Are next gen consoles feasible?

[joker454]

Really, I should reformulate this question. Do you think Sony, MS and Nintendo and possibly even Apple (they have the finances to do it) will release new consoles in the traditional mould of the past 30 years or so (since the NES)?

For now the traditional mould works but that will eventually go the way of the dodo as well. The pc/laptop situation is the clue here. People only need a device good enough to do what they need. For ages you needed a pc because a laptop was too much of a compromise. Now though laptops have gotten strong enough to where they are good enough for many to do everything they need, so the need for a pc dissapears. Consoles are the same, right now they are required to get acceptable quality graphics in gaming so people stick with them. But one day in the future there will be a level of "acceptable graphics" reached by consoles (we may get close to that next gen) to which the benefits of going past that aren't substantial enough for people to care. Eventually tablets or some other device will be able to recreate that level of graphics, and boom consoles will no longer be needed. It's like the pc/console situation with gamers, for many the graphical difference isn't big enough on pc's so they stick with consoles. Eventually the same thing will happen to consoles and they will be replaced by some other device. It's still a ways away but it will eventually happen.

Hence why it is so brutally imporant for the big boys like MS, Sony, Apple, etc, to establish vast software/service ecosystems today that lock people in due to the content they provide. It won't be hardware that ties people to a company anymore, it will be the software and services. Apple is ahead of the game here but Microsoft will join then in a huge way next year, they are in a better position than most people realize. Sony will need to partner with a software company otherwise in the future they will be squeezed out. Nintendo I suspect in the future will no longer be a hardware company and instead will be a value add "software service" on another providers ecosystem.

dont see that really having been a problem this generation. At all, once again.

It did force some big changes on the industry, notably the rise of the uber publisher like EA, Activision and Ubisoft. Even uber publishers though can only manage so much risk, and right now the costs of development are the biggest risk. I can't imagine them going up substantially more next gen unless there is a huge influx of new core gamers.

 

[ERP]

Hence why it is so brutally imporant for the big boys like MS, Sony, Apple, etc, to establish vast software/service ecosystems today that lock people in due to the content they provide. It won't be hardware that ties people to a company anymore, it will be the software and services. Apple is ahead of the game here but Microsoft will join then in a huge way next year, they are in a better position than most people realize. Sony will need to partner with a software company otherwise in the future they will be squeezed out. Nintendo I suspect in the future will no longer be a hardware company and instead will be a value add "software service" on another providers ecosystem.

Exactly it's about users not units at this point.
We're moving towards a software as a service model, it's just a question of how quickly.

 

[Shifty Geezer]

Maybe they'll make shorter games, or use smaller levels. I saw an article recently that said only a tiny percentage of gamers finish any particular game.

That's true. One issue is too little time, too many games. Episodic content makes more sense. Developers can produce smaller games for each episode, with less cost, and being able to gauge interest as to whether a sequel continuing the story is economically viable or not. The downside there is people who want to finish the game only to find it's discontinued partway through. Or worse, TV-style game creation with no story arc and the whole thing being made up as they go along.

 

[Ghostz]

I don't think it'll be that bad. You'll spend more time and thus money, condensing high quality source art to fit in today's consoles. I think if you can minimize the costs between the source art to in-game assets, the costs of man power's time in the art pipeline would peak.

So, I'm not ready to express the doom-n-gloom like everyone else. If I had to take a stab at what it costs to make a Pixar film ( which is about 60 million ) it has dramatically reduce compared to a cost inflation target - which is still about 60 million - even through the quality has gotten higher year after year.

 

[stiftl]

The hardware should be cheaper this round, especially for Sony , unless they develop a brand new chip again. Plus the Blu-Ray gamble isn´t needed again.

My hope is that Sony does a Cell 2 with whatever needed GPU, this time planned instead a stop gap solution.

That Microsoft goes back to the x86 and creates a more PC/XBOX friendly platform and helps to build a stronger PC gaming platform.

That Nintendo wins this round again

And that Apple just goes away forever and takes all it´s brushed arrogant aluminum with it, and it´s users

This man speaks the truth

 

[Shifty Geezer]

I don't think it'll be that bad. You'll spend more time and thus money, condensing high quality source art to fit in today's consoles. I think if you can minimize the costs between the source art to in-game assets, the costs of man power's time in the art pipeline would peak.

Reducing asset quality isn't a big problem. It takes many hours to model and paint an asset. Shrinking a texture is chaning a setting when exporting. Generating a hires displacement map is the click of a button. A $10 million game isn't $5 million asset creation and $5 million reducing the assets down to console quality, such that next-gen, that $5 million reduction cost will be saved!

