Development legend talks Rage, PC, consoles and future hardware

[coloumb]

Cool 20 minute video with the legend - John Carmack - at E3:

http://www.computerandvideogames.co...wares-john-carmack-20-minute-video-interview/

 

[KaOTiK]

He says:
"“When we started on the game six years ago, I looked at the consoles and said ‘These are as good as the PC’, and our development strategy was to develop live on all the platforms. And now we’re looking at PCs that have ten times the horsepower of the consoles. I’m making a large change in my direction just saying ‘We should be building things efficiently on the PC and then deploying on the consoles.’ And we didn’t make that as crisp of a distinction as we could have."

Amazing such a smart guy didn't realize this ahead of time lol

 

[andylawcc]

cool vid, look forward to Rage...

 

[WaitingForNehalem]

I appreciate him taking the time to explain why PCs aren't being pushed but his argument as to why a graphics powerhouse PC exclusive isn't feasible due to it costing $100 million and not being profitable seems a bit exaggerated. Crytek, a much smaller company, released both Far Cry and Crysis exclusively on the PC both with budgets much less than $100 million.

 

[ImpulsE69]

The way I look at it is bigger companies are lazier and have much more red tape. If you think about it from a time perspective -

Older games made by smaller companies took a long time to make, yet didn't cost them that much (in general). On the same note newer games from larger companies take nearly as long or longer to make, and cost millions more due to the many more people working on the game and marketing. None of that inherently improves a game. We all know throwing more money at a turd doesn't make it shine any brighter.

There is also the aspect that smaller companies are generally new and trying to get into the business and at that point most if not all are doing it for the love of it. Larger companies tend not to be this way. Maybe the grunts at the bottom are, but they are pressured by unrealistic timelines and budgets much more than a small company tends to be. These companies only care about the bottom line. Over the years, I honestly believe many companies have gotten to the point that they don't even care at all if the product is any good or buggy because of return policies on software. Once a game is purchased, whether good or bad, they've made their money. Sure reviews, etc could deter people from purchasing, but it's been shown time and time again that even shitty games make money if marketed right. I think "we" as a more technically inclined community tend to forget that most people don't care about the bleeding edge, or read reviews or if you can see a jagged line on the screen.

Before you go into the "games today have many more lines of code" argument, that extra code really is only engine and graphics improvements which falls on 1 or 2 departments. Once an engine is developed that engine is used for many projects typically.

Larger companies just are not efficient. Small companies learn to work with what they have to get the best of it.

---all of this is a generalization and opinion.

 

[Zenoth]

The way I look at it is bigger companies are lazier and have much more red tape. If you think about it from a time perspective -

Older games made by smaller companies took a long time to make, yet didn't cost them that much (in general). On the same note newer games from larger companies take nearly as long or longer to make, and cost millions more due to the many more people working on the game and marketing. None of that inherently improves a game. We all know throwing more money at a turd doesn't make it shine any brighter.

There is also the aspect that smaller companies are generally new and trying to get into the business and at that point most if not all are doing it for the love of it. Larger companies tend not to be this way. Maybe the grunts at the bottom are, but they are pressured by unrealistic timelines and budgets much more than a small company tends to be. These companies only care about the bottom line. Over the years, I honestly believe many companies have gotten to the point that they don't even care at all if the product is any good or buggy because of return policies on software. Once a game is purchased, whether good or bad, they've made their money. Sure reviews, etc could deter people from purchasing, but it's been shown time and time again that even shitty games make money if marketed right. I think "we" as a more technically inclined community tend to forget that most people don't care about the bleeding edge, or read reviews or if you can see a jagged line on the screen.

Before you go into the "games today have many more lines of code" argument, that extra code really is only engine and graphics improvements which falls on 1 or 2 departments. Once an engine is developed that engine is used for many projects typically.

Larger companies just are not efficient. Small companies learn to work with what they have to get the best of it.

--- all of this is a generalization and opinion.

To which I agree to a large extent.

