Do gamers really want photorealistic graphics? [update 1]
by Conrad Quilty-Harper 
{ Dec 14th 2005 at 9:00AM }
Epic has released a bunch of new screenshots for its upcoming game Unreal Tournament 2007, which will be the fourth in the series. Naturally, the graphics are astounding and the example above doesn't look too far off the high end CGI example we posted yesterday. We could just leave you to drool all over your keyboard at the next next-generation of graphics, but that'd be too easy. Instead we're going to recount a couple of the comments left at the previous CGI graphics article.
As Anon points out, "at the rate the price of games are going, the amount of time taken to model this will have to pay the artists 10 fold". We touched on this earlier, where we looked at Nintendo's patent for displacement mapping graphics. The problem of increasing programming and artist time is one that a vast section of the industry is going to have to address. As budgets and graphical expectations (from the greedy gaming public) rise out of control, publishers and developers are going to have to make a choice: graphics or gameplay.
Jake believes that "Half-Life 2, and all those other games, in fact every single "photorealistic" game look cartoony. No matter how many polygons they get in, Alyx still looks cartoony." What Jake's saying here is that photorealistic graphics will never move past the "Uncanny Valley". This is a principle of robotics that addresses the emotional response we have to human-like robots. As the realism of the machine (or game) approaches a certain point "then the non-human characteristics will be the ones that stand out" and we'll start to feel uneasy (around the robot) or detatched emotionally (from the game character).
It's clear that there are lot of problems involved within the field of super realistic graphics, two of which we've looked at here. Other factors include the quality of gaming hardware, slower progress in the field of animation and most importantly: the need to focus on gameplay. After all, don't we play games to have fun?
[Via digg]
Update: fixed some strange errors in the post (commas/apostrophes were broken).
Reader Comments (Page 1 of 1)
Timothy Dorr @ Dec 18th 2005 9:42PM
Ah, here we go. This is what I was talking about:
http://www.unrealtechnology.com/screens/character_creation1.jpg
http://www.unrealtechnology.com/screens/character_creation2.jpg
http://www.unrealtechnology.com/screens/character_creation3.jpg
Timothy Dorr @ Dec 18th 2005 9:42PM
Wasn't the Unreal Engine 3.0's claim to fame the fact that it *doesn't* take additional designer time to get the same results? Remember EA quoting $30m per next-gen game and CliffyB or Epic's CEO or somebody scoffing at that?
Most games are modeled and textured with very high detail in high-end graphics programs. They are then scaled down to fit the hardware they're running on. The detail is normally already there, you just don't scale it down as much for these new games. Look around on the Unreal Engine site and there should be some pictures of a high detail model and one scaled down, both of which look essentially the same.
Charles @ Dec 18th 2005 9:42PM
Perhaps the biggest problem that you didn't mention is the problem of AI. Even the most complex game AI's still only allow for a tiny set of actions. As character start to look more like real people, these scripted actions and responses are going to become more and more jarring.
Gimbal @ Dec 18th 2005 9:42PM
I would like photorealistic graphics, yes. Necessity being the mother of invention that it is, the tool set for creating photorealistic graphics will decrease some of the development time. My only problem with photorealistic graphics right now is how rare you see it properly paired with realistic motion.
I don't care how photorealistic it looks, if a character's feet float all over the ground or has that herky-jerky motion in its animation, you are just defeating the realism.
The last thing I want to see is a jaw-droppingly realitic character at a full run suddenly change direction 180 degrees without any consequence of inertia.
I mean, come on...even Super Mario had to deal with inertia.
But in the end I guess I can't really complain too much. Let them rocket their way to hyper-realism. Once they get there they can finally be done with that and concentrate on bringing the physics and evironmental interaction up to match it.
Clutch @ Dec 18th 2005 9:42PM
I think we look at the whole "graphics debate" issue the wrong way. Do we as gamers really want photorealistic graphics? Sure, I think we'd all like our gaming systems, be they console or computer, to be capable of that feat, but I for one certainly wouldn't want or expect every game to be like that. There are games that simply would not benefit and might very well suffer from this kind of visual representation, and I think that it's silly for people to act as though every game would look the same. One might cite the ubiquity of cel-shading about two years ago as a counterargument, but I pray designers have learned their lessons from that.
