Developing the Art of Fortnite

hey everybody my name's Pete I'm an art director at Epic Games I grew up drawing got into computer graphics while I was exploring architectural and technical drawing as a high school student I graduated from Savannah College of Art and Design in 2000 I've been making games since then and I came to epic in 2005 because I really wanted to make sci-fi art I began as a high poly model er I was focused on props and environment art and most of my work was for Unreal Tournament 3 and the Gears of War series you can see on the screen a bunch of the stuff that I made back in the day and in 2011 we built a small team to develop fortnight and I've been on the team since then I wanted to ask you guys who all has played fortnight wow that's pretty awesome so over the

years for a night has been worked on by a ton of amazing game developers I wanted to come here today and present a lot of the work that shaped fortnight into what it is today and I'm gonna start with our gritty more serious art style that we had way back at the start and talk about how we hit our initial stylized approach and then how we iterated and refined our environments and characters through our launch period of four and I say of the world last summer so the original concept of fortnight I'm gonna read this it's a survival horror game that makes third-person action built around social fortifications in a dynamic day/night cycle scavenged by day defend at night it's minecraft Meets The Walking Dead in an HBO styled world where loners die quickly and extinction is just a heartbeat away social survival work together or die alone these are some of

the original mood boards and reference that we that we put together filter on that concept we had the idea that the world was going to be scary and overwhelming shrouded in mist and be really claustrophobic and lead to a lot of just really high tension high-intensity gameplay I'm pretty sure when I was putting this deck together and going through all of our old stuff this is the very first fortnight screenshot ever these are our original white box prototype meshes that were made to prove out the building system and our walls and how these things would fit together on a modular grid so that we could construct the game we were really excited about getting going and getting straight into development so we went straight into doing concept art and paint overs we went with a lot of the

typical things you would you would expect from the initial mood boards of a lot of grey and brown a lot of really industrial looking things pretty drab but also you know the intent was there to reinforce the initial mood we also may have invented the stranger things creature up in the sky don't tell anyone one of the things we knew we wanted to do was make really really awesome creatures and we started off with this concept of the creatures that come out of the Mist and are there to basically eat your soul so that's what that guy does is he is the soultaker and we've got some other cool original concepts of some of the original creature ideas including the the guy who later became the Smasher you can see him down there with his big tree trunk hammer fist thing and part of the initial brief of the project too is that we wanted to have a zombie a zombie

element we weren't quite sold on that idea because we were like it's 2011 everybody's making zombie games are we really gonna do a zombie game we got to make it unique we got to make it our own so we came up with the idea that the zombies would also be a part of this whole mist supernatural element so the zombies were basically all possessed by the mist and you can see they're you know rotting flesh and missed legs coming out of the front and the guy's skull has popped out of his mouth and he's definitely looking pretty freaky we also started building up our environment art as well to prove out that we could take those initial prototype meshes and make them into usable assets to build a world with so these are our first environment shots out of Unreal Engine 3 so as we were putting this all together we got to a

point in the development of the game where we knew that we were doing this dynamic day/night cycle we had dynamic building and destruction and one of the things that I had started looking at trying to figure out was how are we gonna make a realistic immersive version of breaking down walls you know the idea of I'm pic acting into a wall and it's gonna start to erode away the wall right where I'm hitting it and there's gonna be structural beams and wires and insulation and all the crap that's inside a wall they were like how do you do that I still don't even know how I would do that so we were thinking about that and we were also thinking about what we were gonna do for the long-term of the project and as much as the initial art approach was really cool and something that we all liked a lot we were all passionate about it everybody was excited about it we knew it wasn't the right choice we were thinking about long term games games as a service free

to play games you know these types of things that people would engage with for hundreds of hours if not years of engagement and what we saw as a lot of those games tended to be brighter or more stylized and we even had some data from back in the Gears of War days that people would gravitate towards our brighter multiplayer maps as opposed to the ones that were more of a gray brown D saturated bleak type of feel so we went back to the drawing board on some concept art and one of our guys came up with this amazing piece of art that the moment I saw this I was like hell yeah we have got to make this it was such a cool combination of taking some of the core feel of the project and that mix of like apocalyptic stuff and dirty stuff but also putting this really fun stylized sweeping gestural form to the main shape of the building and

