Photorealism Is Crucial To Games

Global illumination.
Volumetric clouds.
Sub-surface scattering.

These are words that make me hot.

But I know this feeling is forbidden. I should care about games, not the empty pursuit of photorealism. But oh my, it’s so exciting, and not empty. In fact, I think that right now photorealism is becoming crucial to games, and that we should celebrate it.

I love Unreal Engine 4. I like watching the CryEngine toad. I love Koola. I installed Unity on the strength of seeing the trailer with the cool seventies sci-fi porn man (I couldn’t program. Still can’t.)

I love realtime shadows, and I can’t bear it when something in view doesn’t throw them. I love examining wood grain in The Vanishing of Ethan Carter. I love the translucency of tropical fronds in Far Cry 3. I love the contrasts in sheen as the camera pans across a puddle on wet tarmac in GTA V. (I used to harbour the same feelings for the Gouraud shading in TIE Fighter.)

I love the thought that some nerd programmers have gone out into the real world just to look around and break it down into a series of visual components. I imagine them figuring out which are critical to making a realistic scene, given the power of their PC at the office, and then how to emulate it in code. I love that in doing so they felt the warmth of the sun on their pallid skins and the wind among the hairs on their forearms.

For me, photorealism is an inextricable part of my enjoyment of games. Sorry. I love Metal Gear Solid 4: Ground Zeroes: that lighting, Snake’s fluid animation, 60 frames a second. Photorealism in games isn’t just something static. It has to move. I love the slow-motion crunching and rending of steel in BeamNG. But it’s especially the reality I know from it being captured on film. For me, the muddy Polaroid-ish look of P.T. really pops: a potent blend of grimy horror flick and the intimacy of instant photographs.

I think I can see some good reasons why it’s not right to enjoy this stuff so much. Photorealism is an expensive distraction from what makes games valuable, right? It costs so much to produce but it doesn’t specifically lead to great progressive design. It’s dumb window dressing. Anti-art. It’s a quick wow; the lowest common denominator for the greatest price.

It also serves the ludo-industry complex, funding hundreds-strong farms of content producers at Electronic Arts and Ubisoft, and it forces you to buy new graphics cards. But there’s always another level of AA or a sharper texture resolution, just out of reach. The framerate can always be higher. Approaching reality only accentuates imperfection, and big business cashes in.

I don’t care. Being honest about it, photorealism is all I ever wanted from games. And I’m going to risk saying that it’s pretty likely that, at least at some point, it’s all you wanted from games, too. Certainly, for my 10-year-old son, visual fidelity is one of the first things he points out about a game. Anything that looks off is immediately noted, a black mark for which the game has to work hard to make up.

Now, I also love stylised games, because I love anything with considered and skilful art direction. And I’m arguing here that realism is also art directed. The judicious application of a little ambient occlusion here, depth of field there, and yes, even chromatic aberration. I think of those programmers, distilling reality with such care and attention, picking through their materials and applying them, just so. Their decisions and deep craft is worthy of deep appreciation.

Alien Isolation’s look is just marvellous, isn’t it, with those plump shadows and soft focus. The future through 1970s cinematography. I really like Batman: Arkham Knight’s, too, with its rivulets down a scratched metal breastplate. I feel I can reach into both games and feel them, and both speak purely to their themes and settings. Come on, Batman’s rippling stretched cape, streaming with rainwater!

They also give so much to look at, to enclose oneself within. A walk through them is to observe and soak in an atmosphere. Today, so many games are exactly about this. Not just Dear Esther and its progeny (incidentally, Everybody’s Gone to the Rapture shows The Chinese Room is right on my wavelength, because it’s all about making a postcard Shropshire village), but also all those survival and horror games, where reality is the ruse we’re meant to be buying into.

I know, yes, Minecraft and The Long Dark and White Night (but check out those shadows in White Night). Minecraft’s in a world of its own, and it’s probably more about creativity than survival. The Long Dark looks beautiful to me, it really does, but I would also love to explore a more visually rich world. It might even help me brave that damn wolf that keeps hanging around the door to my hideout. ARK: Survival Evolved definitely benefits from all that lushness (even if it tends to kill the performance on my otherwise game-friendly PC). That richness makes you want to explore, because it provides intrinsic reward in just going out into the world.

Horror games also tend to revolve around detail, from Alone in the Dark on. Going back to P.T., could that heavy atmosphere be drawn in such a small environment without all the lifelike detail? Not that it only applies to horror games: could Los Santos feel as special without the tufts of grass growing between concrete pavement slabs, different coloured headlights, and highway-side drainage ditches?

But photorealism dates so fast, and it’s so greedy for power, isn’t it? I certainly concede the costs, from the poor bastard artists having to draw grass and mud textures to the fact that even the best graphics card on the market will be superseded in months. Yet, many of the other great things today about games, such as the extraordinary breadth of visual styles to indies getting to play around with the latest and best 3D engines, are the result of the top end pushing further. This stuff really does trickle down. With the pursuit of photorealism everyone benefits. It’s our duty to want better.

124 Comments

  1. frightlever says:

    So to sum up, you love photorealism in games and also love games which aren’t photo-realistic. Thanks for sharing!

    • Dewal says:

      *Upvote*

      I mean, what’s the point ?
      Creating a false debate by an alarming title and a false question, answered by a mild opinion.

      Nobody “hates” photorealism, nobody is against it. Everybody loves a good view with beautiful light, sharp shadows and rich detailed vegetation. Look how everyone was disappointed when Watch_Dog and Witcher 3 weren’t as beautiful as advertised !

      The popular opinion you tried to go against is just that photorealism shouldn’t be pursued in spite of everything else. That photorealistic graphism doesn’t automatically make a good game (someone should call Crytek about that, by the way).

      Moreover, with photorealism, you either aim to the max or you fail. Chosing this kind of art-style mean first, that you accept that it will get old real fast and then that, if you fail at attaining the standards of your era, it will look ugly.

      So yeah, photorealism is a good thing an can sometime be truly amazing. But it is far from crucial and should only be pursued from the people that can achieve it without sacrificing the other parts of their game.

      • Synesthesia says:

        thank you.

      • MattMk1 says:

        I wasn’t at all disappointed that Witcher 3 didn’t look like the early builds they showed.

        And the only problem I had with Watch_Dogs was that it was boring.

        • Dewal says:

          I didn’t even played it in ultra but found it beautiful too, so I agree with you.
          But it seems to me that the louder crowd was disappointed.

          But either for those that loved its graphics that for those that expected it to be better, the arguments only serve to shows that the mainstream opinion (which is what the author tries to go against) actually cares about pretty and realistic graphics.

      • Paroxysm says:

        I love photorealistic techniques applied to stylize, surreal, etc styles. Like the look of a lot of nintendo’s current games. The cartoonish figures look so solid but it’s due to a lot of techniques developed in the pursuit of realism.

        But yes this article is silly :P I like games that look like a variety of things.

      • snowgim says:

        Also to note is that gameplay and graphics are usually handle by two completely different teams (sometimes even different companies), so a beautiful game with terrible gameplay just means that the gameplay team failed to deliver, it’s rarely the art teams fault.

    • alms says:

      Think I’m joining the perplexed party there.

      There’s probably an interesting debate to be had about photorealism but this post struggles to find a focus IMO.