Forrest @ Dec 18th 2005 9:42PM
According to "Raising the Bar" (the book on HL2), Valve intentionally rolled back the level or realism in their character design to get away from the uncanny valley thing.
Timmay! @ Dec 18th 2005 9:42PM
Does the question even need to be asked? It's gameplay! Why are some of the older games still considered the best, when they have graphically beefed up counterparts on next gen systems? It's because the gameplay was top notch and made the whole experience enjoyable.
How many times have we heard this in a review, "It looks pretty, but the gameplay is non-existent."
t-ready @ Dec 18th 2005 9:42PM
The [H]'s article on the direction that Valve's taking with their cinematic effects kind of deals with the point you're trying to make as well. Ie. intentionally blurring-down (is that a word?) the details in order to acheive an effect that actually draws the user deeper into the environment.
Less details = more realistic gameplay? I'm definitely interested to see the implementation of the rest of the effects they describe in the future...
I completely agree, though, that focussing on the graphics & not the gameplay has definitely brought us to where we are now; however, the sad fact of the matter is that right now, eye-candy sells.
Truth @ Dec 18th 2005 9:42PM
I have 3D Studio Max 6 for my computer at home that I use to model characters like the pics #1 posted and anything I can think of. It cost about $1200, good thing Boeing paid for it. It is an awsome program just go to www.autodesk.com to view the product. You need a fast computer and some modeling experience to help run it. But once you get the hang of it you can program you own games, model anything you can think of. They have version 8 out now.
But anyways back on subject. They will never let games get to the point where you can no longer tell the difference between a game and reality. The reasons why: Dev cost, Time & Lawsuits.
vrf @ Dec 18th 2005 9:42PM
I agree with Gimbal that I'd far rather have something "behave" realistically than look hyper-realistic. Eye-candy sucks if nothing behaves in a realistic manner.
Unfortunately, you need smart coders and programmers to do that. Artists are much easier to find.
Morder @ Dec 18th 2005 9:42PM
it's all about the atari, nes & snes...graphics doesn't mean anything, it's all about the game play...
if a game is fun, or challenging, it doesn't really matter what the graphics look like. All the graphics do is capture the players attention to buy the game, "whoa, look at that!" kinda crap, then when they find out what a crappy game it is, it collects dust in a drawer or something.
yes, graphics are important, but not so much so that it's what the game should be about, realistic anything, just make the game fun
Jeremy Wright @ Dec 18th 2005 9:42PM
As a note, there is nothing to say that increased detail means increased costs or time. That's what toolkits, engines and 3rd parties are for.
As an example, some of the effects in LOTR would have cost more than 30M$ 10 years ago. They could have been achieved, but at a stupidly expensive cost.
However, the actual shots cost less than 1M$ per shot. And the detail was 100 times what thye would have achieved 10 years ago.
In 10 more years, the abilities of special effects crews will be unparalleled.
Companies like ILM already release toolkits, as does Disney. These will become more and more prevalent in the gaming industry. The graphics engine will only be part of it. You'll be grass, tree and wind engines, sun, atmosphere and gravity engines. You'll buy facial expression and voice-to-facial-movement engines.
These will add 3-5M$ in costs to a game, yes, but they'll also mean you can go from just an engine to full-out environments in 2-3 months (and, if you hire a modelling 3rd party for an extra 2-3M$ you can even have all your models done in that time frame).
Toss in AI engines (which are starting to come into their own), new scripting engines and the games of tomorrow's tomorrow will be available for roughly the same cost as today's games. Probably slightly more, really, but that extra cost will be offset by much higher distributions as well.
One of the things too many people forget is that yes, while many games these days cost 3-5 times what they did 5 years ago there are also 50-100 times as many gamers out there.
The challenge for developers isn't "how do I afford this", it's "how do I build my company to the point where I can make a truly high-end game".
You wouldn't start a movie studio and expect to make LOTR as your first movie. You'd do a half dozen budget flicks, then a half dozen mid-range flicks and then one or two blockbusters.
The gaming industry will go the same way, not only because Hollywood designed the model 40 years ago, but because it makes sense. Business is business.
Sasso @ Dec 18th 2005 9:42PM
I am glad that this topic has been opened up.