getting bright sunlight and blue sky and a lot of just really beautiful elements mixed in with how kind of horrifying the core idea was so we got into more building more building sketches and you guys can see if you've played battle royale that's the first version of tomato town right there he was actually a clown instead of a tomato guy and we started doing environment props these started to reflect some of the same shapes that we were seeing in that big huge sweeping building form we we have a lot of these props kind of as they're shown here today but we've pulled back from some of the you know human bits inside trash cans is not something we're doing anymore we we we declared pretty early on that we wanted to shoot for a teen rating so if you want a teen rating stay away from dismembered body parts and blood and your game as part of

figuring this out we took a bunch of the existing props that we had and started just putting freeform deformations on them inside 3d studio max this felt pretty weird and unnatural to me given how I sort of thought about things back then and how these are all objects that were designed and engineered to be a certain way and then we're just taking them and warping them out and making them look kind of funny and broken but eventually we got used to it we also took our photo source textures and we started doing a lot of Photoshop work and cut out filters and layering the cutout filtered versions on top of themselves over and over to get more clear color separations into the shapes of the original photo source textures and also you know ramping up the colors and brightness and saturation and some of the other initial modular building pieces we took and

started bringing in some of that really cool expressive shaped language into the window frames and door frames even though we knew we were going to stick with the grid based approach where all of our modular pieces have these you know very rigid square edges we felt we could bring that feeling of the original concept art into the world by just bringing that that gestural form into all of the shapes that exist inside each sort of atomic unit so pretty quickly we were able to get from that initial exterior feel to this and the interior to this and all of this was starting to feel a lot brighter a lot more fun it's still very it was still very set in in the lighting and post process and how the fog is it didn't it

this hasn't proven out you know the full style of where we're at today but we kept making some content in this in this style we got our initial tomato town pizza clown guy and we also pretty early on felt the need to prove out the idea of building out tons of props to reinforce like the believability of the space and figure out what are we actually going to fill this world with that players can harvest and gather their resources and can we actually make this world you know as it gets more and more cartoony still feel believable and natural and something where players have a mapping in their mind of I understand where I am because these are clearly kitchen props these are the first stylized character concepts that were done for fortnight they're really cool but also weren't quite hitting the sort

of Tina rated feel or the sort of endearing quality that we were going for we did like the guy with the skull hoodie so we took him and refined him a bit and we got to our initial concept for a first fortnight guy and what we liked about this was just how clean his silhouette got and how a lot of the sort of grimy details were cleaned up but he still had you know a little bit of that scavenger kind of badass feel so we went in and we modelled him and it turned out really cool seeing this guy with these really cool elongated proportions and feeling like a really really cool stylized scavenger so we got him in the game and there he is we also went back over our creatures and instead of making them terrifying they're a little bit you know they're still scary to some degree but they have more of a sense of mischief

about them and more we used simpler simpler line work in them and we started bringing in some of the swirling shapes like you can see on the guy's claw hands more of that swirl shape that reinforces the idea that they were coming from the coming from the mist we also toyed with the idea of doing more of a standard zombie guy in a sort of funky way but we eventually realized we loved our original husk design so much that we had to do a stylized pass on him as well so you can see we cleaned up all the gore and gross stuff we cleaned up the silhouette we simplified a lot of the details but we had to keep the skin hoodie which was one of the I guess the iconic feature of this guy is a zombie design and that's that's been there since since the beginning so we modelled this guy out and we modelled out the rest of our creatures and we got everything up and running in Yui 3 and

after we got to that point we announced the game at the BGA's in 2011 and we had a cinematic trailer for that shortly after that we constructed a sizzle reel internally and this is like the oldest real gameplay video I could find a fortnight and I wanted to just share it with you guys it's two minutes and a half long and I'm just going to let it play [Music] [Music]

[Music] [Music] [Music]