So, I'm not ready to express the doom-n-gloom like everyone else. If I had to take a stab at what it costs to make a Pixar film ( which is about 60 million ) it has dramatically reduce compared to a cost inflation target - which is still about 60 million - even through the quality has gotten higher year after year.

That's true because tools improve. however, we have historical precedent in the games industry. There was discussion as to what BluRay would enable, and the likes of Laa-Yosh and friends said filling 50GB with content would be prohibitively expensive. This has proven to be right, with most games using a DVD's worth of assets. Generating three times the amount of content will cost three times as much. If someone creates some automated procedural content systems, those costs could be contained somewhat. Well, we were saying exactly the same thing this generation, and those magic procedural content technologies haven't appeared. I'm still optimistic about procedural content, but it's not a guaranteed solution. What we know for certain is, until some new tech appears, the cost of creating bigger, more detailed games will go up. Alternatively budgets will be capped and the games will be similar to the current games assetwise, but improve on lighting and shading etc. , which wouldn't look too bad. Chances are companies would compete to look the best, and there'd be something of a budget war driving up costs .

 

[austinwisetx]

Hi everybody. Thank you so much for all the ideas and suggestions gathered in this thread...it will of great help.

 

[Arwin]

Reducing asset quality isn't a big problem. It takes many hours to model and paint an asset. Shrinking a texture is chaning a setting when exporting. Generating a hires displacement map is the click of a button. A $10 million game isn't $5 million asset creation and $5 million reducing the assets down to console quality, such that next-gen, that $5 million reduction cost will be saved!

Maybe not, but I would be interested to see how much effort is put on this. If you see to what lengths companies go to to create tools that allow artists to test their art for engine burden and see their polygon and texture budget for art on that location, I think you may be underestimating this part somewhat? Given infinite resources, there are a lot of objects that could simply be scanned into your computer these days.

That's true because tools improve. however, we have historical precedent in the games industry. There was discussion as to what BluRay would enable, and the likes of Laa-Yosh and friends said filling 50GB with content would be prohibitively expensive. This has proven to be right, with most games using a DVD's worth of assets. Generating three times the amount of content will cost three times as much. If someone creates some automated procedural content systems, those costs could be contained somewhat. Well, we were saying exactly the same thing this generation, and those magic procedural content technologies haven't appeared. I'm still optimistic about procedural content, but it's not a guaranteed solution. What we know for certain is, until some new tech appears, the cost of creating bigger, more detailed games will go up. Alternatively budgets will be capped and the games will be similar to the current games assetwise, but improve on lighting and shading etc. , which wouldn't look too bad. Chances are companies would compete to look the best, and there'd be something of a budget war driving up costs .

I would quality this a little. There are a number of titles that stay within the size of a DVD because there is a multi-platform advantage to doing so. In some cases, you will see sound (and cut-scene video quality) reduced a fair bit on multi-platform titles because it's easier to have one set of art for both (or sometimes even all, hurting high end PC owners) platforms. On the other size, there are a fair number of high profile Microsoft releases that come on two discs, as well as high-profile Sony titles that use a big chunk of a BD's capacity.

Personally I believe next-gen will see a much higher emphasis on more efficient art creation. The constant rising costs of art-creation are partly because developers could afford to do it that way. Eventually though the costs should end up balancing out with those of movies, ranging from very low to very high, depending on the price of the sets and the actors (content creators) involved.

Crucially though the industry will need to learn that this whole range exists and the whole range is commercially viable, not just the high-end titles. Of course there are good examples of this already, and I have little doubt that we'll see a lot more growth in the small-to-mid size game range with relatively small teams involved (in that sense I am obviously not agreeing with a certain high-profile industry figure)

 

[Shifty Geezer]

Correct me if I'm wrong, but aren't the majority of larger titles made up of video? Do we have any ISOs or analyses that show games with 10+ GBs of models, textures and sound effects, rather than just massive video and audio streams? I'm not sure even the massive flagship titles push the amount of content that far, let alone this gen as a whole requiring 10x the storage of last gen. And it's not just multiplatforming that results in limited assets. Many PS3 exclusives aren't filling up the BRD with tens of GBs of levels and characters and whatnot. A Google of ISO sizes shows things like Uncharted 2 on a 22 GB disc includes 5 languages at 1 GB a pop, and duplicated data. I cannot think of any game off-hand that is bigger and more detailed on PS3 than a comparable game on 360, enabled by BRD.