      • rpsdemn says:

        Not sure why people don’t know this but there is a huge discussion happening in gaming dev circles saying that photorealism is dead, pointless, silly, wasteful of resources etc and that players should embrace cartoonish, stylized graphics as that is the direction they insist on heading. The truth is probably that it costs money and time and cuts into the profit margin but the reality is that gamers not only expect it, it is required for the deep foray into simulations smart companies are about to make. Between home VR about to happen and the open world / sandbox / survival genre exploding, the smart companies will go deeper into realism than ever before -not less. Anyway, this debate is why this article is here.

        • pepperfez says:

          Non-photorealistic art styles are a cynical corporate cash grab?

    • qrter says:

      It’s painful, isn’t it? It’s a tossed off internet comment stretched to short essay length. It’s embarrassing, really.

    • Eclipse says:

      thanks for completely missing the author’s point!

    • Rumpelstilskin says:

      I agree that making it specifically about photorealism detracts from a potentially more interesting question whether games should have good-looking graphics in general. And my opinion is that yes, appealing looks is important. It doesn’t have to be phorealistic, but if it’s cheap and simplistic, it better be very cleverly stylized to get away with that. I can still play something like Technobabylon, but I would have appreciated if it had looked like Stasis or Broken Age.

  2. Capt. Bumchum McMerryweather says:

    I personally think that what you’re saying is horseshit. Photorealism is necessary for realistic games, yes. But Okami, Shovel Knight, Abe New n Tasty, Mirror’s Edge, and pretty much every game that isn’t a shooter or a walking simulator would have suffered greatly from being photorealistic.

    But then, everyone is entitled to their opinion of course!

    • Capt. Bumchum McMerryweather says:

      I’d like to add that detail and realistic are not mutually exclusive. The detail, love and effort that goes into something like Bioshock, and say World of Warcraft is amazing and lovely, and you don’t need something that looks like a bloody movie to achieve this.

    • dsch says:

      All points already addressed in the article. Knee, meet jerk.

      • Eight Rooks says:

        Yep, they’re addressed by him explicitly saying “No-one really cares what Minecraft looks like, right?” and “Every game could use more detail, more, more, more, doesn’t matter what the art direction is”. So no, not kneejerk at all.

        That being said, Eidolon is almost one of the greatest games I’ve ever played, only it’s held back in large part by the fact its world is so god damned boring, what with it being almost entirely devoid of detail. Okay, that’s a walking sim – but Okami was actually supposed to be rendered in a realistic style to begin with, and its art design was as much down to technological constraints as it was “Well, not everything has to look like a photo”. I don’t like the tone of the article, basically, but I can’t wholly disagree with what it’s saying.

        • Dewal says:

          What you said about Eidolon being boring and without detail has nothing to do about being realistic. Bioshock Infinite & Dishonnored are beautiful and living games, full of details, and aren’t photorealistic.

          So I think that if you wish so, you have the right to “wholly disagree with what it’s saying”. :p

          • Eight Rooks says:

            But as far as I can tell he doesn’t mean photorealistic as in “exactly like a photo”, a one-to-one mirror of reality, and he qualifies that by saying that even ultra-“realistic” graphics are still stylised in some way – choice of (simulated) camera lens, lighting, colour tinting, particle effects and so on. I read it as: photorealism is great because it’s more about verisimilitude than actually replicating a photo, and therefore every kind of game in every kind of art style can benefit from the same principles that guide, say, Assetto Corsa or Gran Turismo. More detail on everything, more AA, more SSAO, all the rest of it. I don’t like the way he phrased it, I don’t like the dismissive way he mentions Minecraft, and I don’t like the implication in the article and in his other comments that every videogame is in some sense ephemeral because it can’t ever be “perfectly” rendered. But I can’t say I think his argument is entirely without merit.

    • MrUnimport says:

      Mirror’s Edge wasn’t photorealistic? I rather remember part of the thrill being how close it came at the time, thanks to those intoxicatingly beautiful indirectly-lit scenes, with colour bouncing all around the place and scattering onto white concrete.

  3. BlazeHedgehog says:

    I feel like the path to photo realism is pretty futile. It’s not just greedy, as you put it, but it’s like… kind of THE MOST greedy.

    And we’re already at a point where half of all releases these days are from indie development teams of ten people or less. Photo realism is not an option for those smaller teams.

    And for the big, giant, triple-A (or even quadruple-A) developers? Diminishing returns are setting in very, very quickly. Photo realism costs too much for a result that could actually tank their entire studio if their game doesn’t hit its sales targets.

    I see absolutely no reason to celebrate what is one of the leading reasons that game development studios are shutting their doors every single day. The latest Call of Duty could be made for half of its budget on the Xbox 360 with no visible difference in the mechanical depth of shooting your way through a scripted sequence.

    Technology has outrun all but the very few developers who are foolish enough to burn enough money to stay in the race.

    • Elliot Lannigan says:

      @BladeHedgehog but with the recent new deals on the major engines (Unreal, Cryengine, Unity etc), tiny teams CAN use photorealistic rendering. Everybody’s Gone To The Rapture was made by an almost unbelievably small team, and it has graphics that can easily be considered photorealistic, especially considering it was mentioned in this very love letter to photorealism that we’re commenting on.

      • Devan says:

        @Elliot
        I’d just like to chime in that even if the tools are inexpensive, there is an enormous time (and therefore money) cost in actually designing, creating, testing and debugging graphically realistic content.

      • BlazeHedgehog says:

        Perhaps I cast too wide of a net; I would not really classify The Chinese Room as a true indie developer. They technically fit the classification, but they’re more established.

        I mean, they did an Amnesia game, and I would not classify that as the more specific thought of what indie developers are, even if they do fit the bill of being strictly “independent.”

        I suppose my definition of indie would be “five people or less.” I’m moving the goal posts, sure, but when you have a team of 13 people (as they did on Rapture), suddenly you have multiple artists, multiple coders, etc. I’m willing to bet Rapture cost them more than a million dollars to develop and publish, which, while significantly smaller than your average Call of Duty, is still leaps and bounds above what I’d say the average indie developer is would have to spend.

        I was talking more about like… Dust: An Elysian Tail was made by one guy. I think BroForce was made by two people. Freedom Planet was one person. Slender (the original version that set the internet on fire) was one person. Five Nights at Freddy’s is one person. For the first four years of development, Minecraft was one person.

        It is abjectly ridiculous the amount of money being spent on game development and hardware when so many games coming out don’t even tap in to a tenth of that. A whole industry built on bloated excess, from top to bottom, left to right.

    • PopeRatzo says:

      Further, photorealism means games that require multi-thousand dollar PCs to play. Every game from a studio that lists the requirements as a GTX 970 and i7 processor makes me want to buy all future releases from that studio a little less.

      Game optimization is more important than photorealism. Spend your money where it will matter to gamers.

      I hate elitist stuff like this. “Oh, all games need more detail.” Sure, why not? How about “all games need more fun”?

  4. WiggumEsquilax says:

    Please, more photorealism in AAA games. The more time big publishers spend chasing the graphics dragon, the more great content from indies shines in comparison.

  5. satan says:

    I was going to point out that character hair in GTAV looks ten years out of date, it only stands out because most of the rest of the game looks so good. Then I remembered that GTAV was first released on consoles that are now 10 years old.

    Damn consoles.