[Music] [Music] we really liked that der burger head it's pretty awesome so this is a good time to talk about how that was our that was our initial stab at doing stylized

with respect to the original vision of the game and it was sort of our what would I look back on and consider our starting point we really liked the direction of the game we wanted to keep improving it and instead of trying to go into like a really aggressive you know early access launch in 2012 we decided to hold loop game back and keep working on it one of the big things we did there was we converted the game and all of its assets from running on Unreal Engine 3 to Unreal Engine 4 so this was sort of a big refactor of all of our art and from here I'm going to talk about some of the work we did there on environments and characters so we kept moving forward with environment concept art and continuing to continuing to factor in the cool stylized shaped language but paying a little bit more respect to the fact that our buildings would be on a grid we also knew since we were making

the world a believable place we were gonna need a ton of assets to fill out the random level generation system so we took a shotgun approach to doing concept art and started generating just massive amounts of stuff this is a sheet showing just how many different variants of beds there are in the game and we made all of these as as bespoke props and we we did this this concept art also to prove out that that shape language would continue to work throughout a lot of different ideas sorry so we kept seeing coming up in the concept art this cool like kinked line work and how the silhouettes were really funky but some of the things were getting a little bit too funky so we kind of did a little bit of Direction setting and guide work to kind of set things up in a format where people could wrap head around how to stylize these things better so the idea was if you were

starting with a box or a rectangle you could taper it up you could taper it down you could bulge it you could pinch it and we had a few other examples of sort of squiggly lines that we didn't want to do we got into a really easy way for people to look at a real world object and break it down into its major forms and then get rid of all of the parallel lines and make the shapes feel like they belong in the world of Fortnight related to some of the concept art we were doing so this let us use photo reference instead of having to do concept art for every single thing here's just a couple examples of some things that are typically extremely mundane props in most game worlds but when we were working on this we were really excited about building all this content because using that stylized shaped language and the bright colors and the textures really made these sort of normal everyday things feel much more interesting to look at our awesome stylized shapes didn't

always work when we eventually got to the point of adding crates to the game our crates did not stack which set set some of the level designers off a little bit so we made some alternate versions of our crates that did stack and we were able to solve that problem that way and then we had other examples where the shape language was also kind of fighting with gameplay and gameplay design where we had some weapons and some guns where the style was just getting too exaggerated and too funky where if you looked at the third-person shooting view of of the gameplay of this you can't really tell what the character is aiming at similarly in our player built walls we were able to use our shape language really effectively on some of the broader shapes like our metal panels but some of the finer shapes like the level 1 wood or the level 1 brick like the way we were building these assets we

couldn't really keep consistency in the shape language all the bricks for example are all pretty rectilinear and they don't really exhibit the exact same hurts so this was sort of a consistency bug that kept getting into my brain we also worked through refactoring all of our textures all of the cutout style photo source textures went through a new process where we hand-painted them some of the failings of the cutout filter were that it would sort of just make a noisy mess out of certain things and we just needed the artist touch of going going through these going through these assets and hand painting them to really make them feel like more intentional and more full of character the props themselves were also all being hand painted at that point and one of the elements from the original style and tone of the game was the lived-in world which I was pretty convinced we had to

keep getting the wear and tear and really intentional bits of grunge and dust and scratches and leaky bolts and all the things that add a ton of character to these assets were were things that we stuck with another thing that we've been doing we're still working on actually as a point of Fortnight being a living project we're going through and adding normal maps and spec maps to a lot of the tiling textures that never had any so this is just an example of kind of the potential of where we can go this is also showing that we fully converted over from colored speck to physically based rendering so that was that was also a sort of daunting process because we had manufactured so many props with coloured spec one of the rendering engineers came to me one day and said hey Pete you know fortnight is the only Unreal Engine 4 game using coloured specular and we

don't want to support that in the engine anymore and I was just like oh so we had a million textures to go fix and we did it and it's it's great because it looks way better now but yeah that was fun we've also iterated the the look and feel of our lighting and post process quite a bit over the years we started off with that original kind of heavy depth of field really claustrophobic feel and over time as we've brighten things up and cleaned up the artwork it's just felt very natural to reduce depth of field and crisp in things up over time one of the problems that took a lot of effort to solve and many years to solve was the overall sort of flat look of the game this is a this is a recent screenshot but it's just missing one feature called distance field immune

occlusion and when you turn that feature on you get that so going from that to that was pretty amazing if you're doing a dynamic dynamically lit game in ue4 I would highly recommend using DFA oh it's an amazing feature given the whole game was lit with just a skylight and a directional light that added the extra bit of depth and fake bounce light that the game had been missing for years so I just wanted to take a second and show off some of the other content that we've made we have a bunch of different environment themes and in the save-the-world side of fortnight we have some pretty cool content that took a lot of took a lot of gameplay time and a lot of effort to get to and I feel like a lot of people probably haven't seen it so we have a desert environment that