Well said.

 

[Red Hawk]

I appreciate him taking the time to explain why PCs aren't being pushed but his argument as to why a graphics powerhouse PC exclusive isn't feasible due to it costing $100 million and not being profitable seems a bit exaggerated. Crytek, a much smaller company, released both Far Cry and Crysis exclusively on the PC both with budgets much less than $100 million.

That's not what he said. He said that if you were to make a PC exclusive that cost $100 million dollars to develop, you'd run yourself out of business. He's more implying that as good as Crytek and other top-of-the-line PC developers are, they could still do more if they had the resources, but going that extreme would be unfeasible.

Overall I think what Carmack had to say was very insightful. Hardly anyone can speak from the kind of experience in FPS development that he has.

 

[RussianSensation]

I appreciate him taking the time to explain why PCs aren't being pushed but his argument as to why a graphics powerhouse PC exclusive isn't feasible due to it costing $100 million and not being profitable seems a bit exaggerated. Crytek, a much smaller company, released both Far Cry and Crysis exclusively on the PC both with budgets much less than $100 million.

I think it's implicit in his point that going forward, in order to improve the fidelity of graphics beyond what's currently available, it will cost development studios in the millions of dollars more. This has been stated by a handful of top developers over the last 12 months. This just isn't viable on the PC as an exclusive platform. We have talked about this on our forum on several occasions. BF3 is going to look amazing on the PC, but at the same time it will also probably be the best looking game on the consoles too. If you google BF:BC2 on the consoles and PC, they look nearly identical. So it will be interesting to see how well BF3 on consoles compares to the PC version.

As to your point of FC1 and Crysis 1 pushing the envelope when they were released, the graphics were still far too primitive at the time. Doom 3 was easily one of the best looking games when it came out as well. However, now to go far beyond Crysis 1, you'll probably need 50 artists alone and require hardware that isn't even available (try adding DOF, dynamic lighting and Tessellation and realistic physics and you'll soon find yourself running at 10 FPS with Triple-SLI 580s). There must be a reason why so few developers are pushing the graphics envelope; and it most likely has to do with development time and $. Otherwise, almost all developers would just focus on making the best looking games if it was not that expensive. Surely, this would offer a significant comparative advantage among the competition. We can also deduce from that that the average consumer is pretty satisfied with the current state of graphics or they would have stopped buying the games on consoles and certainly on the PC. Yet, people still buy games since 2007 when Crysis 1 was released and only a handful of modern games look as good or better (Metro 2033, Witcher 2).

Finally, we don't know how much $ Crytek really made from Crysis 1. It could be that they spent a lot of $ upfront and barely made any $ off Crysis 1, but built up their reputation and brand so that they could secure development on consoles and finally make $ in the 2nd round. But already established developers aren't going to spend 3-4 years to design a game to "break" into the console market, since they are already there!

Also, Carmack mentions how it was extremely important for the game to maintain 60 fps. He doesn't deny the fact that they could have pushed the graphics far beyond what's currently available in Rage 3D, but then it wouldn't be running at 60 FPS for many years to come. Unlike Crysis 1 where you didn't really need 60 fps for smooth gameplay, id games are famous for the fact that they run very smooth and look good on release date. No one is going to care about Rade 3D if it runs at 15-20 FPS on a GTX580 today and it takes 5 years to max it out to 60 fps. I also agree with his point that while graphics are important, gameplay / creativity / storyline, etc. are far more important when it comes to making a hit game. I find that PC gamers tend to focus on graphics way too much and then complain that the game runs smoothly on a single GTX580. Not everyone wants to spend $500 on a GPU every 12 months.

 

[Aikouka]

I wonder if the industry would be better off if the majority of companies would stop working on their own engines and simply license one from id, Epic, Crytec, etc. Using UE3 has been fairly popular as of late... especially given how these engines typically support cross-platform (PC & console) play.

