The Evolution of Video Game Graphics Is Complicated

foreign you know I think the PlayStation 4 was one of the most influential consoles of its generation and possibly of all time and no this isn't another console Wars video the Xbox one was pretty cool too but the reason I bring up the PlayStation 4 in particular is because I think it happened to have some of the best looking games for a console during its shelf life [Music] foreign God of War Horizon zero Dawn Uncharted 4 and the list just goes on all of which were technical Marvels at the time of each respective release with articles frequently praising how each game pushed the bounds of what a box from 2013 and quite frankly Hardware from 2012 could do I distinctly remember reading how Uncharted 4 might have the best console Graphics of all time when it came out only for Horizon zero Dawn to be given the exact same title one year later and God of War a year after that and The Last of Us Part two two years after that and that's not to say Xbox hasn't been killing it also things have honestly been looking great since Forza Motorsport 7 which was an absolutely beautiful game for the hardware it ran on and Forza Horizon 5 which let's face it looks so good and ran so well every YouTuber and their mother couldn't resist using it for an 8K gaming Benchmark at some point in the past I mean look at this [ __ ] it's freaking beautiful not to mention all the cross-platform games that have come out and completely shattered all expectations when it came to Fidelity I mean Modern Warfare 2019 came out and damn near looked like Naughty Dog developed it on the weekends which is pretty surprising considering Activision prefers to spend his time doing [Music] other things and you also have games like Red Dead Redemption 2 cyberpunk 2077 when it works Metro Exodus and so much more like seriously if you told me in 2013 that this Boeing 747 would be able to run games as gorgeous as this I'd call you insane so the obvious thought becomes we as Gamers have been notoriously bad at predicting the pace of graphical advancement in the past I mean every year it seems that the quote-unquote best looking game ever comes out and everyone thinks that's as far as we'll get as a species but then the same thing happens again next year and everyone is like oh so beautiful so pretty but I think it's time someone finally asks are these Graphics actually getting better I mean I can't be the only one who has noticed that the year-over-year improvements in game Graphics aren't as big as a jump as they used to be in the past a game from 2002 looks nothing like a game from 2006 whereas you could reasonably make the argument that games from 2018 looked pretty much the same as they do now and even if they don't why has Graphics progression gotten slow enough that people could reasonably think they do without sounding crazy is it even possible to make modern games look better than they do now short answer yes long answer yes but probably not in the way you'd expect allow me to explain and fair warning this video is going to get pretty complicated so you might want to grab a seat so first let's lay down a couple of ground rules and establishments you know to just completely weed out any of the common misconceptions the average gamer might have about Graphics in general the overall goal of this video is to be more prediction and less speculation so it's pretty important that I get these points out of the way even if you're a person who's already well aware of them those being number one Embrace history but don't be blinded by it while looking to the past is extremely important for pushing the medium forward either because we look for inspiration from those who came before us or because we try to use it as a predictor of the pace of graphical Evolution we cannot assume things will play out the same way they always have just because Graphics have got x times better in the 40-some years since 1980 that doesn't necessarily mean we can expect that same jump from now to 2060. graphical evolution is far from a linear function number two Graphics aren't magic this might seem a bit counterintuitive to someone who hasn't really thought about it but behind every single good looking game is hundreds upon thousands of hours of human labor and Manpower attributed to the concept artists that bring text to life the sculptors that model realistic assets and software like zbrush the animators that rig and make 3D models move and debatably the most important the software Engineers that design the game engines every creator has to work with all of that stuff takes Innovation genius and coordination never in the history of game development has making a good looking game taking more time and more work than today so simply assuming that Graphics will get better solely as a function of time without taking into account everything else that goes into it would be no better than thinking that flying cars with a natural progression of automobiles while the concept is really cool creative and ambitious it's also very naive ignorant and ridiculous and number three there are limits we will one day reach we are only as fast as our Hardware assuming everything else manages to fall into place perfectly we will still be limited by the technology and methodology we have around us it's really easy to believe that there is no ceiling when it comes to how far we could push graphics and that they'll just always keep getting better but this couldn't be further from the truth it might seem reasonable especially in the very narrow view of time a single human has but it's not the full story there will come a day where Graphics will be the best they could possibly be and will never get better just like eventually the fastest 100 meter Sprinter will be born in fact some say he has already Graphics will stifle in the same decade we invent our best graphical technology the question is