we've we worked on and we're working on more of this content and we also built a full tile set of sort of Transylvania inspired Halloween themed assets and this was also a perfect testing ground for the new volumetric fog in ue4 which is one of the most fun features to play around with it's very hard to not overdo it we we used it pretty liberally in this this environment but we were also kind of like his volumetric fog the next lens flare like we got to be we got to be careful here so the next thing we're going to talk through is character art one of the things that we knew from the start that

we wanted was a lot of diversity in the game characters are an expression of the player and how everybody is different and we knew we weren't going to go with a full-on character customization system that lets you morph your face or change your skin tone and all that kind of stuff we wanted to do more curated art so we've been working very hard over the years to always make sure that we're we're building out a lot of different looking people so that we can have a lot of different representation when we started fortnight was not about any single character every character was this scavenger person a survivor in the sort of bleak world and some of those initial designs were really cool but we also knew we wanted a ton of variety and our initial approach to getting there back in the Yui 3 version was character customization you saw some of the

results of this in the video we played earlier where everybody was running around with the burger hat what this sort of led to was a bunch of really really silly goofy stuff we didn't really have a cohesive way of doing tons and tons of scavenger gear as as the thing that was going to define the IP so we did wacky hats silly backpacks and after we went through some of these accessory designs we manufactured a sea of characters that are all things that are kind of still they're still in limbo in our in our in our content basically they're all they're all not really enough meat to build an IP from there we could get like the irony of being a beekeeper or a Canadian Mountie or a raver girl or

whatever idea we came up with we were kind of thinking about this in the same way that we thought of the environment content where we felt like we're just gonna make everything and it'll be cool but what we found was we got a lot of feedback from publishing groups and a lot of people who were looking at the game feeling like the characters really weren't heroic enough or cool enough to really be appealing to a wide mass-market audience and we did sort of a key art exercise to look at them together and you know they look nice from a technical perspective but they're kind of a hodgepodge from a design point of view so our solution to this was we converted the game from more of a player way type of experience to more of a class-based design in our first classes that we started looking at were quarter the gameplay were shooting and building and we started designing these outfits on our existing body that we had and we

felt like they were okay but they weren't really solving the problem that we were we were facing which is making really cool awesome heroes so we came up with the idea to do new body types and here was some of the first past stuff that we came up with for our soldier and we we worked with a lot of different people on these concepts these what these revisions were all very anime they were a little too noisy they just they just didn't feel they felt cool but not like the game we wanted to make and eventually we got to this guy who you might recognize jonesy this was like another aha moment in concept art in the development of our game where we saw this guy and we're like hell yeah that is it so we took that and just pasted him into a screenshot of the world just to verify we weren't we weren't crazy and we said yep that feels pretty good so let's go

make him and at the same time we were making him we started building out concepts for the Builder guy and we wanted that to be on the large body type so we started prototyping what that would look like and doing silhouettes and started building out a dinosaur chart and eventually we got to full concept on the large body type constructor and we looked at any we said man and this is a nice piece of art but we're kind of back to square one where he's like kind of cool but not really cool and we weren't quite sold like the soldier was kind of automatically cool because he's a soldier but this guy was like I don't know so we kind of took a step back and started to think a little bit more high-level about what would a whole entire cast of characters look like and what are those things look like in a lot of other properties and we looked at everything we looked at every

reference point you could imagine we looked at all the movies looked at all the games and the things that kind of stuck with us as we were looking through things were the Ghostbusters the Ghostbusters property was really cool because well we didn't like that they were all the same guy wearing the same outfit we knew he didn't want to do that but we really liked the idea that they were fighting magic with technology and we also looked at the really cool sense of diversity that GI Joe has where you can have a ninja with a soldier with a American Indian but it's we felt like this was presented in a cooler way than our old sense of variety was and then with the matrix all of those characters have this amazing sense of style and they feel really high fashion and they feel unique from one another but also cohesive for the way that my mind was working around the idea of cohesiveness just the idea that they all had sunglasses as a thing just tied it all together so we