 

[KaOTiK]

I wonder if the industry would be better off if the majority of companies would stop working on their own engines and simply license one from id, Epic, Crytec, etc. Using UE3 has been fairly popular as of late... especially given how these engines typically support cross-platform (PC & console) play.

There is a lot more to it then that. A lot of devs uses a lot of custom in house software for a lot of things in their pipeline. Engines are generally chosen on the basis of how easily if at all can they make their custom software work with an engine and if that engine is tailored to what they need for the game. Then there is the engines licenses cost to consider, in some cases it is cheaper for them to develop their own age if they are planning to use it for a few games then to license one.

The big plus to use a license engine is you can hit the ground running mostly even if you have to customize the engine for your own needs a fair bit. It can save a lot of time for a dev to get a game out. That is the main reason you see the big publishing houses all licensing, they want to rush out their next sequel as quick as possible and if they don't already have their own engine made then they just license one instead.

 

[Red Hawk]

As to your point of FC1 and Crysis 1 pushing the envelope when they were released, the graphics were still far too primitive at the time. Doom 3 was easily one of the best looking games when it came out as well. However, now to go far beyond Crysis 1, you'll probably need 50 artists alone and require hardware that isn't even available (try adding DOF, dynamic lighting and Tessellation and realistic physics and you'll soon find yourself running at 10 FPS with Triple-SLI 580s). There must be a reason why so few developers are pushing the graphics envelope; and it most likely has to do with development time and $. Otherwise, almost all developers would just focus on making the best looking games if it was not that expensive. Surely, this would offer a significant comparative advantage among the competition. We can also deduce from that that the average consumer is pretty satisfied with the current state of graphics or they would have stopped buying the games on consoles and certainly on the PC. Yet, people still buy games since 2007 when Crysis 1 was released and only a handful of modern games look as good or better (Metro 2033, Witcher 2).

Epic's latest Unreal Engine demo seemed to do all that at a pretty smooth frame rate and a high resolution (2560x1440, IIRC). They did need Triple-SLI to do so, though.

 

[Aikouka]

There is a lot more to it then that. A lot of devs uses a lot of custom in house software for a lot of things in their pipeline. Engines are generally chosen on the basis of how easily if at all can they make their custom software work with an engine and if that engine is tailored to what they need for the game. Then there is the engines licenses cost to consider, in some cases it is cheaper for them to develop their own age if they are planning to use it for a few games then to license one.

The big plus to use a license engine is you can hit the ground running mostly even if you have to customize the engine for your own needs a fair bit. It can save a lot of time for a dev to get a game out. That is the main reason you see the big publishing houses all licensing, they want to rush out their next sequel as quick as possible and if they don't already have their own engine made then they just license one instead.

Just to clarify, I wasn't trying to state that there's no development necessary once you license an engine. Although, a lot of it will depend on the tools that are provided such as the Unreal Editor and things like that.

It's definitely something where cost will come into play, but you can also consider that previous work done with the engine can be put into use with another title as long as there are similarities. For example, if Gearbox makes a game similar to Borderlands, they may need to rework the entire world (all graphic work for the most part), but the underlying modified UE3 engine is still there. If we don't want to be that specific, there may just be bits and pieces (reuse is always nice ) that can be used.

 

[KaOTiK]

Just to clarify, I wasn't trying to state that there's no development necessary once you license an engine. Although, a lot of it will depend on the tools that are provided such as the Unreal Editor and things like that.

It's definitely something where cost will come into play, but you can also consider that previous work done with the engine can be put into use with another title as long as there are similarities. For example, if Gearbox makes a game similar to Borderlands, they may need to rework the entire world (all graphic work for the most part), but the underlying modified UE3 engine is still there. If we don't want to be that specific, there may just be bits and pieces (reuse is always nice ) that can be used.

I'm no programmer so can't say, just what I heard from them when I did work int he game industry lol.

Usually a studio will stick to the same engine they have already used if they can. Familiarity and already having custom tools made will the main reasons. They will just modify the engine to whatever new requirements they need as long as it is feasible on that engine.