how far away is that day and in what shape will this technology arrive or God forbid has that day already come so keep these three things in mind and when you watch this video I know not many people take them into account when asked about the future of Graphics but I think these three monoliths are cruxes in answering the question so without further Ado let's do this when the first home video game consoles came out people didn't care how the graphics looked they were just amazed that they worked you have to remember this was in the age of VHS in absolutely massive 19-inch television sets the fact that you could even play Pong on your cathode ray tube or CRT TV was enough to make grown men smile so naturally you could only go up from Rock Bottom there were no expectations Graphics weren't so much as a thought on anybody's mind games work but the funny thing about games from the early days is that they were typically created as singular and distinct entities as in no game from the 70s really shared assets or code with one another even if they were in the same series which is a stark contrast to today where pretty much everything is reused whenever it makes sense to do so a game for the Atari 2600 for example which released back in September 1977 had to be designed from scratch every single time didn't matter which games they were or which series they were in nowadays it's very rare for a high budget game to ever be truly designed in the same way from quote-unquote scratch heck 99.9 of all modern Indie Games aren't even built from scratch so why you ask the developers in the 70s and even 80s go through the pain of essentially redeveloping every game they ever made from basically nothing well there's only one reason Atari 2600 games like most other games throughout the 70s and 80s were written in Assembly Language which for my computer engineering and science Majors out there you could see why that might be a huge huge problem most of them used Assembly Language okay the game the game boy used Assembly Language we used a program called brief b-r-i-e-f from a company called underwear yeah you can't make this stuff up and and brief was an editor program in Assembly Language you're writing load the accumulator with this load this you would type the stuff in there and it would run it but you see it had to convert it it was it was using the z80 Assembly Language if I if I wrote stuff and had to convert it to that see Assembly Language isn't your normal coding language like C plus plus Java python or whatever else you might have heard of in fact I probably wouldn't even call it code at all for one every single processor comes with a unique Assembly Language from the factory as it was designed this means every chip if significantly distinct from another has a completely different Assembly Language that is not compatible with other processors whereas true software-based coding languages can run on any chip given the right preparation unlike software-based languages Assembly Language is a set of written instructions that are used to physically control the processor literally telling it what to do line by line sometimes binary bit by binary bit it's about as close as you can get to literally typing zeros and ones into a terminal otherwise known as machine code because only machines can reasonably understand it without actually getting there when you write an assembly language you don't have the luxury of defining phrases using algebra or using any form of spoken language or taking any shortcut even something as simple as adding two numbers together and displaying the result in C plus could take 25 lines in Intel and amd's x86 assembly you need to constantly be aware of how your memory is being used and how many operations you're performing simultaneously in assembly the processor doesn't do that for you the training wheels are off you just kind of have to know so imagine trying to change the color of a single Pixel on screen when Mario steps on a Goomba then doing that for every other pixel as he walks across said screen all all in Assembly Language you can see where things might have gotten out of hand to code eloquently in Assembly Language you pretty much had to be a whiz in being able to convert very low level operations like adding and subtracting into high level visual abstractions like colored pixels and animations somehow some way while most people today can't fathom how their online game of Call of Duty could possibly be just a bunch of zeros and ones the developers of old given enough time could probably dissect exactly what's going on from zeros and ones to firing your gun why how because that's all they basically did in the 1980s they had no other choice now by the way see uh was an option but in those days the C compiler would have taken up the whole cartridge there was no there was no memory in the thing except what was in the cartridges a little bit of motherboard memory for some swapping of variables and stuff but you plugged a cartridge in and it would have eaten up the whole cartridge and you could order larger cartridge but you paid for that cartridge you know with someone those cartridges like 20 20 bucks or so out of a a sixty dollar game you know you got a 20 butt cartridge you know coding in a more traditional language like C or C plus plus would have without a doubt made things infinitely easier but there were no coding languages to speak of this is because the coding compiler the set of instructions that could take your C Java or whatever code and convert it into Assembly Language which can then be taken by the assembler and converted back into machine code for the processor to actually use took up way too much storage space you see the only RAM and storage in game console like the NES for instance could reliably use came from the game cartridge not the console itself this means that if a developer wanted to include a high level C