came up with this idea that we were gonna represent all of our class based hero abilities with these sci-fi gadgets that would explain away our ability to do magic in the game we went through a whole entire process of figuring out what these were going to look like what we landed on was the idea of referencing a lot of the a lot of the movies that we grew up with you know at epic a lot of us are 80s kids so a lot of the cool sci-fi movies from back then where things like Robocop and back to the future and Ghostbusters and we tied into a lot of those materials and shape designs and really trying to make things feel like they could have been built in Doc Brown's garage as opposed to coming from you know a sci-fi Factory or HR Giger or years for you know doing this sort of typical we have one of the rules

in all of our sci-fi is we don't have any of the tech lines with a 45 degree angle cut and then the line continues like yeah so we we stayed away from what we felt were kind of tropes sci-fi and tried to do something that felt more like a pop culture reference and so after going back over the constructor where we got to was this version where now he's got gadgets in gear and a bunch of cool stuff on him and we felt like yes this is pretty awesome he's even got his cool deployable base powerup device so we went through all of our existing heroes and we built all the models based off of the concept art that we had and incorporating all their new gadgets like our ninja has his jump boots and our Outlander has the power punch you know rocket-powered fist glove thing and our constructors got his power pack

and our commando the commando kind of got the got the short straw and only got grenades but we still made them cool sci-fi blue glow techy grenades what we found was going back and looking at looking at all these heroes and looking at the concept that they came from we did them all in this style that was from the original Jonesy concept and this style was sort of an attempt to get to a lot of creases and planarization 's in the face and we went through and we diagrammed how this could work on face Anatomy for any person or any any hero sort of a starting off as a generic guy what were the what were the creases and planarization x' that you would make how would you simplify a nose how would you simplify the lips to really fit into for tonight style and so our first version

of this came out came out pretty cool is how how we felt about it and we stuck with this for quite a while and we built out a bunch of our heroes in this style and the more time we spent with them the more we kind of realized like hey you know maybe they're not really as good-looking as they should be like maybe that what we did with the planarization and the cuts of the face it wasn't really looking attractive or appealing it was mostly done to try and fit in with a lot of the edgy angular style that had been done in the rest of the game so just pulling like our constructor head into Photoshop and doing a very quick brush over the the cheek bone was kind of like alright this feels way better we gotta go work through all of these based designs and make everything feel more natural and we even went over some of our

character concepts and reinvented them entirely with more of the eye towards the fashion idea of just fake making them feel like they had cooler more relatable clothes and making them a little bit less sort of videogame II she is a good example of where we reduced the number of planes in the face quite a bit and just had the sense of a crease on the jaw and just a little bit of a crease over the crest of the nose and a little bit of planarization in the hair and that was enough to make her really feel like she fit with the rest of the game but also letting her face be beautiful at the same time and we went and did this over her soldier as well and reworked his body proportions and reworked the face but we still weren't totally sure if we had nailed it yet we wanted to do everything we could to be positive that we were on the right track with our style so we kept we took the

newer soldier face and we started doing a bunch of paint overs and a bunch of new style explorations and we started showing these to more and more people and getting people's feedback and testing them and getting data around what we were doing I'm sure you guys are disappointed we didn't go at number four and it turned out that we landed on something pretty close to the original take on that guy and and we weren't as far off as we thought all we really needed to do was go through and clean up clean up the excessive edging and excessive planarization and rework the proportions but sticking to the same initial character designs so you can see the the guy on the right is definitely

comes across as more appealing and more endearing whereas the one on the left like when we made it we were like sweet this guy's awesome but after taking the time to go over it once more we're like oh we had that artist moment of like that thing I made that I thought looked awesome I looked at it a year later and I was like what the was I thinking like the new thing is so much better so we also take a got into expression testing just to verify that this stuff was going to work in the long run we didn't have a ton of animation work going on at the time so we had to kind of force this through the pipeline to see see how well things would really work and using that newer shape newer smoother Anatomy to inform the new the new bodies helped out a lot as well we cleaned up a lot of the edgy angular creases and like the calf

muscles and the forearm muscles and kept we kept things angular and we kept our silhouettes really efficient but also we're able to find the appeal that we were missing before at the same time we were working through our female body types and our medium female and our smaller female were we're definitely easier and more natural to make for us than the larger body type who required a bit of iteration when we first started concept art for penny or female constructor we were set on the idea of making her come with the male constructor in PvP by making the silhouette be exactly the same and this idea seemed it seemed like it was a sound idea from a gameplay perspective and we went back and forth on concepts for this and we tried to solve solve her by you know doing a bunch of different face concepts and a