  6. Elliot Lannigan says:

    I’m with you 100%, and I see no need to be apologetic about our love. That seems about as absurd to me as lambasting the producers of a game with famously great sound design (Bioshock and Halo come to mind off the cuff) for not devoting those resources to improving the story, level design etc. Of course it’s possible to have great games with simplistic, non-photorealistic graphics, but that does not in anyway necessitate that there’s something wrong with pursuing and loving photorealism; on the contrary this relentless, passionate pursuit is something beautiful.

    • Devan says:

      What I don’t like about this article is that it seems to seek out sympathetic responses like yours by creating the impression that an appreciation for realism is widely repressed as some kind of taboo.

      At the same time the headline baits in readers with the title “Photorealism is crucial for games”, but the article doesn’t say anything about why or in what way it is crucial to games.
      I came here because that title is an interesting and unexpected claim and I genuinely want to see how it could be substantiated because I can’t think of any way in which it is, truly, crucial.

      Instead we get a cutesy love note to pretty graphics. I don’t see anyone arguing that photorealism is a bad thing; it is simply more or less important to different people. I DO see someone claiming that photorealism is crucial to games, even though it is not intrinsic to them and it has been solidly demonstrated that games can meet their goals (of fun, learning, storytelling, etc) without realistic graphics. If someone can make a valid argument about that then I’m interested in seeing it, but this article doesn’t deliver. The title should be changed.

      • DanMan says:

        But it is kind of a taboo, isn’t it? Whenever someone complains about graphics not meeting their standards/expectations, someone comes along and goes:

        gameplay > graphics. STFU nub.

        Or something along those lines. Like you’re not allowed to express your preference for nice graphics.

        • JFS says:

          I don’t think that’s true. It’s okay to like nice graphics. That’s not a taboo in any way. Being a single-minded sucker for GRAFX might be, but that’s different. Also, the OP has a point. This article is strangely titled.

        • Elliot Lannigan says:

          Precisely.

        • Turkey says:

          People usually attack the complainers cause they start the debate by shitting on the game with such amazing conversation starters as “Looks like a ps2 game.”

      • alms says:

        I think his point about it being crucial is that “it trickles down”, i.e. you can have interesting art styles other than photorealism because there has been a race to photorealism.

        It’s not really a compelling argument as far as I’m concerned though: let’s imagine for one absurd minute that photorealism was never a thing, would have the race for faster, more powerful hardware not taken place?

        Honestly, I don’t think so.

    • LexW1 says:

      That’s a really odd comparison.

      The issue with photorealistic graphics is that they spend an utterly disproportionate part of the game’s budget.

      I am not aware of any games where sound-design spent a disproportionate amount of the game’s budget.

      Don’t get me wrong, I love photorealistic graphics (when they work, which is about 35% of the time, frankly – rather lower than the 60% or so of the time I’d say more stylized looks “work”), but it seems like a completely false analogy you’re making. Either that or you just don’t understand the problem.

      (A further issue is that photorealistic graphics tend to require increasingly expensive hardware – I know a lot of the writers here have £1200 quid+ computers, sometimes a lot more expensive than that, which absolutely makes sense for a game reviewer, but my main objection to the pursuit of these graphics, myself, is that they lead to games demanding more and more expensive hardware – we actually caught a break from that for a while, when non-photorealistic graphics were forced “into style” by the declining power of consoles compared to PCs, but it seems like the whole race has heated up again – maybe it’ll cool in a year or three as the consoles hit their maximum capability, but not before I’ve had to not buy games because I don’t have the £170+ quid graphics card and £700+ PC you need to run those games at 30fps on settings which don’t look like grey mud.

      • Frank says:

        Yup, the problem with fancy graphics is that it puts up barriers to entry for developers and players alike. Fancy graphics are great as long as they don’t crowd out better things.

  7. Premium User Badge

    Andy_Panthro says:

    Better graphics at this point are surely a case of diminishing returns aren’t they? We’re at the point where we’re getting realistic hair physics and such, which is nice but very superficial. It’s a lot of money and processing power into something which has such a minimal improvement on a game.

    You mention ARK, which looks great, but it has very static terrain, some can be destroyed but it lacks the transformable terrain of Minecraft (slightly different sorts of games of course).

    As for the original Alone in the Dark, it wasn’t really detail and photorealism they were going for was it? The 3D models in that were surely a stylistic choice.

    If we moved no closer to photorealism than we are now I would be happy, but I have to admit I do enjoy seeing those improvements. We’re so far away from when I was a kid it’s incredible.

    • GWOP says:

      I believe that there is a general lack of destructible terrain in games is because it messes with the navigation mesh for the AI, not because it’s graphically intensive.

      • Premium User Badge

        Andy_Panthro says:

        Can the navigation mesh not be redrawn? Surely that’s not an insurmountable problem? (I don’t know anything about this though, so perhaps it might be)

        • GWOP says:

          Oh yeah, definitely. Dynamic navmesh generated at runtime will do that, but I think it gets a bit processor intensive as the size of the area decreases.

    • LexW1 says:

      I think there’s a whole layer of photorealism that’s still mostly untouched (and not hair physics), particularly to do with lighting. Lighting in pretty much all games is still fake, and looks fake, really, really, really fake in a lot of cases (heavily mod’d Skyrim can look more real than a lot of modern games oddly enough). So there’s that challenge at the very least.

      Then there’s animation. Which is still fake-as-fuck. It could hardly be any fake-r. Animation is in a similar place now to where 3D graphics were in the ’90s, really. So that’s a whole other layer of realism that they could be doing. Physics in general, too.

      Hair-physics is a funny diversion.

      • Premium User Badge

        Andy_Panthro says:

        Lighting and shadows have been getting better surely? perhaps not as quickly as increasing texture detail or polygon counts, but certainly better.

        Animation on the other hand is a good candidate for more investment. No point in realistic models and environments if the characters look like puppets.

        • SuicideKing says:

          Lighting will be solved when you can do real time ray tracing. If you run LuxMark, you’ll see how far we are from that.

        • LexW1 says:

          Getting better? Oh, sure.

          Anywhere near where they could be? Nope. Still very far. As SucideKing says, Luxmark shows that, and even that’s not really the endgame, lighting-wise. We’re probably a couple of decades out from lighting that’s basically indistinguishable from reality (and lighting issues get more obvious as resolutions go up, note).

          It is incredible how far we’ve come, of course. When I were a lad, you were lucky if your polygons were filled in! (Though that does raise an interesting point – sometimes less is more – some vector-graphics-style games looked better than the filled-poly ones of later years – and we see this again – some games trying really hard for photorealism look distinctly inferior to ones sacrificing some of that realism for style).

      • DanMan says:

        Couldn’t agree more. Lighting and animation are the two most critical areas for me as well when it comes to graphics. Doing both of them well usually results in a very lifelike experience.

        Like the author, I get annoyed when there’s a dynamic light source, but it doesn’t cast any shadows on objects nearby. We also don’t have AO everywhere yet, because of performance reasons. It adds so much though.

        Animation and collision detection go hand in hand for me. It looks so stupid when characters just go blindly through their motions without any kind of sense/response towards their surroundings. The result? Fists and weapons that clip through other characters and objects as if they’re not even there. Doesn’t look vivid at all. More like animatronics from the local fairground’s Tunnel of Horrors. Just worse.

  8. ResonanceCascade says:

    I would gladly give up some texture detail and polygons to have shadows that don’t look pixelated or show weird artifacts in motion. It’s always particularly egregious on facial shadows. Any graphics experts out there — armchair or otherwise — know why that is? It seems like even high-end AAA games struggle with this.

    idtech 4 didn’t have the issue way back in ’04, but it used weird harsh shadows, so I’m assuming it has something do with softer shadowing techniques.