It seems to me that we are very likely to see a very similar thing in video games that we did in fine arts. If you go back and look at the progression of painting, before the advent of photography, you will see that there was a steady quest to duplicate reality. Perspective, proportion, light and color were all developed over time to yield better more realistic results. This continued until the advent of photography. When it was finally possible to duplicate reality perfectly without the great amount of time spent, the alure was gone. And then there was a backlash - cubism, dada, modernism.
It seems to me that games are on the very same path. Right now it is very time consuming to render reality in the game world. I just don't believe it will be that way for long. Advancements such as this displacement mapping patent are just the beginning. I believe in the near future hyper-realistic video games will be common and as easy to produce the art for as it is to snap a picture.
As an artist I am excited at that notion. Not because I would like to see hyper-realism in games, just the opposite in fact. I think we must reach the goal of total realism before any deep creativity can be explored. I think at that time we will see some truely expressive and inovative artwork in games.
Iheartgraphics @ Dec 18th 2005 9:42PM
I don't feel bad for developers (having to spend all of that money boo hoo) in the slightest.
I don't care how many developers it takes and how many hours they have to put into it. Theres new potential team members with wicked skills being belched out of schools every semester...find one!
I don't care about having to dish out $100 a title to see graphics that are photo realistic.
If you have someone that has a PC that can push that kind of graphics...chances are money isn't really an issue with them.
I want photo realistic zombies to scare the s*** out of me and give me long term psychological damage.
I want to feel like I'm there, I'm the player, in a sports game.
I want to feel like I'm actually racing at 220mph.
WHY do I want these things?--BECAUSE--these games get me the closest I will ever be to these situations in real life!
P.S. Developers know great graphics doesn't mean a great game...just to let ya know...there doesn't have to be a compromise.
I WANT INSANE GRAPHICS!!..why is that so wrong?!
ZaBlanc @ Dec 18th 2005 9:42PM
Graphics are impressive, and an easy way to market your game (see every WWII game released in the last few months) -- but once you've plunked down the money, the Publisher will always be happy, but you'll only be if the game is GOOD.
I'll take my $25-$30 DS games, with their fun factor, over a $60 PC or XBox game anyday now.
Chris @ Dec 18th 2005 9:42PM
I think it's a bit silly to suggest that there always has to be a choice been graphics and gameplay. As the industry evolves, publishers will start sharing and/or licensing physics engines, character models and AI technologies to each other. It's already happening on some levels. Once all of these components become commodities, then game developers can increasingly focus on gameplay.
I am not sure the problem with the industry is the race towards realistic graphics; rather, it's the fact that the market is flooded with so many games. It used to be that the A-List title were obvious. You HAD to play Goldeneye; you HAD to play Metal Gear Solid; you HAD to play Zelda. That's changed. There's too much out there. And there's rarely anything to falls into the "must-play" category.
This is likely because games just don't offer anything new these days. With few exceptions, developers are generally under the gun to deliver games for a specific date--usually the Christmas season--so that the game's publisher can get the game on shelves and hopefully meet earning expectations. Good games take time to get to market. And I'd rather see a developer hold back the game to do it right -- and make it extraordinary -- than barf it up in time for Christmas.
Zack @ Dec 18th 2005 9:42PM
Why can't we have best of both? At least have like 60% gameplay, % 40 graphics. And instead of just focusing on photorealistic wrinkles on humans, why not on world-building?
Why not have say, GTA style space opera-themed RPGs where you go streaming from planet to planet seamlessly without any load time ala Dungeon Siege, and much like entering a building? Meaning, you pilot a spaceship, it'll enter that planet you desire to visit, planet becomes bigger as you move towards her, you enter atmosphere, once atmosphere breached, you began to see cities, landscapes etc. and you park right ahead at nearby landing pad. And it's part of the gameplay!
XanderC @ Dec 18th 2005 9:42PM
Honestly, I think the games that try to be photo-realistic look like crap.
Seriously, if I want to see something that looks like it does in real life, then I'll pick up a damned camra. The whole race for "photo-realism" thing does nothing but stagnate the industry and make all games look exactly the same.
As far as I'm concerned, Rez, Electroplankton, and ICO look WAY better than Half Life 2.