language compile so that they would have an easier time during development they'd need to put it straight on the cartridge increasing the amount of storage it would need to carry along with any other additional resources they might need to run it this would often lead to quite laughable situations where the C language compiler might require two or even three times more storage than the game itself and in a world where larger storage cartridges cost more money that cost would have had to fall directly on the consumer likely taking the average 40 cost of an NES game cartridge back in 1983 up to a 60 or even 80 dollar premium a price that most game companies rightfully assumed most people wouldn't Fork up so they bit the bullet Assembly Language it was I had a manager at micropros who said that c was going to be the language of the future so so the um my boss decided that we were going to do uh we were going to learn this and now this was in the summer and we were up in a south of New York or New Jersey I think he decided to have this class in the summer every night for two weeks and and he ended up giving up my good summer evenings I could be out having barbecues and you know the backyard and everything so I go down there and he said your teachers see you in the evening so if they try to see and and then we never used it okay then I go to cinema where in California and they're doing everything in C it wasn't until the late 1980s when big storage cartridges started getting cheap and affordable that some game console makers began to include proprietary coding compilers that developers could just stick into their Dev kits to avoid the hassle of Assembly Language altogether they started making games using C or even Atari basic then having the console itself convert the code into the much more messy hard to manage and non-reusable Assembly Language it needed in case you missed that point right there is absolutely massive at some point developers could make their games using a unified coding language across multiple devices and having the console itself interpret that code instead of having to exclusively design a game for a single piece of hardware and starting over every time they wanted to make something new so what did that mean practically well it meant for the first time in game development history it was now feasible to reuse assets from older games and develop for multiple platforms simultaneously so if a company wanted to say port a game over to PC or develop for multiple consoles it was now possible though still nowhere near as easy as it is to do today and that's saying a lot considering the concept of a decent Port still flies way over the head of EA the reason console and even PC exclusivity was so excruciatingly bad throughout the 70s 80s and early 90s wasn't entirely because Nintendo wanted to make sure you bought their console over everyone else's sure that was a part of it but it also wasn't that clear-cut the whole concept of exclusivity was essentially stumbled onto by accident because of the limitations brought on by the physical Hardware developing for two systems at once or even porting over to PC in those times would it have been a port at all it would have been an entirely separate Redevelopment process an expense that all companies opted not to partake in especially since they couldn't even guarantee that both versions were would even turn out the same but in comes the late 80s and early 90s and all of a sudden developers now have the option to work with a unified language since console and console cartridges are now coming standard with their own means of running traditional code so it's no shocker that by the late 1990s cross-platform games were essentially born and never before in history was building upon old work or creating ports Easier by the way I got myself a really nice Mac because I'm I'm doing stuff all the time on it I got the uh the M1 silicon thing on it normally I buy the cheaper one but I deserve a good one that I and and it compiles so fast now I and when I go to build it it takes almost no time I would say over and over again though the really the job has gotten much easier I think contiguous memory is one of the biggest things not having to divide things into little Banks of memory and put it in little spots you know just being able to say here's a list of the graphics and it would figure out where to put it itself in the cartridge you know that that part of it saved an awful lot of time and believe it or not this revolution while often overlooked and under explained was one of the pivotal reasons game Graphics took such a massive leap in the mid-1990s on top of being able to create a backlog of code that could be repurposed if need be developers had also been granted the many luxuries and quality of life improvements that came with developing in a traditional language they no longer had to deal with memory switching or pipelines or direct instruction of the processor and could instead put a little bit more of that Focus toward actually making their game and there are no two better examples of this than ID software's doom and Nintendo's Super Mario 64. and quick side note I know these two games weren't technically the first 3D excuse me 2.5 D and 3D games released ever but they were among the most influential which is why I decided to talk about them C doom and Super Mario 64 have one major thing in common they were both coded in C in fact the entire Suite of Nintendo 64 launch titles back in 1996 and 97 were primarily coded in C so what was once considered near impossible rendering a complex Game World in full 3d with polygons Vector space and proper Collision detection somehow in what seemed to be the span of a single console generation became a staple design feature of every major game release I couldn't imagine trying to make a full 3D game like Super Mario 64 in Assembly Language [ __ ] would just be