. I'm on board there. And I can't disagree with your second theory - in fact it's one of my bigger fears for the game industry.

But like I said, that's really outside of my original point. Which is, subjective, I suppose, but I do think that the next gen consoles will look signicantly more handsome than our current ones, and they will sell in part because of them.

What recent mods have you worked on, btw? Unreal I assume? The last thing I actually released was Riftwar, but I had few projects in the hopper before playing with Torque2D instead.

[JP]

The interesting thing about Kismet is that it *doesn't* offer any incredible new functionality... it's merely a nice GUI to assist level designers in the same sort of Event-Tag scripting that they've always done. Which is still quite nice, don't get me wrong. I'll take better tools over more bells and whistles almost any day. But again it's not going to lead to "more interactivity" unless developers figure out new ways to harness what's already been there.

The physics has been re-engineered, but dedicated physics hardware will certainly *not* be required. The engine *may* end up supporting physics cards once they materialize, but I get the impression that's just Epic hedging their bets in case it ends up being a big new thing years down the road. Even then, though, I don't see it contributing much on the interactive side. Read Carmack's recent QuakeCon keynote, he knows what he's talking about... it'll be used for cloth, hair, grass, particles etc. We already do simple, good-enough simulation for this sort of thing where needed, and it's (usually) tertiary feedback as far as gameplay is concerned. Which makes me want to do a game where you cut and style people's hair, just to be perverse.

Physics over network play is going to continue to be problematic, I suspect, because it's a very fine-grain, data-rich simulation that doesn't degrade very gracefully under latency, but there are people who are much better qualified than me to comment on that.