    • Geebs says:

      I’m no expert, but I’ll try to help a bit:

      Soft shadows are difficult because of three things:

      1) you need to do an additional pass to render everything that casts a shadow and, for some techniques, everything that needs to receive a shadow. This can be very tricky – something well outside of the main camera view can be casting shadows into the visible area, so you might need to draw a lot of stuff. This can be expensive.

      2) most shadows in games are created by drawing the scene, and recording the distance from the light, from the viewpoint of the light. This allows you to infer that anything further away from the light source than what’s in this picture must be in shadow. This information is stored in a texture. The texture needs to be as small as possible for speed, and it also needs to be as big as possible to keep the shadow detailed. The trade-off usually looks bad under some conditions and good under other conditions.

      3) and this is the biggie: depth precision is a problem. Prior to the most recent generations, GPUs used very imprecise numbers because they’re much cheaper in terms of resources. Lighting a face from a long way away may mean that you can’t cast a shadow from, say, cheek to nose or chin to neck because the GPU doesn’t have suffcient numerical precision to tell that one is further from the light than the other – they’re all within one in-game metre of each other and that’s the smallest depth difference you’re set up to resolve. Limit the scope sufficiently to do face shadows properly, and suddenly that tree over there doesn’t cast a shadow any more because it’s out of shot. Falling between these two extremes, there’s a bunch of situations where shadows look horrible because the angle at which they’re cast across a surface introduces local problems in depth precision, causing horrible flickering bands.

      TL:DR – shadows are hard in terms of resource management and always end up as a least-worst compromise. Magical new shadowing techniques to solve all of these problems get invented regularly and only ever look good in totally artificial scenes.

      • ResonanceCascade says:

        Thanks for the explanation! I’ll just assume this won’t be fixed until 2049, when realtime path/ray tracing is actually viable.

        • Xerophyte says:

          As a person who works on a real time ray and path-tracer for a living, I rather hope it won’t take that long.

          • ResonanceCascade says:

            I hope so too! The Brigade demos look amazing.

          • Stellar Duck says:

            Wouldn’t that just mean you have steady work for 30 years?

      • Premium User Badge

        teije says:

        Very interesting – thanks!

      • Xerophyte says:

        Depth precision hasn’t really been an issue for a number of years. It’s limited, to be sure, but with a 24-bit depth buffer you have 16 million depth slices. That’s a lot of precision, and that’s not counting cascades. The resulting errors are very slight.

        The problem is that while we have 16 million depth slices we can’t quickly render an image with have 16 million x 16 million pixels of resolution in the other two dimensions…

        • Geebs says:

          I disagree. Z-fighting is still a huge problem for any open-world scale of draw distance, and a basic shadow mapping setup is pretty much guaranteed to have issues with depth aliasing unless you kludge in a bias term, even with 32-bit depth textures.

          Plus, if you want those 16 million slices evenly distributed with respect to distance from the camera, you’re going to have to throw out the hardware depth buffer.

          • Geebs says:

            (Of course, you’re completely right about the limitations in shadow map resolution. Also, I’m out of my depth ;-) )

          • Xerophyte says:

            Yeah, you’re definitely right that depth precision is a problem in a lot of scenarios. I didn’t intend to say that it’s a solved problem in general.

            I meant to say that I don’t think it’s a problem for shadow maps, specifically. Shadow acne still happens, but far as I’m aware it’s generally more because of coverage estimation inaccuracies introduced by the assorted filtering techniques than precision inaccuracies in the buffer. Even a small PCF kernel should blur the data enough to obliterate any FP precision issues.

    • Xerophyte says:

      Shadows in real time graphics are a pretty complicated question — my old thesis advisor literally wrote a book on it a few years back, Real-Time Shadows — but I’ll try to give an overview.

      The short answer is: game graphics are done in a way that makes checking whether or not there’s anything between a light source and a point visible to a player impossible, and all the ways people have invented to get around that uncomfortable fact are terrible hacks.

      The longer answer means I have to get technical, sorry.

      When rendering an image in a game the basic operation by which stuff goes from 3D scene to 2D image is that the game takes a triangle in 3D space, figures out where that 3D triangle maps to on the 2D screen (the vertex shader), figures out which pixels are in the 2D triangle and finally computes a color for each of those pixels (the pixel or fragment shader). This method of drawing 3D is called rasterization and GPUs are built primarily to this very, very quickly for a whole lot of triangles and pixels. They do this by treating the computations for each triangle and pixel as independent of all the nearby triangles and pixels and then doing several thousand of them in parallel.

      The problem with rasterization is that the speed is achieved by throwing away any contextual information as soon as possible. When shading (=computing the color for) a pixel we know where it is, we know what the object surface is like at that particular point and we can send in some relevant light sources that affect it. We know fuck-all about the rest of the world. Shadows and reflections and such are now extremely hard since they are entirely dependent on what the world looks like from the perspective of our pixel, which we just threw away. To make it work we need to use some crude approximation that is simple enough that we can use it in our thousands of per-pixel calculations but still says enough about the world nearby to get some sort of shadow. The shadows we get will be “bad” in some sense but can run in more than .2 FPS. There are, traditionally, 2 main methods for doing so.

      The first is called “shadow volumes” and it’s what Quake and the other older id games used. Here we figure out what the volume of space that is in shadow looks like, draw that volume to a separate image (technically: to a stencil buffer), and finally use that image to figure out if a pixel in the real rendering is in the volume or not. One problem with this approach is that without some very complicated tweaks you only get hard-edged shadows and those look very unrealistic. More severe is that we need to figure out what the volume of shadow cast by our stuff looks like so we can draw it, and doing so gets really complicated the more triangles we have. With today’s model detail it’s basically not a workable idea.

      The second approach is called “shadow maps” and works by first drawing an image as seen from the shadow-casting light source. When rendering a pixel in the real image we can use the image of the scene as seen from the light source to figure out if there was anything else closer to the light than our pixel and therefore blocking said light. The main problem with shadow maps is that the accuracy of our shadow is limited by the resolution of the shadow map: if it’s too low-res we’ll see the pixels of the shadow map in the shadows, and the more we zoom in on something the blockier the shadow gets. One advantage of shadow maps is that we can get more realistic soft shadows quite easily; for a given computational budget they’ll almost always look better than shadow volumes. Most important is that getting a shadow map just involves taking the 3D triangles we had before and drawing them, which is exactly the thing that the GPU is really, really good at. The shadows in pretty much every modern game are based on this, with some tweaks in order to combat the resolution artifacts. It’s just very, very hard to find a shadow mapping solution that works well everywhere.

      There’s also a class of approaches to shadows — baking, ray and cone tracing, etc — that don’t have any of these problems but they only work if the light and the geometry are static. They are why we’ll often have nice shadows on the environment and crap shadows on the characters in games that don’t have day-night cycles (or have them but cheat with a static sun). They’re also why the Unreal Engine architectural demos look a whole damned lot better than any Unreal game: shadows are easy if they never change.

      The “real” solution to shadows is to not throw away the 3D context when rendering our pixels and use the actual 3D data to check for occluders. It’s possible to do this for static geometry nowadays by precomputing a bunch of additional information, but it’s just too computationally expensive to be done for animated geometry right now.