bunch of different outfits and a bunch of variants there but ultimately what we had to do was get away from the like fact that she was a larger set woman but had these really skinny spindly legs she just looked really unnatural so we went back to the drawing board and did some new concepts we had a more natural form to the body we got rid of the idea that we had to match the PvP mode so precisely and this was a sort of first pass concept of penny that was more country girl and then we pulled in the sort of cool jacket and some of the tech tech elements on the leg and gave her more of a cool heavy utility belt and here was our final final version of penny so we were really proud of this character it took a lot of development

but ultimately we landed in a place that that felt really good we had a ton of positive fan feedback on this character and after we going back and redoing everyone in the game this was our final cast of Heroes that we shipped with last summer so I'm getting close to the end I wanted to define or wanted to just talk through a few final thoughts one of the big things that I sort of learned on fortnight was we defined a lot of rules and for me being a first-time art lead and art director I didn't really know how to form a rule and then figure out how to break a rule things like the shape language became kind of an obsession for me given my background and in modeling I was very concerned with form more so than the overall picture

and I struggled to see the forest through the trees when I was judging individual models so we tried really really hard to force that edgy style into everything and ultimately it made it take way longer for us to get to where we needed to be with our characters it's understanding that the techniques you use to develop a style can be used don't need to be used everywhere took a long time for me to understand like if you want the impression that the game is fun and wacky and vibrant and funky with all of the shapes that it has you just have to use them in the most prominent places like players are constantly interacting with door frames and windows and things like that which is where we would really have to push those shapes and that style versus feeling the necessity to apply that rule to every single asset so understanding that the art style is the sum of the

parts not repeating the same idea over and over another takeaway for me on fortnight is just the idea of getting out of getting out of my comfort zone I think given the types of things that I was used to making before I worked on fortnight I never really saw myself getting into getting so into stylized art I thought that when I came on the team at the start it was going to be something that we would turn around quickly and ship quickly but what it turned into for me was is an avenue to learn how to do a whole new kind of art so it's been a real journey exploring art style figuring out what works what does it work and the payoff is now we have an art style that people look at and people consider it unique and that feels really rewarding to myself at the team and my final

thought for you guys is the biggest thing the biggest test on for tonight was basically we knew that we knew that we believed in this game the team believed in the game I believed in the game and I believed in the team and the biggest point there is don't give up when you're working on something that is taking forever and is a slog of redoing and redoing and redoing I just wanted to tell you guys that finding a way to persist and finding a way to do what you can to take control of the little things and find a way to make them better for yourself while you're going through the process of having to slow down to speed up and having to redo things and having to get through like big demos or other other things that are forcing functions for

you to be maybe not necessarily doing the exact thing that you would prioritize I would just encourage you that if you believe in that project just stick with it and don't give up so I want to thank all the people who've worked so hard to make Fortnight happen and I want to thank you all for coming to see my presentation thank you [Applause] [Music] I have time for questions I finished slightly earlier than I thought I would but anybody have anything I believe you're meant to go to the microphones is what I was told just one to me barely speak up a little so my

question is you shared some great examples of all the iteration you guys did over and over to kind of land on that style did that was that factored into the schedule from the get-go or the philosophy of how you're approaching the game or did it kind of emerge organically and schedules had to change and also how did you incorporate feedback in that whole iterative process where there was internally or externally and can you give examples of that that is a very good question building fortnight was a very organic process within epic there there were a lot of things that came up along the way where the team was asked to sort of change target or involve the project along with the fact that epic as an organization was evolving from being a company that shipped boxed products to being its own publisher so there were inflection points along the development timeline where it was like your schedule

is basically thrown out because we know we're just doing a whole new thing now I know that's a little bit vague but you know things like we're gonna do we got an opportunity to do a game informer cover and something like 2014 so we automatically added time to our schedule of developing the game to deal with the fact that we were creating demo assets all of a sudden when otherwise we might not have normally taken the time to do that or we did a staged demo at one point with Apple to show off the metal API and that that was the whole thing of like alright so the people who were making the game are making that demo so now we kind of need more time to ultimately deliver and just the way that at that epoch tends to work is people will be allocated to a project and then there might be a project that has a great opportunity to do it a demo or do something and those people might shift