King Bigos @ Dec 18th 2005 9:42PM
Do gamers really want photorealistic graphics?
Yes, yes we do.
32_Footsteps @ Dec 18th 2005 9:42PM
Personally, I'm trying to figure out what it means that ESPN actually discussed this a couple weeks ago.
http://sports.espn.go.com/espn/page2/story?page=hruby/051122
It's all about the game play to me - but I know full well that there are enough superficial people in this world that graphics are an obsession of the industry.
emehrkay @ Dec 18th 2005 9:42PM
as a tech fan, i would like to see it, but when would it happen. with this round of consoles, i was expecting "toy story-like" graphics straight from the gate. like the one guy said, everything is modeled with very high detail, poly count, then made to fit whats available. i think photorealistic is far far away. are there many artists that can do photorealistic?
Reeve @ Dec 18th 2005 9:42PM
I'd like to touch on something #4 said. I too feel that realistic/believable animation is more important than realistic graphics (then again, I'm an animator, so call me biased if you want). However, great gameplay is still king.
However, there are a number of graphics snobs amongst gamers (I should know, my fiance being one of them), so there will always be a need to appeal to that sector. Just so long as there are still fantastic games out there-- exquisitely rendered graphics or no-- then I have no complaints.
Matt @ Dec 18th 2005 9:42PM
If photorealistic graphics was that important to people hand helds like the GBA,NDS,& PSP wouldnt be doing good.
Mullinator @ Dec 18th 2005 9:42PM
#4 Gimbal
I totally agree, I was just re-playing Zelda OoT the other day and I noticed that despite the simplicity of the 3d models the animations that Link has are surprisingly more realistic than some of what I have seen on many modern games. Just look at how Link gets out of bed when you start a new game, I was actually very impressed.
Stakker @ Dec 18th 2005 9:42PM
What really makes game characters feel "realistic" or belivable and immersive is ANIMATION. It doesn't matter if the character is made of cardboard boxes, if it moves like a human being, it starts to feel alive.
Right now we're seeing too many games where high-poly normal mapped characters run around switching from animationloop1 to animationloop2 robotic-like, sliding and running against walls and doors, not reacting to environment at all, feet sliding or floating over ground, feet not touching steps on chairs etc etc etc... This is where the real improvement in next-gen should happen. I just fear that it won't, because it doesn't translate well to screenshots :-/
matt @ Dec 18th 2005 9:42PM
My biggest fear isnt with photo realism per se. I personally have no problem with it, but instead I fear that with the fevered focus by everyone to want to achieve photo-reaism (developers and gamers) we will lose all of the whimsical or truly artisitc games.
Will we ever see another Katamari? Will we ever see another Dig-dug? What about a Parrapa the Rappa?
Probably not.
Mazinkaiser @ Dec 18th 2005 9:42PM
I'd prefer companies to put more effort into making games fun again. Sure great graphics are awesome, but that seems to have become the main focus. Make that damn games fun, then work on the cosmetics! Look at GBA, the games look like SNES but they're fun! I'd rather play an ugly fun game then an eye candy bore.
Lagomorpho @ Dec 18th 2005 9:42PM
matt, I think that's what Sasso was getting at. Once we achieve photo-realism, there's really no place to go so we'll go back to simpler and more abstract games.
All this talk of animation vs CGI reminds me of this computer animation class I took. My professor had this java applet where you could manipulate two characters made of geometric shapes and they looked more like people walking than some of the people I've seen in games today. here's a link: http://mrl.nyu.edu/~perlin/experiments/emotive-actors
There's lots of other neat stuff there if you look around.
Truth @ Dec 18th 2005 9:42PM
UT2007 graphics already look good enough. You really don't need to update them more. Just focus MORE on AI, atmosphere, animation and gamplay.
matt @ Dec 18th 2005 9:42PM
@27
I think Sasso and I are getting at slightly different points. While he is saying that the whimsical games will come after realism. I am saying that we have had artistic and whimsical games throughout gaming up until this coming point.
I think towards the end of this generation and the next one we have seen and will see less and less of these games form Sony and Microsoft. (I dunno about Nintendo, but if tradition holds they can keep the artistic and whimsical alive.)
