Rayman Legends: The Design Process Within the UbiArt Framework

[Music] okay hello hey I'm supposed to tell you to turn off your cell phones and things like that so please do that and I'm supposed to also tell you just sort of like scoot in and make room for more people but I see that's not really necessary so good job okay I really appreciate you guys all being here this is actually my first time talking at GDC so the fact that any of you are here at all really means something to me thanks so I'm guessing there are a lot of artists and designers in the room that's kind of what I put on the you know schedule so I hope you are and if you're anything like the people I've worked with or me for example there are times when you're working on a project and it becomes difficult to sort of mesh a clear artistic vision with sort of a gameplay direction and combine them into something that works for everybody and make sort of a harmonious whole for the game and there can be a little bit of a

clash between artists and designers and one thing that we've done with our workflow to create the game rayman legends' is build a tool and sort of a pipeline in a way of working that tries to sort of mitigate this issue a little bit and make the process smoother for everybody but also make it really fast intuitive and flexible so that we can produce a lot and produce quality a lot faster than normal so Who am I why am I here talking to you why do you care I'm a game and level designer at Ubisoft Montpellier in southern France I've had the pleasure of working with them for about three years now I've been working exclusively on Rayman titles in the UV art framework which is our tool that we used to build them so I've got quite a bit of experience using this tool pretty much exclusively from my career so I have a lot of experience that I'd like to share with you guys and hopefully you'll take something interesting away from it what's the plan I'm gonna start by just giving you a little bit of a rundown of our design process in a nutshell so you can sort of see why it's

an interesting process or maybe why it's a little bit different from the ones that you're familiar with then of course what's the tech why do we use this tool what have we built to use and how does it help us follow this process that we set for ourselves and then of course I'm going to talk about Raymond it's my only frame of reference so I'm going to talk about some useful systems and tools that we have case studies from the actual production of the game and a little bit of a post-mortem to give you some insight into things that we know went right and things that we know could have gone a little bit better and I'm going to end the presentation by giving you a little bit of information about a couple of other games that are being made in the framework that aren't exactly Raymond so I'm gonna let you guys in on a little secret okay the secret to designing Raymond is actually quite similar to the secret of playing Raymond which is rapid iterative

failure so basically what I mean by that is with our tool with our process we can design so quickly we can try a bunch of different things so fast that we can fail so much more often it sounds awesome right you fail about 40 times a day but then you get one win before you go home for the night and that's enough because tomorrow you have something to work from and it's really great to be able to put all these ideas in front of somebody and prove concepts and disprove concepts really fast and know that you're moving in the right direction so one of these core philosophies we have to work from is no idea is sacred I've seen a lot of people I still see a lot of people do this and I still do this myself sometimes that people hold on to an idea it's like no no but trust me this is the best thing we need this it's gonna make or break the game and they just put like little bandages you know it's like your arm just got cut off but you're putting a bandaid on it it's not really gonna change much it's already messed up it's not gonna happen so why are you trying to fix it and so

sometimes it's just about learning when to let go of an idea and move on to something more promising and that's usually what happens for us is we're able to create so many different little ideas and different executions of the same idea that we learn to let go a lot quicker and move on in the right direction without too much fuss I don't want to waste anybody's time except for my own because I mean I know why I'm wasting it I have an idea I want to see if it works so I'm okay spending a bit of extra time to prove or disprove it but I don't like the idea of briefing two programmers in an artist on an idea of mine and then telling them hey invest two weeks of your time to make a prototype to prove or disprove whether or not this is a great idea because then after two weeks it sucks and it's all my fault and I wasted their time when they could be doing something more important so I'm perfectly happy to spend a little bit of extra time doing a bit of art doing a bit of scripting and some sort of messy combination in our engine to

make something that works or doesn't work rather than actually having to go to someone that doesn't need to be involved yet that allows the programmers in the artists to actually be investigating things that we know are going to be in the game later larger systems things that are really important to investigate in the long term and only when I've proven my concept to a creative director or somebody else that's important do we actually invest the time to make it we also tend to work together but separately at the same time what we basically do is because you know when you're making a level right you you build a level design layer and then it's time for the artist the artist to start doing is art Pass and it's like okay give me a level it's my term to work on it's okay fine but then you're kind of got some downtime and every time you want to work on it you have to sort of like pass it back and forth unless you're working in a more structured way I'm at least that's how I've done it in the past which is a shame so what we do is we basically say this is the level design layer right this is where I'm building my level all the gameplay elements in the game plane no