from this project to that project so the development timeline of the game was I'm not quite sure how intentional anybody can really say it is in hindsight and the feedback process whether you're doing it internally or bringing in X like how did that factor into your just your iterative process oh sure we had a few moments where we you know before we were fully I think fully sure that we were gonna be our own publisher we had some publishers come in and look at the game but at some point all of the feedback was driven internally and then we also had probably two to three years of beta testing and user experience testing in our UX lab inside epoch so we were getting feedback through both of those mechanisms of mechanisms as well some of the character are we we put in two we put a bunch of images into a survey and we had people go through and kind of rank you know do you how do you

feel about this character is it I hate it is it mad Swiper I need to cool is it awesome so we went through that on a bunch of our character art and you know for a lot of that stuff that was in that middle space it was kind of it was kind of 5050 and we were like alright we need to do better than that thank you mm-hmm hello hi over here so my question is with such an organic way for you to arrive to the style that you got in with it being such a long process how do you ensure that the way that you stylize the environment in the way that you stylize the characters is consistent and also do you find any challenges there were perhaps more challenging just based on the fact that you were working in a stylized environment sure I think the one of the

big things for me was just learning that I didn't have to make the characters feel exactly like the environment that we could get away with having them take on a bit of their own style which wasn't apparent to me at the beginning I was always like when I was a kid I would watch cartoons and see like okay I can tell that there's a there's a water color painted background but I know exactly what the rock on this frame is going to move because it's clearly cell painted right and that inconsistency it would be like a little bit of a deal-breaker for me just in how I have a little bit of an analytical side on top of you know my more artistic side so learning that we could do the characters in their own style was was the secret to making them good because I think the early iterations we were trying too hard to make them analytically cohesive with the environment and was the second part

of your question again can you repeat it do you sorry what were there any challenges for you that were even more challenging just based on the fact that it was a stylized IP for example maybe in building modular sets for example for the environment yeah I mean going with that shape language definitely complicated some things you know you have to be really concerned with your collision and how players move through the world doorways for example you couldn't have them be wide at the bottom and then at the top because a player will run in and their head my clip where it's angling up there's all kinds of like tiny little gotchas that we went through but overall I think a lot of it felt felt pretty natural as an evolution of what we'd been doing in the past I made a lot of modular environment sets for Gears of War and UT so we had a lot of experience there and we were able to

get get off the ground pretty quickly with for tonight with all that knowledge we already had do you find yourself having to standardize a lot of things like doorways just to so it would fit all of your characters or you guys we yeah we have like a minimum kind of a minimum size like we want to make sure that all the big characters can get through any portal and then we have other we have other standardized sizes for AI as well and the PvE game so there's there are definitely a lot of rules there that had to be learned over time just one more quick question you mentioned earlier that you're trying to transition from your hand-painted textures into adding normal maps to them how are you finally that transition are you having are you running into a lot of issues where like maybe the normal maps are not really helping sell the the look I guess or the appeal of the textures we haven't done tons of them yet so I don't

have a thorough answer but I think I know what you're getting at because ultimately it's probably gonna be a little bit of a rabbit hole where a lot of the diffuse textures have light and shadow already painted in tonight and we're gonna have to pull back from that and probably let the artists have a little freedom to interpret those textures and say alright well this brick wall isn't going to be exactly the same anymore I'm kind of redoing it with a new I'll do a model in ZBrush and we'll bake it out again and so there there will probably be a lot of that in that work okay thank you hmm I love the look of the game with look you arrived at the success is really well deserved I think there's a pretty simple question and you almost touched on it scaling of props is a conscious decision to keep things maybe a little larger than would be natural compared to the characters mm-hmm how did you our props are huge yeah they're

enormous it's actually pretty ridiculous and a lot of our marketing materials required us to like go be like alright let's maybe make this rock or this car a little bit smaller and the intent there is you know you can only run so many so many actors in a level and you also have the gameplay consideration of every little thing a player is running to to harvest is a new gameplay objective so we definitely had a point where some of the props that we were doing were too small and we just worked hand in hand with the level designers to figure out like okay there's really nothing smaller than a mailbox in this game because there's no point to it it's just gonna slow people down or get in the way which you know in some respects it makes it so you can't as finely detail certain areas as you might want to but the benefit is that it keeps your levels pretty clean and easy