I think that the whimsical and artistic will be abandonded for pure realism on most games. Granted we may have a gem here and there of a artsy game, but they will not be as prevalent as previous generations. Even the last few years of this generation only had a gem here and there. Katamari, Ico, Wind waker, umm....I cant even think of one for the Xbox.
I dont think their will be a re-surgance of those games after the "realism" bug gains its strong foothold. I think market and marketing forices will push the realism factor full force.
Now the only market I think the non-realism games still have a chance to be a big percentage of those made would be the handheld market or maybe the new Nintendo Revolution. I think the Zelda:TP videos and images have shown us we can have extreme beauty yet still have whimsical. The Rev design, Nintendo's history and the controller lead me to believe that will be the one beckon of light for those types of games.
I would really like to see the re-surgence of the "hand-drawn" artwork in next gen consoles and next gen games. Castlevania and Tales of Symphonia and Veutiful joe did this really well. That has always been something I loved and have missed from this last gen.
Voodoo M @ Dec 18th 2005 9:42PM
I think what's happening here is that we need to unpack some concepts a little, so we understand what different people are talking about. I think there are two different things being discussed here. The first is the idea of a "virtual world". This might not incorporate any actual ‘games' whatsoever. The freedom to walk around a museum, or perform at a concert would not traditionally be considered a game. However, such experiences are interesting to people, and I think we would want any virtual re-creation to offer as much richness and fidelity as possible.
The second concept is that of a game. It is worth noting that this in itself is a difficult concept to define, although some (notably various philosophers), have tried. Some "games" are simply things we'd like to experience as closely as possible, but are unlikely to, or unable to in real life. Paintballing, in real life, is an approximation of war, without the danger of death. For many people, they would want the approximation to be as close as possible to the real experience. However, some games we play, even within the 'realistic' virtual space, are likely to be more focused on 'traditional' gameplay mechanics. Often a ‘realistic’ or simulation approach dulls the experience in many cases. The best approach may be to capture the essence of the experience, without the unnecessary and distracting micro-detail.
As far as virtual worlds are concerned, it is my hope that a lot of the construction work will be shared, leaving creators to focus on building interesting in-world experiences, whether real or unreal. A good example of this, and one of my current favourite 'games' is Google Earth, which will hopefully develop into a rich virtual re-creation of the real world, but without the frustrating constraints of our real lives.
Sorry for the long post.
rudimentalist @ Dec 18th 2005 9:42PM
Sasso, your comment was very insightful. Those "post-reality" games are something I'd really like to live to see.
theKiko @ Dec 18th 2005 9:42PM
Gameplay>Graphics
Always.
Darren Tilley @ Dec 18th 2005 9:42PM
I wish people wouldn't keep saying "focus on gameplay, not graphics".. It's such a ridiculous argument, and if you knew anything about games production, you will realise that the art team has very little, if any, influence over gameplay choices.
Most games are made like a factory line, the assets are decided at the start and the Art team makes them. If the Art Leader decides that they want realism, that is what's made. This choice doesn't have any affect on gameplay, so why is it so wrong? Artists of any medium tend to experiment with Photorealism before taking it in another direction. Why are games any different?
If games were made by one person, focussing on graphics would be at the expense of gameplay, but games are made by teams, and the art team's choices won't damage or help the decisions made by the designers.
Jay @ Dec 18th 2005 9:42PM
I dont give a rats arse about 3D graphics. now, a quality 2D graphic game will just drop my jaw. Metal Slug series pour example.
2D is greater than 3D. 3D is an embarrassment of something that's trying to look real and failing, badly.
epobirs @ Dec 18th 2005 9:42PM
I want realistic graphics in situations where it fits the setting of the game. This can include rendering of objects that don't exist in such a way as to make them seem real. Such as a light saber that shows signs of long use and wear when you examine it closely.
Part of the problem of realistic people and animals is that the investment in developers' time goes beyond the visual. Part of what makes an ultrarealistic CGI character fail is the lack of the behaviors we expect from a living being. So many games have NPCs that just do nothing in particular until you actually talk to them, at which point they go from aimless zombie to spastic fit because the developer thought this would make them more believable. I recall one game [Sword of Camalot?] where a child sat on a barrel like a puppet with its strings cut until you spoke to him and he flailed his limbs about like he having a seizure. That was an extreme example but has always stood out to me as a perfect example of bad NPC design. As it is, game environments like cities tend to be sparsely populated because it's so much work to produce each inhabitant (even without concern for the hardware keeping them going) and can make the game a needle in a haystack exercise for the player trying to find the NPCs critical to advancing the game.
