Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 24, 2014
arrowPress Releases
October 24, 2014
PR Newswire
View All

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Interview: Crytek's Yerli On Why Graphics Still Matter
Interview: Crytek's Yerli On Why Graphics Still Matter Exclusive
April 9, 2010 | By Leigh Alexander

Following Crytek and EA Partners' unveiling of Crysis 2 in New York City, Gamasutra had the opportunity to chat briefly with Crytek co-founder Cevat Yerli to get more depth on Crytek's approach to the game.

Crysis was a game that a relatively narrow vertex of the market could experience due to its high-end PC requirements, and as such -- especially as Crysis 2 will be the showcase for the latest incarnation of the CryEngine development platform -- is it difficult for the game's developers to innovate on design when such a mandate has been placed on tech sophistication?

"Not really," Yerli tells us. "Creating IP is a challenge in any regards, whether this is a new IP -- or, as in Crysis 2, a reboot of an IP for different platforms. Challenge is always provided by the fact that we want to make an awesome game.

"We're known as a company that is about the leading edge in the tools and middleware business, and that's one arm of us," he continues. "The other arm -- probably the right arm -- is actually making games. It would be bad if you weren't making games, because then the tech wouldn't actually be proven as a game tool. What we're talking about absolutely works."

"Since the development of Far Cry, for me, it was always about getting efficient in designers' hands and empowering creativity," Yerli continues. "Technology fills in for what you don't do as a designer... you had to abstract the designer from the details, from the technology. There had to be an intelligence behind it."

People refer to those kinds of tools as "procedural," he says, "but it was always about abstracting an intelligent tool design."

Crytek will release Crysis 2 into an environment that wants to declare the high-budget console game a dinosaur in the face of digital and social gaming. A game like FarmVille, as is the common refrain of investors and venture capitalists, can garner millions of users and plenty of revenue without focusing on graphics at all.

"I think graphics matter a lot," Yerli asserts. "It depends on the type of game. If graphics and technology, AI and physics provide you a better experience but a challenge, then ultimately it does matter. You can get an experience out of Crysis 2 you just don't get anywhere else."

He poses: "The call to action for developers is look at your core experience and ask yourself: Is it worthwhile, the need [for] technology? Or is it just a matter of writing a Flash 3D engine and moving into other markets?"

"In my world of gamemaking, I want to deliver blockbuster, highly interactive, highly intense experiences for people that feel like they're in a Hollywood movie," Yerli explains. "And that's just not possible on another social platform where graphics don't matter."

It's "completely fair enough" that these markets exist, he says, noting the "huge potential" in other platforms and other gaming paradigms. "But a statement to say 'graphics don't matter' is to look at what you want to do, really. If I want to make an iPhone game, I can say it doesn't matter. If I want to make a Crysis 2 kind of game, I don't think there's a right or wrong answer to this question."

Moreover, Yerli sees "a lot of opportunity for higher end games" in the middle ground. "I believe if you look at emerging markets that are very big and highly growing... [there is] a potential strategic marriage of those platforms that can result in something bigger."

He urges: "Leave the tradition behind and look forward to what you can do with new technology... instead of looking at it as extremes."

Related Jobs

Digital Extremes
Digital Extremes — LONDON, Ontario, Canada

Digital Extremes
Digital Extremes — London, Ontario, Canada

Generalist Programmers
Petroglyph Games
Petroglyph Games — Las Vegas, Nevada, United States

Unity Engineer
University of Central Florida, School of Visual Arts and Design
University of Central Florida, School of Visual Arts and Design — Orlando, Florida, United States

Assistant Professor in Digital Media (Game Design)


Tim Carter
profile image
At this point, I'm more interested in standardized, fairly easy-to-use tools, with a solid well-established pipeline, and a variety of gametype frameworks I can plug in fairly easily. This supports artists, and art - as in the input of the artistic human hand - is probably the most important element that is now going to be your visual interest driver.

Tim Johnston
profile image
I second that Tim. TOOLS! We have enough graphic fidelity within the current game engine tech to create incredible experiences. So many are still so far out of reach learning-curve wise for most designers. Trying to learn Maya or 3D studio let alone a proprietary world building tool like UnrealEd or whatever is daunting and discouraging. I'm trying to do this now and, while the depth of Maya is incredible, so much of it is really not utilized if you are making game assets. Even a UI mod that simplified the GUI and only exposes those functions you would use for modeling/texturing would be great.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

E Zachary Knight
profile image
While yes it would be nice to have a standard set of tools, you really have to watch out for monopolies in tool production. If all game companies used the same 3d animation tools, the same world building tools etc, you will see a huge rise in price for those tools. Currently I think the cost of a 3D Studio Max license to be on the insane side of things. I can't imagine just how much more it would cost if it were the only tool used.