  9. keithburgun says:

    Firstly, I want to say that I am very, very happy someone had the audacity to say this out loud. Bravo on that. I think it’s an idea that infects the minds of many, many people, even many who think that they’re totally immune to it. So I think shining more light on this question is a purely positive thing.

    With that said, the idea suggested by the title of this article is, of course, dead wrong. The article itself proves that by mentioning stuff like Minecraft. If you think Minecraft is good, then how is photorealism “crucial”? You could have said important, but you said crucial.

    Crucial: decisive or critical, especially in the success or failure of something.

    Basically, if something is crucial, it’s make or break. And then you go on to say that it’s not make or break. I guess maybe this could just be a case of wanting an attention-getting headline?

    Anyway, I think the implicit message here is that interactive systems can’t actually have value by merit of their rules alone, a claim which millions of boardgamers know to be false.

    Further making this claim absurd is that “photorealism” is a moving target. So let me get this straight – if a 2015 AAA game is fun, because of its photorealism, does that mean that in 2020, that game will *no longer be fun* because there are more-photorealistic-looking games out there?

    Again, I’m really happy someone is saying this stuff out loud, because these are ideas that I think are very common, under the surface, but which cannot stand the light of day.

    • Alex Wiltshire says:

      Hello Keith! I also love Auro! Thanks for the comment.

      So I hoped I was clearer that I believe the pursuit of photorealism is crucial to games as a whole, not on an individual basis. So I don’t think it makes any given game ‘good’, or that games that aren’t photorealistic are not good. That would be crazy! I also tried to say that a good argument against the whole photorealism agenda is that it doesn’t support progressive design in any direct sense, by which I meant the systems that you’re so good at designing. That stuff is also crucial! I really hope I don’t imply that the pursuit of photorealism means that systems aren’t important. Just that photorealism is a cool thing that we can appreciate about games.

      How much one appreciates systems or visuals is up to taste, in my opinion. I very much appreciate deeply systemic games, and for a long time felt they were the One True Way, the thing that makes videogames so different and amazing compared to any other creative medium. I guess I’ve softened now, and I can also completely understand someone else who mostly plays in order to be able to see amazing sights. I once saw that as shallow, but now see there’s a complexity of thought behind photorealism, which is so often written off as empty.

      And finally, yes, of course photorealism is a moving target. That was what I was trying to broach by noting that you’ll always want more, things always go out of date. But that doesn’t mean that seeing something newly photorealistic in a game, whether in a UE4 game that’s coming out tomorrow, or way back when you saw the first Unreal Tournament, it doesn’t make it worthless, just because it will date. It’s part of the continuum, it’ll lead on to further advances, and it’s OK for it to excite us, even as we also know that it’s a fleeting thing. Actually, I think that fragility is kind of beautiful.

      • fucrate says:

        Change “Photorealism” to “Skinny girls with huge tits” and my reaction to your article is the same. Your main interest in games begins and ends with surface level wow factor and it is lowest common denominator.

        Thank you for being the only journalist around here willing to say “I am a boring person and my tastes haven’t evolved much beyond that of a ten year old boy.” It’s a statement made far too rarely here and I think it’s important for us hipsters to remember that most gamers are just as vapid.

        • ResonanceCascade says:

          The path to indie game success: angrily trolling RPS writers from an account that links to your game company. Thank you for the hearty laugh. :)

          • internisus says:

            Sarcasm is not trolling.

            Do you really believe that game developers should walk on eggshells around journalists, afraid to speak their minds as readers in case an article’s author decides to add them to a blacklist of games and developers to be either shunned or deliberately besmirched in the future, regardless of the actual quality of their work? Do you really want to suggest that the writers of RPS have such base and juvenile ethical standards as that? Do you really see the world as being full of people too scared to talk to each other lest they offend and invite petty retaliation?

          • internisus says:

            On second read, it might not even be sarcasm. This might be a developer who sincerely appreciates the disheartening wake-up call.

          • ResonanceCascade says:

            Of course I’m neither suggesting that RPS has the power to “blacklist” anybody, nor that they should or would. I’m saying that it’s pretty stupid to go out of your way to randomly be a massive dick to someone who also works in your small industry, especially from an account that literally links to your business. It’s like wearing a sign that says “I’M BOTH MEAN AND AN IDIOT,” which is why I got a laugh out of it.

          • fucrate says:

            Eh, I’ve got more twitter followers than Alex, I think that makes it okay to be a sarcastic asshole to him.

          • ResonanceCascade says:

            @fucrate

            That’s a good point.

        • TillEulenspiegel says:

          Change “Photorealism” to “Skinny girls with huge tits” and my reaction to your article is the same.

          Yeah, pretty much this. Or to be slightly more generous, an equivalent article about special effects in movies.

          I’ve always been baffled by people who gush about “special effects”. Sure, sometimes they’re pretty, but…I just don’t care. That’s not why I watch movies. If you try to do fancy effects and they’re awful, yes that’s bad. But just give me a well written, well acted film with tight editing, and spare me the distractions.

      • Consumatopia says:

        I miss hand drawn animated movies. OP seems to assume there is some kind of dichotomy between “visuals” and “systems”. So it’s worth thinking about animated movies, e.g. pure visuals, no interactivity. Even in that domain, we have too much photorealism. Yes, sometimes photorealistic techniques can be used for stylized animation, e.g. Tangled, and I’m not saying that looks bad, but it doesn’t look better than the old hand-drawn cartoons looked.

        Or consider something like Guilty Gear Xrd, which looks so good precisely because they worked so hard to, in their own words, “kill everything 3D”, e.g. fixed camera angle, keyframes only.

    • Frank says:

      Um the title can still be fine. Read it not as “all games need photorealism”, but rather as “we need games that strive for photorealism”.

  10. Juan Carlo says:

    I think the path to photorealism is inevitable, but once achieved I’m not sure if many games will be made with photorealistic graphics, partly just because non-stylized, photorealistic violence would be infinitely more disturbing (although I’m sure some people stomach it), but also because photorealism is kind of boring. Apart from games or virtual worlds that are meant to mimic the real world for whatever reason, once we reach photorealism I suspect games will become way more stylized. And if VR becomes common, I suspect people will demand that virtual worlds have some degree of stylization just because it’d be incredibly disturbing to put on a helmet and be in a virtual world that looks and acts exactly like the real world (if only because it’d start to blur the distinction between the virtual and the real).

    I suspect photorealism will affect games similarly to the way it affected art. Photorealism was a major goal for many centuries, but once cameras came along art started getting way more stylized. Not sure if we’d have something like cubism, for example, had the camera not been invented. People will get tired of the novelty of photorealism in games fairly quickly, I think, and will inevitably start to want things that they can’t find in the real world.

    • LexW1 says:

      I think it’s very likely you’ll be right, for the most.

      Games play a somewhat different role to “on the wall”-type art, of course, so I think we’ll see photorealistic games remain common for a long time. As you say, they will be slightly disturbing with VR or particularly AR (which I think will get big sooner than VR, myself, as it’s not this solitary enthusiast experience that VR is), but games will use that.

      Horror games, for example, may well want to go straight for the ultra-real jugular, because you’re going to be more alarmed if something looks like it’s real than looks like it’s cool. Similarly “experiential”-type games, which include a lot of FPSes, actually, will want to go for realism, because the feeling of being there is pretty intense.

      Mud-n-blood-type fantasies like The Witcher would also want realism rather than stylization.