unbearable development of the original Super Mario Bros in 1985 reportedly took a team of eight people three years and over 15 000 lines of code all four game you could be in a [ __ ] hour and that the same team could probably code in literally a day in modern C plus plus needless to say the introduction of high-level programming was a big win for game Graphics it didn't necessarily make things look better but it made it easier to make things look better but it wasn't only a win for game Graphics don't think I brought up Doom for no reason doom on top of being the progenitor of the modern FPS genre was also one of the earliest examples of game creation using a game engine and oh boy this is gonna get wild real quick and now of course like with unity it's like no comparison because you can actually run Unity right on the PC without hooking up to any platform and see what it plays like on your PC screen and then you can say I want to see what it would look like on an iPhone and it will give you the ratio screen for the iPhone and and for the iPad and you can see what it looks like and then to be able to go back to a to a game from two years ago when it's time to update it and load it in and maybe make a few changes and get it back on the store in a reasonable period of time you know so um and then when you finish the game to be able to get on multiple platforms like I mentioned earlier one of the benefits of using a traditional language to create games is the ability to easily reuse old work if need be a young ID software realized this back in 1990 and decided that reusing assets where it makes sense would be their business model so between 1990 and up until just before Doom's launch in 1993 you can easily see this progression at work hovertank 3D catacomb 3D Wolfenstein 3D and doom all showed that very natural progression of Graphics we are also used to today and despite the fact that they were pumping out like four puzzle Platformers in between each of their major FPS releases and a span of only three years in literally 13 games their FPS games went from looking like this to this the reason all four of these games looked and played so similar is because they basically were the same game with a couple years of optimizations and improvements put behind them and the craziest part is it's software kept going they took it a step further by the time Doom released in 1993 they had already come to realize that they might want an easier way to create the next Doom game even easier than simply reusing C code or going through old libraries and thus the idtech one engine was born a way not only to reuse code but literally Aid in level design optimization modeling and game physics if using Assembly Language is here and using a traditional language is here then using a game engine built on a traditional language is here absolutely groundbreaking and this massive leap was only possible because of the improvements made to compilers storage technology and processing speed all together soon after a little-known program known as Unreal Engine 1 used to power the FPS game unreal in 1998 was created by Indie Studio epic games founder Tim Sweeney and well we all know where that story goes by the late 1990s pretty much every game ran on some form of a primitive engine because they made life easier why code in places where you potentially don't have to it's an awful lot easier to use editing tools than to have to hard code every edit you wish to make so obviously every game company and their mother pivoted towards developing their own engines or using someone else's where legally viable so I call all of that the entirety of that overly informational history I just gave you a basic explanation of what Graphics are and a little taste of the Innovation it takes to push them forward sorry for all the technical information but now that you have a good idea of how all of this works allow me to take you into the field [Music] okay so where can Graphics go from here well allow me to point out a little cycle that you might have missed from the first section that should more clearly explain exactly where we are on the graphics timeline I personally like to call this cycle limitation Innovation adoption Perfection and so far it has been defined by three key ages from the 1970s to about 1990 we have the age of assembly and from 1990 to about 2000 we have the age of traditional code and from 2000 to the present we have the age of engines each defining roughly the primary Way games were developed in the industry and despite being completely different in terms of length they all share the same life cycle the graphics life cycle limitation Innovation adoption perfection in the age of assembly we were initially limited by a lack of reliable and cheap processors themselves so innovations that brought these items to the masses through arcade machines and consoles led to the adoption of Assembly Language as the industry standard for game development and up until Assembly Language became the limitation itself in the age of traditional code it was perfected Beyond reasonable efforts in the late 1980s and in the age of traditional code larger storage sizes on machines allowed for languages like C to be adopted industry-wide compiler and all leading to pretty much all games being developed in a traditional language and the flexibility of game development growing 100 times seemingly overnight I also tend to call this the age of 3D graphics and Rapid Fire game development but just as all Cycles it must continue and this is where we find ourselves today when the age of engines began we were limited by traditional languages and how slow they could be to create large and immersive experiences so some companies decided to create repositories of code and develop interactive tools designed to be used in game development on a large scale a practice that wasn't really feasible in the 80s this led to the