When it comes to things like compression video and audio to lower quality to fit on DVD, that's not resulting in cheaper production costs. The cost of those cutscenes is the same, just it's compressed more. So DVD vs. BRD isn't saving money, and BRD isn't being used fully because it costs too much. Now next gen BRD will be a good fit because textures can be 4x the size etc., but we will be affected by cost limitations and we'll be waiting on advances in content creation to enable more and better content for the same costs. I'm unsure if we'll see enough processing power to shift enough content creation to the game engine in realtime to allevaite much by way of costs. I hope so, but then I hoped so this gen, with examples of goblins all varying by piecing together aprts, but that never happened.

 

[Laa-Yosh]

If I had to take a stab at what it costs to make a Pixar film ( which is about 60 million ) it has dramatically reduce compared to a cost inflation target - which is still about 60 million - even through the quality has gotten higher year after year.

Average Pixar movie costs way more than that, 90-130 million, and they're working with medium detailed assets compared to others. Actually a lot of that money is spent on R&D.

A more appropriate example would be Rango, that was about $135 million and has highly detailed artwork. They had about 120 characters, a city, some outdoor environments and underground caves.
Or I can bring up Avatar with the rumored 300-400 million cost...

 

[Laa-Yosh]

Rage is going to be the first game with art content in the range of 20-30 GB. Appropriately it took 5 years to build and I don't even dare to try to guess the costs.

 

[Rangers]

Rage is going to be the first game with art content in the range of 20-30 GB. Appropriately it took 5 years to build and I don't even dare to try to guess the costs.

Yeah but it's megatexture, a little bit different from the norm.

Also, whatever the costs it will be profitable otherwise ID will stop making games, yet I am pretty sure Rage 2 previews will be popping up soon enough. Take the constant game industry rising dev costs doom and gloom with salt.

Sure, there are rising dev costs, and I'm sure there's a winnowing factor there where small devs are priced out (or down to XBLA/PSN/iOS more likely) but until the flow of games slows to a problematic point I'm not worried the least.

Interestingly there probably may be SOME slowing of releases now, believe I read for July NPD there was 17 releases vs 26 last year. Then again that's one month so I dont know if it's true overall it's just my perception. but I think we could probably use some lessening the pace of releases anyway, I dont see it as a bad thing unless it goes too far, more a natural thing. There is such thing as too many triple A games that we dont have time to play.

For this fall for example I am good, better than most years with Rage, BF3, Halo Remake, and Gears 3. And I'm not even interested in Uncharted 3, Skyrim, MW3, Resistance 3, Assassins Creed Revelations, Deus Ex, Batman Arkham City, all huge games. Oh and I forgot about Madden. And didn't mention who knows how many "AA" games.

 

[Shifty Geezer]

Yeah but it's megatexture, a little bit different from the norm.

Only in format. 20GBs of texture detail is 20 GBs, whether one massive texture or thousands of small ones. Rage took a while though as iD had to develop their technologies, so we can't count every 20GB asset game as taking 5 years or similar.

Also, whatever the costs it will be profitable otherwise ID will stop making games

It may be profitable. If it isn't, perhaps iD will have to stop making games because they'll be bankrupt?

Also, where one dev can get a success by being the biggest, most expensive game, the moment other devs compete, that pushes the price of every game up. After all, why should Joe Gamer buy the 8GB Game X when for the same $60 price, he can get the 30GB Game Y? Not that people buy on sizes, but on the appearances those increased sizes afford, they can. Without an increase in the sale price, it won't be possible to fund more expensive game creations across the board.

Sure, there are rising dev costs, and I'm sure there's a winnowing factor there where small devs are priced out (or down to XBLA/PSN/iOS more likely) but until the flow of games slows to a problematic point I'm not worried the least.

I don't think anyone's suggesting the end of games because they cost too much. Just that there is an upper price limit on what devs/publishers can afford, which will define experiences and what the consoles will need to cope with.

 

[Laa-Yosh]

That's exactly why I've pointed it out. Other games in this generation simply do not have need for such high amounts of art assets because they're not uniquely textured or modeled.

Costs are a different issue, and you can't deny that they are very very high nowadays.
Just look at LA Noire, it sold what, 4 million copies? Yet the studio basically went bankrupt, now that they couldn't find a new publisher to finance their new project (the entire industry abandoned Team Bondi after their recent scandal). They're closing shop and selling their assets because they're out of money.
So it's not just doom and gloom talk, the average game is pretty expensive already and something like LA Noire is barely able to get in the black for the developers....