      But yes, you’re right that a lot of popular/mainstream-type games will probably stay away from full realism – I doubt they’ll go to the heights of abstract-ness that art has, of course.

  11. DrMcCoy says:

    And I’m going to risk saying that it’s pretty likely that, at least at some point, it’s all you wanted from games, too

    Sure, when I was younger and more ignorant than I am now. But I also scoffed at art that’s was not straight up realistic portraits then. Because I was an ignorant little shit.

    • Scurra says:

      This. Although I’m not even sure I went through the photorealism stage myself – but that maybe because I was brought up in a household that went to art galleries? So I appreciated impressionism and abstraction (and even surrealism) before I really started to think about what I was seeing on my gaming screen.
      (And also, perhaps, that Tetris on the Gameboy taught me that all you actually needed were grey-and-white falling blocks.)

      • caff says:

        How nice it is to see a couple of people echo my thoughts, consdering the parallel between the art world and the game world. Conversations such as these must have been taking place in the 19th century between traditional artists and photographers.

        I’d agree with the comment above that higher fidelity drives indies to better results, but ultimately I think the interactive mechanic of a game can play a more important part in the overall experience.

  12. TillEulenspiegel says:

    I’ve said it before: Quake III was and still is my dividing line for “good enough” 3D graphics. It’s fine. Until you can get absolutely indistinguishable realistic graphics for VR, I really truly do not care about marginal improvements.

    And we’re nowhere near genuine photorealism. Real-time videogames are still only inching towards movie CGI, which itself still looks terrible and unrealistic in many contexts. Pick your favorite massive budget Hollywood blockbuster of the past few years, and look at how silly any wannabe-realistic CGI character is.

  13. dsch says:

    An article that strays from the current “indie” orthodoxy. Predictably received by the crowd.

    • JFS says:

      You do know that this is RPS? A gaming site with a certain direction? “Anti-Hipster” much?

      • pepperfez says:

        I think at this point it’s trivially true that everyone is anti-hipster, given that “hipster” seems only to mean, “Person who cares about some things more than I care about them/”

      • dsch says:

        Eventually the worm will turn and photorealism will become a hipster value. This article is simply ahead of its time.

    • internisus says:

      And you, a representative of the straw man orthodoxy.

      • dsch says:

        If it’s a straw man, it’s a peculiarly alive and kicking one, judging by the comments.

        • internisus says:

          Or another possibility is that a lot of people with diverse perspectives and concerns happen to be voicing negative reactions to this article and you’re making an enormous assumption by lumping them all under a ridiculous banner that you call “the indie agenda.”

        • qrter says:

          Although there certainly are people disagreeing on the importance of photorealistic graphics in games, I see at least as many people commenting on how thin the actual article is on content – it basically reads as “I like photorealism in games because I like it”, and not a lot more.

          • dsch says:

            @qrter. I read its thinness as a protest against the thinness of games journalism in general, except games journalism does not recognise its own weakness. Games journalism has not yet reached the level of games criticism because all its categories are still naively given. For example, what is the usual argument for indie games? The one positive value that is agreed upon is “gameplay” — as opposed to “graphics,” or “hype,” or “AAAness,” or whatever. But this concept of gameplay is simply imported wholesale from the naive distinction of design vs. programming vs. graphics, which eventually comes from the division of labour during the production of games. (Another category, “narrative,” is simply imported from the analogy with film.)

            Only recently have we begun to see this picture complicated, such as in the reception of Dark Souls, and games like Dear Esther (despite being not very good) have started to challenge the traditional focus on gameplay as the site of innovation/creativity.

            So what I think this article does is precisely to insist on the dumb ideal of photorealism as something that exceeds the present discourse of games journalism. There is something that photorealism appeals to that the indie consensus is unable to account for, and it is provocations like this that will hopefully stimulate games criticism.

  14. Hensler says:

    How the hell were you able to publish this on the modern Rock, Paper, Shotgun? Did it involve sneaking into the server room under a cardboard box?

  15. Freud says:

    Our brains are wonderful constructions that are constantly looking for patterns to make sense of. That’s why we can imagine a world of Dwarves seeing all these ASCII symbols.

    But it’s nice to see photorealism being pushed. Anyone who has played P.T knows how immersive and scary something that looks almost real can be. Even if it was just a looping corridor with a few rooms it had a sense of place.

  16. trn says:

    I don’t mind reading an opinion piece, but I do balk at being told what I think. The arrogance here does not suit you, sir.

    I almost exclusively play top-down and isometric games, I suffer terrible motion sickness playing 1st and 3rd person games. Photorealism? Please. Give me Dungeon of the Endless, Stasis, Torchlight 2 any day.

    As far as I am concerned, more devs need to pursue art styles that convey something vital about the nature of their game. Photorealism is not always going to accomplish this.

    Anyway, this is just like my opinion, man.

  17. CrazyMoai says:

    So, you love photorealistic graphics but do not mind stylised graphics if those are good.

    Me myself like pleaseant graphics, pleaseant being anything that is not uncomforting to watch, could not be more subjective. My standard however has improved and many old titles have lost appeal to me due to graphics, specially many early 3D games that pursued reality and have aged badly.

    Neither of us needed more than 2 paragraphs to explain his point of view.

  18. Philomelle says:

    I’m honestly not sure what you’re even talking. You discuss how crucial photorealism is, but then you post screenshots that aren’t photorealistic (the background on the header image is chock-full of stylized filters) and you list games that aim to not be realistic. The Vanishing of Ethan Carter used its realistic-looking textures to invoke a particular mood – the game is meant to feel like it’s stuck in a single moment, so its lighting and movement do not function in a realistic way. Arkham is chock-full of stylized fog. Far Cry 3 uses visual filters to invoke the feeling of being stuck in a drug haze.

    I get that you want games to look good and their worlds to be immersive, but the way you convey it only tells me you have no idea what photorealism means.

    • pepperfez says:

      Does “photorealism” actually mean anything useful? I mean, outside of the reaction against abstract expressionism, does it convey anything more useful than, e.g., “high-definition”?

  19. Chiron says:

    Graphics for a long time have been “good enough” for me, as long as I can tell what everything is and theres no pixelated edges I’m fine with it.

    Animations need more work a lot of the time mind, and AI needs a severe kick up the arse. As for plot… well, its been very clear for a while that what the target audience for most games seems to be is the Multiplayer crowd which is used to paper over the cracks in terms of AI and Plot.

  20. geldonyetich says:

    Personally, I prize good stylization over photorealistic graphics. I would rather play another Twinson Little Big Adventure or Wind Waker than a dozen Call Of Duties or Crisii.

  21. Laurentius says:

    Oh yeah, baby, talk to me more !

    I don’t know what people are in arms about tbh, I play Master of Magic to death every two months, just google if you wan to see how graphics of this game looks like but I am totally with Alex on this, I love these splendind graphics and sometimes I can look at it for hours and make screenshots. And yes I also loved Gouraud shading in Tie Fighter, no wait I still do!, flying around Star Destroyer was magical. And I keep games with shiny graphics on my HDD and sometimes just launching them only to experience this shiny graphic and not play them at all.

  22. MrNash says:

    Photorealism is fine and dandy, but there’s a time and a place for it. I’ve never really gravitated toward it myself, preferring much more varied visual approaches.