In your example with Boarderlands. GB would most definitely use their already modified engine and do whatever tweaks that are needed in addition to it.

 

[Texashiker]

He says:
"“When we started on the game six years ago,

Usually, but not always, the longer it takes to develop a game, the crappier it is.

Doom and Quake both had fast development times. Over a decade later and they are considered classics,

Now take a look at duke nukem forever, that took over a decade, several companies to develop and the demo is getting pretty bad reviews.

There was an article awhile back that talked about software development times, as compared to the amount of people working on the project. The author said the smaller the team, the better. This led to less hands in the cookie jar, and less people in the decision making.

 

[wuliheron]

I've heard him talk about all this stuff before. The Id Tech 4 engine used in Doom 3 was the first megatexture engine and allowed the game to run well on even wimpy computers. It was an efficient engine, just like the original Doom and Quake engines, but the power of the hardware increased much faster making the graphics obsolete overnight. Games like Crysis exploited the power of the hardware more by anticipating it would come down in price.

If Carmack is going to continue leveraging the efficiency of the megatexture engines he needs to shift the focus towards maximizing the artwork and game dynamics, while still maintaining the high frames per second. With Rage he's managed just that producing a game with new dynamics such as racing, almost infinite textures and detail, yet can be played on an iPhone and still look great.

 

[WaitingForNehalem]

I think it's implicit in his point that going forward, in order to improve the fidelity of graphics beyond what's currently available, it will cost development studios in the millions of dollars more. This has been stated by a handful of top developers over the last 12 months. This just isn't viable on the PC as an exclusive platform. We have talked about this on our forum on several occasions. BF3 is going to look amazing on the PC, but at the same time it will also probably be the best looking game on the consoles too. If you google BF:BC2 on the consoles and PC, they look nearly identical. So it will be interesting to see how well BF3 on consoles compares to the PC version.

As to your point of FC1 and Crysis 1 pushing the envelope when they were released, the graphics were still far too primitive at the time. Doom 3 was easily one of the best looking games when it came out as well. However, now to go far beyond Crysis 1, you'll probably need 50 artists alone and require hardware that isn't even available (try adding DOF, dynamic lighting and Tessellation and realistic physics and you'll soon find yourself running at 10 FPS with Triple-SLI 580s). There must be a reason why so few developers are pushing the graphics envelope; and it most likely has to do with development time and $. Otherwise, almost all developers would just focus on making the best looking games if it was not that expensive. Surely, this would offer a significant comparative advantage among the competition. We can also deduce from that that the average consumer is pretty satisfied with the current state of graphics or they would have stopped buying the games on consoles and certainly on the PC. Yet, people still buy games since 2007 when Crysis 1 was released and only a handful of modern games look as good or better (Metro 2033, Witcher 2).

Finally, we don't know how much $ Crytek really made from Crysis 1. It could be that they spent a lot of $ upfront and barely made any $ off Crysis 1, but built up their reputation and brand so that they could secure development on consoles and finally make $ in the 2nd round. But already established developers aren't going to spend 3-4 years to design a game to "break" into the console market, since they are already there!

Also, Carmack mentions how it was extremely important for the game to maintain 60 fps. He doesn't deny the fact that they could have pushed the graphics far beyond what's currently available in Rage 3D, but then it wouldn't be running at 60 FPS for many years to come. Unlike Crysis 1 where you didn't really need 60 fps for smooth gameplay, id games are famous for the fact that they run very smooth and look good on release date. No one is going to care about Rade 3D if it runs at 15-20 FPS on a GTX580 today and it takes 5 years to max it out to 60 fps. I also agree with his point that while graphics are important, gameplay / creativity / storyline, etc. are far more important when it comes to making a hit game. I find that PC gamers tend to focus on graphics way too much and then complain that the game runs smoothly on a single GTX580. Not everyone wants to spend $500 on a GPU every 12 months.