just the stuff that I need right then there's the art layer it's in the exact same level but it's only the art elements that complement the visual aspect of what I've created and that's inside of the root file of course so you have one seam which is the entire level and then you have sub scenes that are manipulable individually by different people and when they send their update and I cannot I can get the new data and open it when I open the whole file their part is updated and I'm still working on my thing so nobody has slowed down the process and that's extremely important because that means that once I stumble upon that little gem where it's like ah this is the idea that's working let's move on from there and everybody approves it the artists can start working on that already while I'm continuing to flesh it out and work on new things and also we need to remain flexible because later when you're debugging or when you're sort of re-evaluating work you've already done a while ago you want to reserve the right to be able to adjust things without

having to adjust all the artwork to on Raymond it's especially important for us design with the artist in mind we can't just make something that we think plays cool oh it looks a lot fun whatever no but trust me the inputs are great and I'm teaching the player exactly what I need to so it's gonna be great and then the artist has to go in there and say oh but how am I supposed to decorate this it just doesn't work it looks like crap what are you doing it's like yeah but trust me it's important for the game it's like they're not your mom they don't just come in and clean up your mess because you're the designer and you're all-powerful and you say what you want and it has to go and so we try to design with the artist in mind what they do is they create art kits I mean we've got art this beautiful do you really want to mess it up or do you want to compliment and make the whole game feel better so they give us an art kit something like this for example this is a concept but they can actually extract the artwork from it and make a kit out of it so in this case it's a nice you know flat ground some little hills with flowers on it and you say oh I want to

make a right angle because I want to have a wall jump it just doesn't work it looks ugly it doesn't make any sense and so we have to do is think do we need this wall jump or can I think of an interesting alternative or a totally different idea altogether that complements the kit that I've been given and that's one of the cores of the production on the design side of rayman legends and rayman origins that we're constantly thinking about the rules and the constraints that the artists have given us with the artwork they've created and trying to use it to our advantage rather than seeing it as something that's holding us back this is an example when I was doing pre-production for one of the worlds in the game it was supposed to be an underwater base and I didn't really know what the topology should be like and there wasn't any concept which was rare for us normally the artists do concepts and we build on top of those but here we didn't have the time and I had to investigate before they could actually draw something so I was thinking about what could I do to make interesting topology

that would fit with the the see just the setting that I already knew but it had to be solid and artistic and so you can see that for example the larger pipes are sort of the borders and the smaller pipes are connected things don't feel unbalanced it all sorted it's a solid and on top of that I had an interesting gameplay layer you know it's actually a fun level to play so then I passed that to the concept artist and he paints over it and then of course it looks good I mean it's a rough concept but it says hey that layout actually works not bad so then we go through the process of approving things and one thing leads to another the artists make the final assets we integrated into the level the level design didn't change I defined the topology and at the same time the artist had what they needed to be able to make something beautiful but it's because I designed with the artist in mind from the beginning so how is this magical process done it's not necessarily the easiest thing in the world to say we

make five million prototypes every week throw away four million of them we have one little gem and everyone's happy not everybody has that luxury of course we have a very talented team that's important we work well together but also we have an interesting piece of technology that we developed for Rayman Origins and have been constantly updating alongside the development of future titles so the magic tech is the UB art framework I'm going to show you a little video just to get you into the moon and then we'll continue from there and I'll give you some more details on why it's an interesting tool for me as a designer [Music]

[Music] [Music] [Music] [Music] [Music]

[Music] kind of pretty right not bad so as it said the tool was designed for artists so it's it's it was a 2d game engine it always was so it was designed for it has full HD 60 frames per second all that technical nonsense but it's impressive I mean our games are very fast and fluid we can have a lot of detailed textures in there we have a lot of layers and it still runs very very smoothly and that was one of the key selling points for us when building this tech it needed to run well but its artist friendly at its core I mean what an artist has to do to get an asset from concept into the engine in the playable game all he has to do is paint a concept cut a piece out like a

tree or a rock or something that he likes in the scenery save it out with an alpha Channel drag and drop and it's in the engine and they can just scale move it around whatever change the colors change the distance in the Z everything like that but it's that easy there are other more complicated methods of bringing art into the engine that I'm going to show you that helped me as a designer but it doesn't change the fact that it's extremely easy for artists and designers to create content in this engine because it was built with the artist in mind and it's fast intuitive and flexible in everything that it does so just a little bit of information on the Raymond series because that's where all the examples are gonna be coming from like I said I got the chance to work on origins and legends and back in 2011 when we released origins we had four player co-op a large number of levels HD 2d graphics 60 frames per second all that stuff I said right and of course when you're making a sequel