    Even as games do get more realistic visuals, what actually throws me off, though, is the animation. In that case, it can be far more jarring. I may look at a couple of characters and think they do look surprisingly life-like, but then one of them will move in a somewhat unnatural way and suddenly the illusion is broken.

  23. DanMan says:

    Brother! Where have you been! Can I get an “Amen!”?

    No? Not even an “ey, man!”?

    *sob*

    I’m over here with my Ambient Occlusion and Tesselation, if anyone needs me. Ahhh…. who am I kidding…

  24. Zack Wester says:

    The Thing whit photorealism in my eye is this.
    let say it take 10 sec to model a 15Polygon model.
    it takes 20 sec to model a 25Polygon model, it takes 30 sec to model a 40Polygon model and it takes 1 min to model a 60Polygon model.
    the thing is the more Polygon or in this case the more photorealism something is the longer it takes to make.
    so that Hl2 Room that took 1 day to make might in a photorealism game take a week. Game play wise your spending 20 second in it, so its a cost reward thing and many would question if buying a okey looking game for 30USD or a great looking game for 60USD the great looking game Might even have less rooms to explore.

    Also this question let say you play Battlefiend you see a 4 store apartment block in the lower graphic game you can probably visit every single room in the photorealism(High graphic one) probably not.

  25. SuicideKing says:

    Hmmm. I certainly didn’t care about photo-realism when I was 10, maybe because it didn’t really exist in the 90s. Low-poly dinosaurs in arcades were enough to scare me and make me bail from the seat.

    Yes, Star Citizen or Elite look really nice, I still want my FreeSpace 2. Of course, I want a prettier FreeSpace 2 because it’s been 16 years since the original, and I’d want today’s FreeSpace 3 to look like Elite or SC. Of course, even if it just looks like FS2 + the Source Code Project mods, I’d still play it, because of everything else the game offers.

    And the thing with photorealism is that it demands better of everything else – Arma 3 looks photo-realistic at times but then you see grass pop-in and terrain warp, weird shit going on with rocks, or you get stuck in doorways, and it’s distracting and doesn’t look nice. This used to bother me, but then I started playing with Folk ARPS and I didn’t care anymore.

    Conclusion: “Photo-realism” is okay, but not crucial – art, mechanics, AI, balance, storyline, world detail, etc. matter a lot more.

    Alec doesn’t like MGSV because of a thousand graphics. He likes it because a cute puppy licks his character’s face.

    • Archonsod says:

      I think that’s probably the biggest counter argument. Photo realism doesn’t really matter because in hindsight it’s always photo-realistic. I remember being blown away by Barbarian 2 on the C64 back in the eight bit days. Load it up in an emulator now and the notion of those being good graphics is somewhat laughable, however when I remember my time playing it I recall it having all the fidelity of the Conan movie. You don’t really need ultra-realistic detail, just enough detail to allow imagination to do the rest.

      • SuicideKing says:

        Yes! Exactly this! It’s always photorealistic for the time, because imagination does the rest.

        Heck, I had never played Thief, so I tried it last year – I had to take my headphones off for the zombies in the cave. Without audio, they looked like the least scary thing ever, but the audio drove imagination. And no, I don’t like zombies. :D

        I wonder if the recent race for photorealism kills the imagination part to an extent, which is perhaps why a lot of us who’ve been playing before the turn of the millennium go “meh” now.

        I never played Arma 2, but a lot of people in the community find it had more soul than A3.

        • Rumpelstilskin says:

          > Yes! Exactly this! It’s always photorealistic for the time, because imagination does the rest.
          I can’t quite agree with that. No one even thought of using the word “photorealistic” back at C64 times, because its inadequacy would have been obvious. And today there are games that are realistic enough to be called that, and games that clearly aren’t.

  26. Iajawl says:

    I can’t help but wonder if this article exists just to create controversy.

    • alms says:

      I’m giving them the benefit of doubt and just assuming the post didn’t come out so well.

  27. daphne says:

    Reads like the article was written by the author’s son. I don’t mind the fluff, but this article doesn’t actually say anything about how photorealism is crucial for games. Given the lack of empirical evidence — just look at the top played titles on Steam on any given moment and tell me how many owe their success to photorealism aspirations — you’d think the author would be intellectually responsible enough to make a point.

  28. shrieki says:

    virtual reality sure needs photorealism.games are no more what they used to be … walking simulators and some people even ask for “tourism simulators” and such :D

    but hey people will probably always love to play with abstractions – even if they can have total photorealism.

    but it is a very interesting debate.

  29. Jakkar says:

    It’s important to recall that the technical functions described aren’t ‘vital’ to the experience described.

    Look back to FEAR, for example, a game that achieved a very thick atmosphere through quality lighting, high detail textures and models for its era, and managed to look ‘real enough’ to immerse the player powerfully.

    Slip back further, turn of the century, and compare the colourful bounciness of the original Resident Evil to the muted greys, intense darkness, mist and ‘grit’ of Silent Hill, and know which game truly absorbed you into a compelling atmosphere.

    Further; Halflife’s absymal lighting engine, comic scientists with Einstein hair, aliens painted in primary colours, and compare it to Unreal’s dynamic lighting, inky darkness, its capacity for both the grimly muted and the glowing intensity of fire, acid greens and lambent blues and dusty browns as foes sparkled in the light of explosions or became invisible, dashing from cover to cover in the gloom.

    ‘Photorealism’ isn’t the term I’d use, for all that is truly necessary is lighting, as best-demonstrated some years ago in the revolutionary Float32 lighting mod series for the STALKER games.

    An eye for realism in the way light interacts with a surface, and even the most basic technology to simulate it, was adequate to create that compelling atmosphere in one game while another could lack it entirely.

    The complex, varied, hardware-heavy and expensive-to-develop technologies that increase fidelity today are nice, but they were not necessary.

    I remain convincingly drawn into a well-lit and drawn 3d world from fifteen years ago if the lighting, surfaces and the rest of the game’s features click together properly.

    The more technical tricks, in fact, the more abberrations and mistakes begin to annoy my eyes – evidenced by spending the last hour trying to force features into Mad Max using the NVIDIA control panel to eliminate the dithering and aliasing that defies the game’s in-built FXAA.

    And when a beautiful horizon-shadow at sunset turned into a jagged mass of aliased triangles when I was walking the desert last night, that really knocked me out of the game for a minute >.>

  30. Flit says:

    That’s “MGS V: Ground Zeroes”, not “4”. For a moment I thought you were talking about a PS3 game.

  31. Pizzacheeks McFroogleburgher says:

    VIDEOrealism would be the greater goal. Looking at the current state of the art in games and movies it’s motion that fails to trick the human eye rather than the static drawn imagery..
    The push for better graphics is what drives people to buy new cards and consoles, but games are sold on merit of their gameplay…
    Best graphics does not necessarily equate to photorealism.. There are some simple scenes in the Stanley parable that approach photorealism already.. But witcher 3 has better graphics..
    Oohhhhh.. Ethan carter looks nice too..
    In summary: meh.

  32. SoulForMachine says:

    Thanks for sharing your feelings, kid.

  33. Arithon says:

    Photo realism is all about selling video cards.

    Is it crucial for games? 100 million Minecraft players would suggest not. A GAME is essential for games. Cut-scenes and purdy pictures are window dressing and without an actual GAME, they are nothing more than filler.

  34. thelastpointer says:

    So this is how people enjoy otherwise utterly boring sequels and the latest on-rails military shooters. Look, wood grain!