This non standardization is what has brought about things like the free indie license of Unity, the Free license of Unreal, the $200 license for Torque etc. This competition is what keeps the prices low and the cost of development low.

So if you don't mind tool companies pricing smaller development studios out of existence in the name of making thing easier for you, go ahead and standardize.

Tim Johnston
profile image

Point taken. But, I currently am a web designer and this industry is already very much the same situation. If you want to design something visually, you pretty much have to use photoshop. And, as you may know, a photoshop license (or suite license) is pretty steep. But, its the price we pay for standards, and compatibility when it comes to sharing and collaborating on assets (and also to offset massive piracy). Dont think for a second though I would mind an alternative that was as powerful, yet more affordable. Why couldnt these tool companies create versions of their software designed to stream workflow for specific applications (modeling/texturing only, rigging/animating only) ? This way they could perhaps focus on adding a usability layer that made doing things like say, mirror modeling, so much more intuitive, or unwrapping a model to do a UV or something. Adding "smart" features would simplify and streamline, but the fully featured tool is always available for those power users. This is not a "light" version of a program per se, but rather a version that takes into account a very specific work flow, skill level and intentions, and rearranges the UI, and functions to support them.

E Zachary Knight
profile image

Now that is something I can get behind. Breaking up these tools we use into smaller software packages meant for specific tasks could certainly help out. It could make things a bit cheaper for the studio as well. But I don't really see tool companies doing it though as right now they can sell a 3d Studio or Maya license for every artist in a company. They wouldn't want to sell just certain portions of the whole as those costs would have to be lower.

As for the Photoshop example, Yes it is pretty much the standard. It has been for a while. There was a time when it was just one of many quality graphic and photo editing tools. That era of competition is what led to be the quality it is. Now if only we could get some more competition to bring the cost down.

Merc Hoffner
profile image
Bearing in mind that I'm not a developer of anything gaming in the slightest, I'd ask you lot on the increasing expectations of visual scale? Of course there's only a certain detail level that's really required in a particular settings, and as you say, the attraction is injected by the artistic design: There's not much point in photo real if it's a boring photo. But then expanding the details that are observed over an ever growing expectation of visual assets must be a monstrous challenge. You've all seen it in the movies: After Jurassic Park they basically established they could depict more or less anything to more or less photoreal level in CG, but the expectations of model and scene quantity grew and grew. The Hollywood solution was basically to throw more artists at it - and the result was a ballooning production cost - justified by ever improving sales.

Take a look at pixar - Toy Story cost $30 million and featured about 2 locations (a bit iffy to compare so early in the timeline). Toy Story 2 is clearly more visually complex- or really I should say more asset dense, and cost $90 million. With Nemo they were comfortable with their tools and their process and it cost about the same. From then as the setting/scenario scale started increasing they got more expensive, and so Wall.E, featuring a junk planet and a city of robots cost nearly $200 million.

If Crysis featured an island, Crysis II features a city and presumably Crysis III will feature a planet, each fully decked out in an endless litany of unique and beautiful assets, won't the rising cost of human artists start exceeding the revenue a game can expect?

Let me pose a parallel problem: The robotic/high throughput revolution has Biologists accumulating so much data so rapidly that they can't analyse it all with a human eye for hypothesis, patterns, and meaning. Their solution has been to teach computers to automate basic analysis, with humans checking it over and making high-level analysis. With the advent of systems biology, the problem has become so complex that many are now talking of AI scientists, that can hypothesise, plan, experiment, analyse, interpret and iterate science at all levels, in massive scales with humans watching over. Could the same happen here?

Could someone establish a basis for an AI artist - where a human demonstrates an artistic style, and a computer replicates, varies, and embellishes in large quantity, with a little bit of neural network driven creativity, overseen by a human? If not, then I fear graphical gaming will die under the weight of salaries, even if all those artists are up on established, efficient and functional tools with the detail of an asset fixed at today's level.