to read and people can plot their Coy's of course from point A to point B pretty easily when you have the right scaled props right thank you mm-hmm a great talk by the way and I think big fan of the game have been playing it a lot thanks my question is as an art director you said at the very end you were talking about persistence and how finding that persistence to continue to go even through constant iteration and reiteration is very important so as an art director were there challenges for getting the team on board for reiteration and constant iteration is that one difficult yeah for sure we just basically had to keep talking to everybody and let them know hey this game's gonna come out and it's gonna be awesome and just get everybody hyped up about the future of the game and how we know we were building something really unique like just the idea of going into a world and smashing a bunch of stuff and getting all the getting all the crap

and building a cool fort and all of the you know meta game systems that we built are are fairly compelling once you get into them if you're that type of gamer so we knew we were onto something and I think a lot of the team believed in that thank you I am working in the launch department from our game company you just say every time you improve your character or sometimes you improve your culture you are disgusting with the launch team right and when when when when you are two team have different options how did you convince them or how did they convince you that's a question sure I mean for us so is just about transparency and showing people work in progress and talking talking him through

where it's going and I think every step that we took was it wasn't hard to get agreement Haun as being the right step so it was a fairly organic process I'm sorry that's a little vague but we we just worked hand in hand and eventually we got there okay fine first of all thanks for the amazing talk and my question is what there any artists like either concept artist or 3d artists that were like very specialized in realistic style but was very not getting used to this kind of style how how would you like help them to overcome this obstacle sure so we have we have a number of people who have spent a lot of time doing realism and I'm one of them it's mostly about coaching through the ideas of simplification exaggeration if you're

gonna put a bolt on a thing that bolt should probably be five times the size of a real-world bolt and one of the benefits we had was we built a ton of content early on with a really small team who all kind of got the style so we had a ton of stuff that we could point people to and say well the thing you're building is kind of similar to this so why don't we look at this example you just had tons of visual examples for people to look at and follow and learn from so most people were we're pretty able to make a transition you know it was definitely there's a different level of speed certain artists will take to do that but everyone who's come onto the team has been able to get up to speed and be productive within a couple of weeks thank you so much mm-hmm just a quick question thanks great talk and thank you love the art style um it's fascinating to see the progression especially how you smooth the face and zip reminding me a lot of

stop-motion like like men do so my fanboy question is any chance you'll release like an art book over a website showing our progression or toys oh man we'll get there yeah we're we don't have a ton of people in the department to to take on those types of things but we have a ton of source material and I'm really excited about the idea of doing that so I'm right there with you I hope it happens any last questions did you make some significant changes in the engine's rendering to deal with still-life visuals the big feature that we got for fortnight was the one I talked about a little bit in the lighting which was the distance field ambient occlusion I think if it wasn't for fortnight we probably would not have come up with that solution it was definitely custom-made for what we were doing on fortnight and one of

the other one of the other features that we really special requested about a year ago or so was the volumetric fog so that really improved a lot of the depth and the image and being able to see fog that actually shadows in the world was was pretty dramatic but for most of the course of the development we were working with all of the engine builds that came from the engine team without a lot of custom features for stylized in fact most of what we incorporated was more more of the features that are geared towards filmic and photo-real and where that sort of is going for us in the game is we're maintaining the style of our content but I think the rendering style of the game from a technical perspective is actually skewing more realistic and closer to becoming real-time rendered Pixar film style of visual where we're going to get a little bit more material

definition and material separation and all the things that you would associate with more of a high-end rendered thing thank you hi I love the author by the way fantastic looking love when individual artists are working on iterating things that they keep ownership of the asset they were iterated through or did you tend to swap over artists it's a good question it's a little bit of both I try my hardest to let an artist keep ownership of a thing I think that's a sort of a man at the management question of how do you how do you coach somebody to to get it to the point where it fits into the game and is shippable and their making they learn from that asset so that the next one they make they're doing it at a higher quality level and faster so I try very hard to make sure people hold on to

their work and but there is also some times where the schedule just doesn't allow for it and you're like all right we're gonna pass this thing off to this person and this task might go over here so you know over over the many years we worked on it a little bit of both happen but my personal preference is to make sure we're giving the feedback to the artists so that they know how to do the work correctly thank you hmm anything else awesome thank you guys so much for coming [Applause] you