  35. CookPassBabtridge says:

    Coming late to the party and stepping into the Lions Den I know. I don’t think PR graphics are essential to all games, but they are something I love and want in certain types of experience. I have been gaming since the Commodore 64 days, and grew up watching the massive steps in rendering tech. I love photorealism because it amplifies the sense of the “lucid dream” that games produce – that awesome feeling when you are asleep, realise you are dreaming but can suddenly take control. I’ve been waiting for that experience in gaming ever since.

    The lucid dream factor makes open world games that have a massive dose of exploration particularly mind blowing for me. There is an almost electric charge I get when, for a brief moment, my brain cannot quite tease apart reality and render, and It satisfies something I always wanted as a kid, sat in front of that little beige oblong that was my Commodore. Perhaps it’s something not everyone experiences. It’s different from immersion. Does every game need it? No – but I think games that have amazing world building and a large helping of freedom (even walking simulators like ‘everybody’s gone to the rapture’) really benefit from it.

    Most recently I have had the same sense of creepiness and awesomeness with Unreal Engine 4 tech demos and the Oculus Rift – the edge where your brain knows what you are seeing is not real, but on a physical, more primal level, something tells you it is. In VR, that sense can be easily shattered if you come across something that looks more obviously like a render, like a ghost train where you see the headless horseman is made out of polystyrene. Its lead me to invest heavily in kit for my PC build. No, Its not vital to gaming, but the feeling it gives me is something I really, really want.

  36. Laurentius says:

    Wow, I have never seen that horrible side of RPS community before under such innocent and pretty basic opinion article about graphics in video games. So many people (most of them with massive backlogs) are like “Oh no, Alex Wiltshire is gonna kill all Indies”, “it’s about gameplay man!” and “AAA industry are giving us boring sequels, thx Alex”. It’s just sad.

    • thelastpointer says:

      I don’t think anyone said that this is killing indies. In fact, indies thrive precisely because they emphasize things other than graphics.

  37. Little_Crow says:

    I think the article has a point, but it’s a bit flame-baity.
    It’s not crucial for Games, it certainly isn’t for genres like turn based strategies, but for first and 3rd person shooters – it might very well be.

    New gamers don’t come into the ‘hobby’ blind, they know what GTA, CoD, and AAA titles look like – dump them in front of Doom and I think many would have a hard time getting past the lo-fi graphics.
    I know from playing Combat on the Atari 2600 there is amazing fun to be had with just 160×192 pixels, but someone who has only seen the latest FIFA, Battlefield, etc might take a little longer to get there.

    My missus wouldn’t identify as being a gamer at all, but she loves Syberia and Broken Sword. I’ve attempted to introduce her to other adventures, like those from Wadjet Eye, but she can’t ever get over the pixelated graphics.
    It’s a shame, as I think she’s missing out on a lot of brilliant games and amazing stories, just because they don’t have photorealistic graphics.

    I honestly don’t believe my missues is that atypical for new gamers, Alex’s 10 year-old being among them. In their particular cases, photorealism really is crucial to games.

  38. Premium User Badge

    Bluerps says:

    I disagree. I don’t think achieving photorealism is unimportant, but I don’t think it is crucial. There are many ways in which games, in general, can develop and get better. Better AI, better level design, better mechanics (whatever that specifically means), better writing, better music, better npcs, better guns, better dogs – and so on. Better graphics technology is just one of these ways, and the others don’t depend on it (though it can be helpful, of course).

  39. Stomphoof says:

    As I have gotten older I have found that I want, more and more, for games to look GOOD.

    Not necessarily photorealistic mind you.

    I like various art styles. Borderlands 2 for example is a favorite as is A Wolf Among Us. GTA V took my breath away.

    But at the end of the day, for me, I will always take story and gameplay over visuals. Heroes of a Broken Land, for example, looks like a game from 1988. Its got pixelated graphics and a very basic style.

    But the game underneath that is amazing.

    But that’s me. I equate playing games to reading a book: I want a story. I want cohesion. I want a world to experience and explore.

    Wizardry 8 is ugly as sin, but fun as hell.

  40. DispleasedEskimo says:

    I feel like calling the importance of photorealism into question is like calling the importance of any visual art style into question. It’s just one artistic decision which supports a given set of gaming experiences. But, it’s more expensive than the others which puts it in a negative light straight off the bat, and also it’s harder to pull off to the extent that we haven’t really seen it yet – so can’t truly say how good it would be.

    Meanwhile, the art styles which can be considered “finished” (ie. they achieved an intended purpose because that purpose wasn’t BE REALISTIC NOW) genuinely could be considered crucial, in my opinion. I for one know I wouldn’t have fallen in love with Team Fortress 2 if not for that iconic art style (that it used to have, zing). Similarly I wouldn’t have spent anywhere near as much money on World of Warcraft if not for Blizzard’s art team. Many-a-time would I buy a month of it, knowing full-well that I wasn’t ever going to get good enough to be accepted into a good raiding guild. Really, all I was doing was paying an entrance fee to get into (a very roomy) virtual art gallery; one with interactivity, satisfying feedback and animation, and a beautifully seamless world.

    So to that end I think photorealism is going to be very important too, for a certain set of games. It’s just more noticeably iterative than the other styles, because it’s right on the cutting-edge of what’s possible. Right now the best it gets is a “oh wow, that’s a step up in impressiveness over the last realistic game I played”, but one day it could be a “WOW, this really makes the experience!”, which is what I think some of the stylised games we already have do.

    • DispleasedEskimo says:

      But then I scroll up and read the “It’s always photorealistic for the time, because imagination does the rest” argument, agree with it, and am not sure what to think!

      I guess it’s important in the sense that an evolutionary arms race is important? Got to keep moving forwards, not because moving forwards is especially good, but because moving backwards or standing still is bad. Or like how a drug addict needs to up the dosage because he builds up a tolerance… erm, not sure how I feel about it now at all.

  41. Shigawire says:

    I would also add that the textile shading in Witcher 3 is the prettiest I’ve ever seen. ^_^

  42. mactier says:

    It’s good, but higher detail doesn’t mean better design or higher believability or higher anything (apart from stats). And of course different design can fulfill different purposes. Including requiring less development time.

    And PC screens have worse colour space, contrast than greyscal (grades of differentiation) than good TVs anyway, so… (Making even Mario 64 look really ace and some modern PC games a bit drab and limited.)

  43. Augure says:

    1st of all, you’re confused not wrong. This feeling you’re describing, as liking photorealism, is about the graphics themselves.

    Why is it that some paintings or CG stills are more immersive than bullshit blockbuster games like Batman, Battlefield and such?

    Because it’s about how you use the techs, meaning the craft, the dynamic, the light, the look, the feel, the filters etc…they account for way more realism than the so-called graphic pursuit of photorealism.

    Also all your example speak to me, we’re on the same, you and thousands of other people. I don’t give a shit and can’t stand a pixels game anymore, I don’t care much for 2D game overall unless incredibly stylised, and I throw away ugly Oculus or 3D game right away if they’re ugly.

    But the problem is that the technology is utter shit. I mean 3D is long overdue as a process that should replace photoshop and video as a way to create experience, ads, music videos, flyers, weird buzz etc…but NOBODY without a bachelor study or 5 years of career can make anything with because the engines and tools are incredible unoptimised, complicated, unintuitive pieces of shit. Yes including VXGI, Flex, any Gameworks, Unity or UE tools etc…