Jesse Tucker
profile image
From Merc. "Could someone establish a basis for an AI artist - where a human demonstrates an artistic style, and a computer replicates, varies, and embellishes in large quantity, with a little bit of neural network driven creativity, overseen by a human?"

This is exactly what people are talking about with tools. Say I want to make a building in a city - how do I go about doing that? Well, First I'm going to have a picture in my mind about the general proportions, what it's made of (glass, concrete, etc.), architectural style - you get the idea. The current way to do this is to grab some available assets (perhaps you have a concrete texture and some premade doors) and start building everything by hand. Depending on how detailed you want the building, this process could take anywhere from an hour to a few days.

Now, let's say you have a "building tool" that takes a flexible algorithm, some assets (like the concrete, premade doors, etc.) and you define roughly what shape your building has. Tweak some sliders for concrete to glass ratio, hit some checkboxes for basic architectural style, and hit "Build". Look over the end result, make sure it does what you want, maybe dress it up with some radio antennae on the roof and you're done!

ken sato
profile image
@ Merc

While you bring up an interesting point, you seem to have abstracted the cost trajectories and development path linearly without taking into account that there is a feedback process in place. Costs per title tend to rise but that really is a problem in process rather than scope. If the scope is properly assessed, a budget is accurately tied to it and cost savings and limitations usually get discussed at this point. Staffing, which I agree is critical, is the major cost to the project with even a junior team member consuming a considerable amount of time AND impacting the development work flow of an established team. The team lead and producer for the project will take this into account according to the work at hand. Sometimes an asset doesn't need to be highly detailed and will have an acceptable low mip, or in another case there will be a trade off in presentation to performance. Both are 'big' in scope because they effect the entire project and therefore have to be balanced carefully.


Tools are critical to work flow which means there has to be some level of functional consistency from tool set to tool set so less time is spent familiarizing and more time spent working. The more 'custom' a tool set is the less likely its features will carry from project to project which is why high end tools are so necessary but also so costly. If you're lucky enough to be at a studio working on more than one project at the same time, a common tool set becomes even more cost effective. Now add that to a stable of studio owned developers that use a common engine, then the cost savings in toto can be considerable...and now you can have tool developers coordinate and work in tandem to customize and optimize the tool set. It does mean that your tools programmers go from least to moderate in project consideration since they will have cross project experience.

Finally, graphics are the principle way to display action that all interactive story telling depends. The 'fidelity' of those graphics, while visually stunning, MUST dovetail into the narrative, game play, and over all ease of use from menu design and flow to scripting cinematics and asset management. Large environments do not necessitate high quality assets but rather high quantity of assets and how they get used by the title. All of this needs to be assessed to how this will affect screen updates, checks, AND fidelity.

Crytek's engine looks impressive, I just need to see it in more and different projects to make an assessment.

steve roger
profile image
"Crytek will release Crysis 2 into an environment that wants to declare the high-budget console game a dinosaur in the face of digital and social gaming. A game like FarmVille, as is the common refrain of investors and venture capitalists, can garner millions of users and plenty of revenue without focusing on graphics at all. "

Just looking at recent big realeases, I don't see that graphical fidelity has taken a backseat to anything. Look at Batman, Infamous, Uncharted 2 to just name a few.

I don't think I agree with the article's premise. Where does this premise come from?

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

T Parks
profile image
Hmm, fyi... the Cry engine has always had a more user friendly interface/workflow than most competing systems. You guys are screaming "TOOLS!" yet crytek probably is probably on the forefront of offering more intuitive tools than anyone else in the industry.

David Cogan
profile image
@ T Parks: Absolutely. I've been learning to use the CryEngine 2 in a hobbyist fashion for the past couple of years now, and it's been an interesting and enjoyable experience. Aside from the gradual learning curve, ease of use and "WYSIWYP" nature of the Sandbox editor, Crytek also provide excellent support to their modding community.

They've released 2 good patches for the game (Crysis 1), 2 SDKs with source code and asset exporters/animation tools etc., and the game itself (and even simply the single-player demo) came with the world editor supplied. There is extensive official documentation available online for learning all aspects of the Sandbox and its extended editors and for working with the mod SDKs to create assets and code. There is also a good wiki incorporating all of the above-mentioned documentation plus additional official and user-created tutorials that are updated regularly.

The tools themselves look to be much-improved for the new iteration, also. The improved editor UI looks amazing. There are some good, narrated demo videos from GDC and the like kicking about in various places, like a good 5-part series over at GameTrailers.