 

[Arwin]

Correct me if I'm wrong, but aren't the majority of larger titles made up of video? Do we have any ISOs or analyses that show games with 10+ GBs of models, textures and sound effects, rather than just massive video and audio streams? I'm not sure even the massive flagship titles push the amount of content that far, let alone this gen as a whole requiring 10x the storage of last gen. And it's not just multiplatforming that results in limited assets. Many PS3 exclusives aren't filling up the BRD with tens of GBs of levels and characters and whatnot. A Google of ISO sizes shows things like Uncharted 2 on a 22 GB disc includes 5 languages at 1 GB a pop, and duplicated data. I cannot think of any game off-hand that is bigger and more detailed on PS3 than a comparable game on 360, enabled by BRD.

I think that Uncharted had about 12GB of graphics data (textures, meshes, animations, etc.). I'm fairly sure that's about what GT5 has in terms of files (DF checked what was being copied to disc, which ended up being 12-13GB or so, and was just tracks, cars, etc.). Forza 3 also needed two discs with one installed to the HDD or you could forget about playing the Nurburgring for instance.

I think God of War 3 is also a very big game. 40GB if you leave everything in. Not every cutscene in that was video, either I think - I remember that you could play 5 different skins and that would reflect in at least the majority of the cutscenes.

When it comes to things like compression video and audio to lower quality to fit on DVD, that's not resulting in cheaper production costs. The cost of those cutscenes is the same, just it's compressed more. So DVD vs. BRD isn't saving money, and BRD isn't being used fully because it costs too much.

While that is true, doing something in high quality video form instead of in-engine does in fact save on production costs (you don't have to optimise those scenes for your engine, or run them in your engine but at a slower framerate to create a normal high framerate video) as well as hide loading times.

Now next gen BRD will be a good fit because textures can be 4x the size etc., but we will be affected by cost limitations and we'll be waiting on advances in content creation to enable more and better content for the same costs. I'm unsure if we'll see enough processing power to shift enough content creation to the game engine in realtime to allevaite much by way of costs.

More interesting would be to compare last gen DVD use with this gen DVD use in terms of mix of assets. Likely some of those could have fitted on CD as well, if you removed languages, videos, etc. That defeats the point a little though? And the biggest single disc game last gen not stuffed with video was a dual layer CD of 9GB (you know which one, it's one of your favorites and was an early example of a sort of megatexture ). Dual Layer BD is only about 5x as much as that.

A megatexture like Rage should in theory allow the creation of varied content faster rather than slower, once the tech and tools for it have stabilised.

I hope so, but then I hoped so this gen, with examples of goblins all varying by piecing together aprts, but that never happened.

There's still the comparison between the highly expensive creation of different animations for Assassin's Creed vs Uncharted 3. Similar advances have come from manual creation of loot to auto-generating loot in several games this gen. And of course user created content is becoming a big thing. With stuff like Kinect, people can even create their own custom animations. And LBP is a poster child for it to some extent, regardless of your frustrations there ... they've basically crowdsourced their level creation very successfully. But also Gran Turismo's track generator is an example of procedurally generated content, never mind ModNation racers (you can get all the F1 tracks for that from users, all Mario tracks, all Gran Turismo tracks, etc.).

I think if you make a good list, there's a lot showing the future of content creation. That's why I think it will become such a big thing next generation. Then it will probably also take the shape that is being discussed here and there in podcasts currently also, in that online interaction with other players will start playing a bigger role. In MMOs for instance most NPCs you encounter are in fact other players, which creates a world with intelligent NPCs doing their thing. Demon's Souls and Journey are also interesting examples of that concept.

I think another crucial point, also in relation to what Laa-Yosh said, is that the resolution went up more than the RAM or disc read-speed should have in comparison. Next gen resolution won't change nearly as much (games can stay at the same resolution basically), but RAM should go up considerably, and hopefully read-speeds go up significantly as well (though more RAM compensates for worse read-speeds some as well of course and vice versa).

The upshot of which should be that texturing budgets can go up significantly.

 

[Laa-Yosh]

. Rage took a while though as iD had to develop their technologies, so we can't count every 20GB asset game as taking 5 years or similar.

How exactly do we count asset space anyway? Rage has terrabytes of source art that gets compressed to that 20GB, and we know they're using far more aggressive compression methods than most other game devs. It's not the same work as an Uncharted game with 20GB of art assets would be, but it's hard to decide how much more it is, since ND's artists may have to spend a lot of time on being clever because of their tech's limitations compared to Rage.

Also, where one dev can get a success by being the biggest, most expensive game, the moment other devs compete, that pushes the price of every game up.

Doom 3 is the perfect example for this. You could not afford not to do normal mapping once it was released, and it raised content creation costs for everyone in the industry. I'm sure a lot of producers and studio heads were cursing id at that time.