From what I've gathered Crytek made a lot of money off of Crysis. Very few games have come close to Crysis's graphics let alone surpassed it. My point is that if a tiny company could make Crysis 4 years ago, then there is no reason other companies can't release better looking games today. Call of Duty: MW2 cost almost $50 million yet it used the same engine as COD 2 and was graphically unimpressive. Crysis cost $22 million and it was made on a new engine. Where is the money going? I certainly appreciate Carmack's insight on game engine development and have a lot of respect for him but I disagree with him on graphics being good enough today. As was pointed out in the BF3 powerpoint, developing graphically advanced games on the PC is good practice for next-gen consoles.

 

[finbarqs]

crysis dx11 is beastly... But it still didn't beat the sandbox'd crysis 1. But excelled in other fields: displacement maps are f'ing amazing in dx11 mode on crysis 2

 

[Stringjam]

From what I've gathered Crytek made a lot of money off of Crysis. Very few games have come close to Crysis's graphics let alone surpassed it. My point is that if a tiny company could make Crysis 4 years ago, then there is no reason other companies can't release better looking games today. Call of Duty: MW2 cost almost $50 million yet it used the same engine as COD 2 and was graphically unimpressive. Crysis cost $22 million and it was made on a new engine. Where is the money going?

I've often wondered this myself. CryTek develops a world-class engine (with an integrated physics engine) and builds a game on it all for $22 million.

Now we're in the era of $100M game development, and I honestly can't figure out where on earth they're spending all this money. Increased labor or overhead costs? Is expensive TV advertising figured into the amount? Maybe I can see a lot more money being spent on games with 50+ hour campaigns where tons of coding / scripting has to be done, but not on a 10 hour FPS.

Maybe somebody who has been involved in development will enlighten us.

 

[Texashiker]

Now we're in the era of $100M game development,

If a game cost that much money to develop, then it "must" be good.

After all, quality is all about how much something cost, right? The more expensive something is, the better it is?

On the flip side of the coin, the best things in life are free.

As far as I am concerned, id software has not made a good game since Quake 3. Quake 2 is ok at best, Quake and Doom brought about a revolution in FPSs. Doom 3 was a flop, at least in my opinion.

But for some reason, people keep looking up to id software like they make great games? id has not made anything outstanding in close to a decade.

Just because a game cost X amount of money to develop, then maybe people expect it to be a great game? From what I have seen and read about rage, I will wait until it hits the $10 mark before I buy. In all honesty, Dead Island looks more fun then rage.

 

[Blitzvogel]

If a game cost that much money to develop, then it "must" be good.

After all, quality is all about how much something cost, right? The more expensive something is, the better it is?

On the flip side of the coin, the best things in life are free.

As far as I am concerned, id software has not made a good game since Quake 3. Quake 2 is ok at best, Quake and Doom brought about a revolution in FPSs. Doom 3 was a flop, at least in my opinion.

But for some reason, people keep looking up to id software like they make great games? id has not made anything outstanding in close to a decade.

Just because a game cost X amount of money to develop, then maybe people expect it to be a great game? From what I have seen and read about rage, I will wait until it hits the $10 mark before I buy. In all honesty, Dead Island looks more fun then rage.

Expectations of Id arise from the warm feelings people have for their old classics, as well as their usual push on the technical side of things for each of their releases.

 

[KeithP]

Carmack says whatever he feels it is in his best interest to say at any given time. That is why he ends up saying stupid things like...

"When we started on the game six years ago, I looked at the consoles and said ‘These are as good as the PC’, and our development strategy was to develop live on all the platforms. And now we’re looking at PCs that have ten times the horsepower of the consoles. I’m making a large change in my direction just saying ‘We should be building things efficiently on the PC and then deploying on the consoles."

That was true 6 years ago (and probably longer) and everyone knew it at the time. I don't know why anyone pays attention to anything he says in interviews.

The games he develops and releases tell us all we need to know. Hey John, quit the yapping and work on releasing a fun, bug-free game for once.

-KeithP