you don't want to just make a direct sequel you want to actually make it meaningful make it something that people want to buy so we wanted to add new features make the engine more robust this is just a small list we have a lot of new additions that we made during the course of transitioning from origins to legends but some big changes we made were online functionality that we didn't have before a lighting system touchscreen gameplay 3d bosses we didn't have support for any 3d objects in the environment before and we added it for rayman legends and the game still runs the same so we wanted to make sure that we're constantly updating our technology to give us the capability to do more than we ever thought that we could with the tool and it also gives us a larger range of things that we can actually do to surprise the player so this is sort of like the the most impressive part of the talk I'm gonna be showing quite a few videos of the tools that we use and explaining sort of why they're interesting from the perspective a designer I think they're kind of gonna

speak for themselves but I'll talk anyway because I like to do it so here's the biggest one it's our game play geometry and for some reason it's called freezes I think it's a French thing but I don't really know and basically you just double-click and you start dragging vertices around it's like manipulating something in a 3d modeling package in up on a plane but what's impressive is that it's graphically complete all the textures are already there as you move the P the vertices around the textures update depending on the angle that the the vertex is at for example so when you have a whole Berta chol face it changes to a cliffside if it's horizontal then you're going to get the grass and of course I can check and see where Ramon's been for the past few steps which allows me to quickly add new elements and improve my level or continue designing the level and working on so that thing was a little bit too high now I brought it down and it feels better and I can move on to the next step so I can bring a platform in and just this ease of use

is really the key to making so many prototypes all day every day and being ok with throwing them away because if it takes you 5 minutes to make an entire screen sequence who cares if you have to throw it away that's business as usual but if you had to draw it out on paper and then convince somebody to help you build something so you can do your sequence and then it takes you 3 weeks and then it doesn't work you're not going to be happy throwing that away exactly I can actually pause while I'm playing too and I can place things or they need to be so I can sort of have real great precision on where my enemies are gonna be placed where my platforms are and then I just press P and I'm back to the game so the most important thing about all of this is the game is live while you're manipulating anything in the environment it's really really important for the speed of iteration for testing your sequences for knowing that what you're doing is actually working or not and everything that we have in this tool just helps to push that feeling of flexibility creativity and allowing

someone to express themselves with minimal hassle and also it's interesting about this visually complete geometry is that just in the level design layer you already have a visually representative prototype of what the final thing might look like you know that it's cliffs and grass and you know when something doesn't look right it's easy for us to fall into the trap of just saying I'm gonna use white box make something that I think could look like a space station and then the artist looks at it and says how am I supposed to make a space station out of this this doesn't make any sense it's too organic it's too this it's too that and what this does is they say these are the things that you can use this is how they work work within those constraints and it changes the way that we have to think about our levels for the better so we have other types of freezes as though we have pipe freezes which are simpler to manipulate because you only need about two or three vertices and everything is already built around it the collision the shape the way that everything looks so it's extremely fast to just build a pipe

Network connect things to one another what you're seeing here actually since we don't have perspective in the game we obviously have to fake it somehow to make the game look nice so we have the ability to add two extremity textures on the same area basically of the fries and we can switch between the two at any time so we just say okay Raymond needs to stand on top of this and perspective wise this looks wrong so you just right click switch texture and then it looks the way it's supposed to it's super fast we also have this thing called meta freezes this is something that we actually designed during the production of Rayman legends so it was an addition later because we realized yeah everything's live even the enemies dying we realized that we were sort of working backwards sometimes we were actually having three or four freezes we had to manipulate at the same time every time and that's not really something you want to do so we did is we combined multiple

freezes in the background the playground the foreground of this platform into one manipulable object what that does is it preserves the flexibility of the tiling of each of the individual freezes without causing unnecessary stretching or compromising the visual quality and at the same time preserving the flexibility of manipulation which is extremely important for us these ones are really cool they look a bit like pipe freezes but they're nothing like pipe freezes they're busy a so they're spline controlled and the collision is dynamically generated based on the spline that you have it's updated in real time obviously and manipulation is extremely easy you can see it's also visually complete like all the other freezes it tiles properly stretches updates you're playing it it's live just like everything else and there are a lot of extra parameters on top of these things that make them even more interesting for prototyping one thing