Now unique virtual texturing may soon become a necessity as well. We'll see how the public reacts to the game...

I don't think anyone's suggesting the end of games because they cost too much. Just that there is an upper price limit on what devs/publishers can afford, which will define experiences and what the consoles will need to cope with.

This is actually an interesting question, I might post some thoughts on this later...

 

[Arwin]

Also interesting for this discussion is the job description Smooth Devil just posted (emphasis mine):

A Santa Monica studio is looking for a Senior Software Developer to implement and improve the existing in-house pipeline and toolset, as well as integrate and maintain various 2D and 3D third party software applications.

Job Duties:

The position involves design, implementation, testing and debugging of productivity-enhancing tools for artists and supervisors. Additionally the role requires strong technical and problem solving skills, experience in software design, graphics, and production workflow. Excellent communication, attention to detail, and strong organizational skills are critical. Productive interaction and discussions with the rest of the development team is paramount.

Job Requirements:

- 3+ years Film and or TV production experience
- 3+ years experience in a software development environment
- Bachelors or Masters in Computer Science, Mathematics or comparable professional experience
- Proficient knowledge of Python, C/C++, MEL, SVN, Bash, Perl
- Significant experience with Linux, OSX, Windows, QT libraries, Boost libraries, OpenGL
- Working knowledge of Maya, Houdini, Nuke and other VFX-related packages and APIs
- Ability to meet timelines and work under pressure while still selecting the best methods and techniques for the positive progression of the facility pipeline and toolset.

http://www.smoothdevil.com/index.php...ob&job_id=1661

closes: 22 Oct 2011

 

[Laa-Yosh]

It isn't anything new. UE3 got big mostly because they had awesome tools that could be used to speed up production, well before the engine was completed.

 

[Shifty Geezer]

I think that Uncharted had about 12GB of graphics data (textures, meshes, animations, etc.). I'm fairly sure that's about what GT5 has in terms of files (DF checked what was being copied to disc, which ended up being 12-13GB or so, and was just tracks, cars, etc.). Forza 3 also needed two discs with one installed to the HDD or you could forget about playing the Nurburgring for instance.

There are always peak examples. One can point to Crysis as see a need for 8 GBs RAM. Someone somewhere will always use fill whatever avaialble resources, but business in general has to be more moderate. GT5 can get away with however much content they want, because Sony will fund PD indefinitely so they can take 5 years to create content, plus they have a guaranteed market so they can be suire of the ROI. Any other racing game won't have that luxury.

More interesting would be to compare last gen DVD use with this gen DVD use in terms of mix of assets. Likely some of those could have fitted on CD as well, if you removed languages, videos, etc. That defeats the point a little though? And the biggest single disc game last gen not stuffed with video was a dual layer CD of 9GB (you know which one, it's one of your favorites and was an early example of a sort of megatexture ). Dual Layer BD is only about 5x as much as that.

Well we had all this discussion back regards BRD, and whether devs would or wouldn't use it. Fringe cases aside, this gen has got by with DVD capacities, same as last gen. The amount of content increase hasn't seen the same degree of growth of aspects of games. That's perhaps played out by the idea (unsubstantiated?) that games this gen are shorter.

A megatexture like Rage should in theory allow the creation of varied content faster rather than slower, once the tech and tools for it have stabilised.

Not so sure about that. Each pixel drawn is a pixel drawn. There'll be saving regards structuring the textures in the level, but there'll could well be a larger cost increase in adding more content. Some devs would try repeating the same texture and using megatexturing as a cheap way of managing the environment, but others will paint unique detail trhoughout, pushing the bar for what consumer expect, and driving costs up.

I think if you make a good list, there's a lot showing the future of content creation. That's why I think it will become such a big thing next generation.

That's a whole other discussion I've been championing already!

How exactly do we count asset space anyway?

That's actually a vital question. The cost is creating the masters. 20 TBs of unique textures compressed down to 8 GBs or 25 GBs is still going to cost the same on the creation side. I guess we'll never have that info.

 

[tabs]

Bringing it back to the Rage taking 5 years... optimizing the workflow is Carmack's big focus. He seems to indicate that while more art is created, it is easier to create using megatexture. The length of time it took to build rage was apparently more to do with it being a very new approach, and some intitial assumptions being off. JC says that subsequent IDTech5 games won't be anywhere near as time-hungry.

How much artist time does it take to scale stuff down for actual use in game from their higher detailed creations? I would imagine there's quite a few versions of most models and textures being made just for this alone, as performance targets are trying to be hit.

What if the artist can just make his hi-res piece then just move on to the next thing?


/edit I see Shifty beat me to it