Photorealism can be used to good effect, especially in combination with other graphics. I just introduced my neice to Pikmin on the GameCube and she was fascinated not only by the game but also by the scenery that incorporated images of real landscapes. A more photorealistic Pikmin sequel on future hardware could be a lot of fun without sacrificing gameplay to graphics.
Clu @ Dec 18th 2005 9:42PM
What gets me is when there is no variety in the characters. They were real enough for me in RE4 when I first started the game, Leon, the cops, all unique, but then I met the first grenade. That was cool until I went into the village and met his six twin brothers who all happen to be wearing the same clothes that day.
And when 3 of them pull back their hands to throw an axe at the same time, they all look like mirror images of each other. Have they been practicing? They look better than synchronized swimmers. The realism is ruined by this, more so than by lo-res textures.
It is the variety and uniqueness, the little imperfections that bring realism, not the detailed textures. I want games where no 2 enemies are alike, is that asking to much?
Jim @ Dec 18th 2005 9:42PM
No thank you. Reality is boring. I want shiny monster fantasy games!
Even movies aren't photorealistic (filters used, special effects, etc. etc. etc.)
hydrogen_wv @ Dec 18th 2005 9:42PM
Argh... I like good graphics, but I like good gameplay better... more specifically good gameplay in a game that is LONG. I feel more justified if I spend my money on a game such as Tales of Symphonia or Baten Kaitos, where I can put at least 50 hours into it. I'm sick of all these games that peak out at about 10 hours. Would you go to an arcade that made you pay $5 an hour? Then why should we have to pay $50 for a 10 hour game? I'd rather have good gameplay in a nice long game with shoddy graphics than a photo-realistic game that is only 10 hours long. Super Mario 64.... The graphics were good for the time, but If I hadn't sold my gamecube i'd probably still play that game.. it's long and it's a load of fun.
revo4life @ Dec 18th 2005 9:42PM
no matter how good the graphics are the characters will still move around like idiots which actually kills the realsim. you people out thier may want to see realism but will developers dish out time and a LOt of money to make the games more graphical? May be with first reales on XBOX360 developers have hit the jackpot but what will happen next year? when everyones played an xbox360 and there are will the games still be in high demand then? probably GTA and some other games but for everyone else down you go. games are actually losing great interest in gamers(especially teenagers) My friend refused an xbox360 for christmas he jus said he didnt want it.( i wasnt surprised). This is where i think nintendo comes in. Im not a fanboy but i think nintendo are pulling off what sony did with the playstation. nintendo are lowering the costs of developing and all that to get the attention of developers. they are letting developers know that the market is going to crash if they dont make a change
The 13th Duke of Wybourne @ Dec 18th 2005 9:42PM
Almost without exception every project I have been involved with has been badly managed; be it poor scheduling, lack of clear objectives, poor design, changing things half way through... the list is endless.
I just learnt how much the project we're just finishing cost to make. I nearly pissed myself laughing because it's utter shite. If we can get rid of the fuckwits that seem to head up most projects I think we could meet the expectations of next-gen without a significant increase in cost.
R .U .N .V .US @ Dec 18th 2005 9:42PM
If graphics start to look too real games would loose something. when it gets to the point that grand theft auto looks like a live action film ,I doubt It will be as fun to beat up an old lady and take her money. Or at least I hope it isnt as much fun.
Goku21 @ Dec 18th 2005 9:42PM
Man everyone knows that a model like that above needs about one week to be done by highly skilled artist who gets his 4000 to 5000 $
a month.
If you think... CGI-cinema takes about 200million $ for a 1 and a half hour film.
Im laughing out loud to everyone who thinks you get a game in the same graphical experience that last 25 hours to get through and only costs 30 million. HA HA you are so motherfuckin dumb ass retardet biatches....
thomas_h @ Dec 18th 2005 9:42PM
#41, what game are you making?