mass adoption of game engines in the absolutely ridiculous jump in visuals and performance they've brought us over the last two decades over time we've slowly inched them forward and pushed the status quo to produce absolutely stunning games and at this point in time I think it's fair to say that we are in fact reaching the end of the Perfection stage of the age of engines I doubt we're ever going to see a truly mind-blowing unbelievable leap in game Graphics anytime soon without another massive breakthrough to send us into the next age a breakthrough that would essentially start the cycle all over again but what exactly would this quote-unquote breakthrough consist of heck what does breakthrough even mean well it would have been really easy for me to just split the age of engines in two and say slow processors was a limitation but then where exactly would that split go as processing power has slowly increased from the early 2000s onward so has the capability of game engines so it's more reasonable to say that their perfection is directly tied to the Perfection of game engines themselves essentially I don't think small and incremental improvements whether that be in Hardware speed or game engines themselves can reasonably count as a breakthrough yeah sure over time these small improvements did add up to a very large change games today looked nothing like they did in 2000 but that took 20 years to happen and the base methodology of how it happened never really changed a breakthrough needs to be a dramatic shift in Hardware so great it literally changes how people go about developing games which is why the age of engines has been so ridiculously long compared to the others for two decades we've been mainly relying on Hardware iteration rather than Hardware Innovation to improve the tools we use to develop games heck the base concept of how 3D models are rendered in a game world still has not changed since the year 2000 we've just gotten better at utilizing the new and faster hardware and unfortunately that breakthrough innovation has probably yet to come either because we didn't feel like we need it yet or because it's just that difficult to think of but as we've seen in recent years Graphics won't be able to piggyback off of physical hardware refinements for much longer pretty soon Innovation will be needed otherwise game Graphics might stagnate today as they are either that or our future consoles and graphics cards will have to get comically large and start drawing more power than our air conditioners oh oh no so back to the question what would this breakthrough consist of well it's really hard to predict what an innovation will look like before it happens so I can't really give you a clear answer we're just going to have to trust the professionals and hope they could think of something soon but there is a good chance it might have already happened and that the times are just too young to realize it and the biggest candidate for this breakthrough is Ray tracing accelerated Hardware now I know you guys are probably tired as hell of the phrase Ray tracing because at this stage it really does just seem like a marketing gimmick that Hardware manufacturers like Sony Nvidia AMD and Microsoft can just slap onto their chips so we have an easier time seeing a blurry image of ourselves in like the five total mirrors we have in a game but in all seriousness it's a legit technology that if properly adopted over the next couple decades can easily be that mystery breakthrough we've been looking for now of course there's no way to actually tell if Ray tracing accelerators are in fact the Golden Goose it could instead be dedicated AI chips or even game streaming believe it or not if your game is able to run on a supercomputer in the cloud and you could just stream it reliably to your TV that might actually be what allows the medium to progress nobody knows and trust me that paid me just as much to say as it did for you to hear the point is it will take something radical to change the course of Graphics we're already on I just so happen to think that Ray tracing as a hardware technology is probably our best bet so now I see you staring at your screen thinking I'm an idiot wondering how such a meaningless thing such as Ray tracing could possibly be my number one Contender for a new age Innovation well the reason you probably feel that way is because most Ray tracing modules in consoles or on PC graphics cards pale in comparison to what the true execution of the rendering technique could do and while this isn't a ray tracing explanation video I think I've done enough technical speak already everyone on YouTube has beaten that topic to death so I'll leave one of my favorite explainers in the description but for the purpose of this video there are only two things you need to know to understand what I'm saying number one the ray tracing units you have in your console or even on your graphics card are nowhere near as powerful as they need to be to actually be able to render a full game using the ray tracing technique alone and add a quote-unquote high enough Ray resolution and when I say nowhere near as powerful I really [ __ ] mean it the CGI in movies like Encanto or Doctor Strange use very sophisticated Ray tracing techniques I won't even pretend to understand with billions and sometimes trillions of tracked Rays they also take weeks to years to render on a supercomputer needless to say your gaming machine probably won't be able to replicate true Ray tracing anytime soon and I think this Reddit post on your screen right now does a great job of explaining the current situation game development is in with Ray tracing technology and why its potential is astronomically high but we should also taper our expectations pause the video and read it if you have time and number two Ray tracing as a technology will improve with Innovations of its own a sort of innovation section you might call it the use of Ray tracing in gaming and graphical Fidelity in general can only go up from what we have right now because right now it's simply not that impressive who knows there might be a world where the ps9 and the Xbox Infinium 80 comes standard with massively powerful AI chords that boost its equally capable Ray tracing core's ability to render or they might just allow us to stream ridiculously high quality rendered assets over a network it probably won't be as good as the CGI we see in movies which can honestly sometimes look better than real life but it'll be a whole heck of a lot better than what we have today so while I hope I'm not wrong on this because it will be very awkward if someone in 2050 watches this back and absolutely nothing I said made any sense it's very possible that the next stage of Graphics might be the age of Ray tracing or the age of AI or the age of streaming or the age of insert any potentially revolutionary technology here only time can tell but granted I overlooked one crucial fact when discussing pretty much all of this and that is photorealism probably and I say this as a big assumption probably isn't the direction gaming is headed in you know if we wanted to completely be fooled that what we were playing were in fact real people we just watch movies instead the direction gaming is heading in could very well be stopped [Music] foreign you see there's a reason most modern art isn't hyper focused on making things look as realistic as possible I mean at some point in time humans were really obsessed with getting that realistic look down pat but once we mastered it and the Advent of cameras came around the whole concept just kind of got boring in this day and age most artists can draw realistic art whenever they choose not because they think it's the most optimal way to convey a message but because it's almost seen as a Rite of Passage to be able to do so understanding the many different forms or styles of art is something that the modern artist takes great pride in but modern artists also aren't afraid to completely break the norm and pivot towards something that almost makes no sense yet resonates with the deepest parts of the human brain art as a medium has grown and it's seen itself go through a multitude of phases and changes it has matured over the Millennia it's been around and it will continue to mature at some point it found itself being pushed more and more towards realism but the second it got close to it it it's like a massive detour and started chasing the human soul instead the most visceral paintings in history usually aren't the ones that are the most realistic but instead the ones that strike a chord with the beholder so what if this entire time gaming was chasing a style photorealism not because it was the best way to make games but because it was all our uninspired monkey brains could think of I mean seriously what if the future of games is stylistm what if the first abstract game has yet to be made but what it is it becomes a hit people often praise the Horizon series for taking an idea so wild and unbelievable and actually fleshing it out this series will probably be the last AAA robot dinosaur series you ever see in your life and yet people love the idea behind it and questioned how no one ever thought to do something so irregular before well it might be because gaming as a medium is too young to realize exactly where it wants to go you have to remember the term gaming in reference to video games has only been in widespread use for the last 50 years at best and that is nothing compared to the Millennia it took for Modern Art to develop or even the 300 years it took for the modern engine to take shape it's really hard to imagine that we aren't living with the final iteration of the things we love today but they're both sad and inspiring reality is gaming is Young and gaming will change fast forward 100 years into the future when gaming is 150 years old and it'll undoubtedly look nothing like what you're used to somehow I doubt the people of 1922 would have been able to imagine a smartphone that responds to touch inputs even if they tried to avoid the graphical limitations brought on by stagnating silicon and or game engines we could move towards Graphics as expressionism like how [ __ ] cool would it be to play a game that looks as unique as Arcane maybe the next age won't be the age of Ray tracing at all it could also be the age of VR or stylist and personally I wouldn't mind that photorealism probably and hopefully isn't the future of game Graphics to me what makes games so fun is there's slight distinction from real life that allows developers to take visual Liberties and make for a more fun experience kind of like how Noah watches the Fast and the Furious for real world driving advice you watch it because it looks cool so when someone asks where will the next jump in video game Graphics take us well I don't know but hopefully you might be the person who does [Music] [Music] yo innocence hopefully you guys enjoyed that video let me know Down Below in the comments what you guys thought of the longer run time and almost documentary style I took this time around also huge huge huge huge huge shout out to Ed magnin founder of Magnum games and game developers since the 1970s who took the time out of his day to be interviewed for this piece he's a really chill dude and I'd recommend checking out his website and his mobile games down below you can also find my sources down there as well as a short explanation on why I decided to make this video in the first place with that being said be sure to subscribe it's free and it means a lot but other than that have a great day and or night and I'll catch you guys on the flip side peace out [Music] just keep me

Channel: Innocenceii

Views: 200,917

Keywords:

Id: l5nxZ8Jycc0

Length: 37min 6sec (2226 seconds)

Published: Thu Feb 02 2023