you can also see here is that we use the same system of manipulation for are collectibles now in Rayman Origins we actually placed every single loom the little collectible guys we used to place every single one by hand and now all we have to do is put a spline to say they start here they end there you trigger it and they just sort of fly to this at the end position so it just makes the whole level design process that little bit less tedious and that makes it fine if I throw it away it didn't take me ten minutes just to place 500 stupid looms it took me two seconds so I can delete them who cares we also have things like the ability to make them grow and retract just with a simple trigger and now of course if you're a designer you can imagine all kinds of interesting things you can do with this and that's not all but it's just one of the many little features that were added that just brings so much more to the table as designer to make creative levels and when it's integrated

into a visually complete package there's no reason not to use it so here's just a little example of a combination of some traditional freezes some busy a freezes some animation and a coin and that's about it it's just a combination of a couple simple systems to build this it wouldn't take more than five minutes for us it's really really fast and I'm talking about visually speaking to the background foreground who cares but the stuff that you're playing on it looks like that when you build it there's no extra work necessary of course every level designer needs white box because there are some prototypes that you just need the freedom to do what you want exactly the way that you want it's cool to have these visually complete freezes but sometimes we don't know where we're going with the world or where the specific sequence fits into the grand scheme of the game as you can see it's still flexible like everything else it's very fast to work with this kind of stuff but what's really powerful for me

personally about these white box freezes is what I can do with them when it comes to prototyping because I can combine them with other elements and I can make my own gameplay basically it's just a mixture of a couple different freezes a couple movement tweens and a trigger and what I've basically simulated is a button that gets pressed and it triggers the door to open and that's just the the speed at which I can make a gameplay element that would normally take an artist in some programmers some time to develop before it can prove that it works I've proven that it worked in about half a day of work and I can easily manipulate it to if the scale is wrong if the shape is a bit wrong or the color one thing I like to do too is white box sketching I'm sorry don't know what happened

I'm gonna continue all this boots up I'm really really sorry I don't know why one thing I like to do is I really take this sort of white box geometry and I sketch what it is that I want it to look like because what I need to do is I have to convince the the artists and everybody of what my my vision is basically what I could do is I could simply make this object a red circle and say it's a dangerous object go and the artists okay well what do you want for this red circle and all that kind of stuff and they have to figure it out but for me I think it's important that it has spikes it needs to be clear why it's dangerous for the player because sightings and feedback are important right and so the ability for me to actually sketch that out makes it a lot more powerful for me to be able to pitch that idea to somebody and and what's interesting is when you look at the final art asset basically it looks relatively similar to what I was able to actually produce with the white box so it makes sense and actually this is one of the things that

I do on a daily basis it's one of the things that is the strongest thing for me as a level designer in this engine is to be able to sketch things out like this so rapidly what you could say is you can just draw it out in Photoshop and drag and drop it into the engine okay why not that's fine that's what it's built for right but for someone like me I'm a hundred percent used to actually working in the engine these freezes they're like it's my Maya you know I don't need to leave the engine to do things and the fact that I can actually do it so fast like this it just makes my life that much easier and it makes more flexible for me to I don't actually have to open another package to make a small change resave reload no I just grabbed a vertex to drag it up does it work cool does it not work okay move it around again I'm very sorry about this I can juggle but basically I mean that's what it comes down to right is everything about

this tool is about empowering us as designers and artists to do our job that much faster than we normally could and that's why all the systems are built that way and honestly working in it is a dream I've worked in other editors before in school for fun whatever but to be fair it just doesn't compare at all to the power that we have with UB our framework I'm very sorry I hope so on my first

time - what are the chances I think we're back for now cross your fingers if not I'll just juggle so what I was talking about and then it turns into something like that when the artist gets their hands on it it's great it's awesome and it's cool that I was able to sort of inspire them in a certain direction even though I'm not an artist myself we have Lua scripting as well so there's a way that we can actually add a little bit more functionality to our visual prototypes it's not just all artsy fartsy but it's simple scripting it's nothing too complicated I picked this up very quickly and I'm not really a programmer so we have is we have simple chunks of behavior like a shake a movement ease-in ease-out things like that and you can just sort of stick one in front of the other it's a series of events and that's how we can do simple

things like moving an object down the screen making a crusher that shakes and comes back we can pretty much prototype everything we need to using these tweens these movement tweens in Lua scripting plus all the tools that I've shown you already in the editor and one thing we also like to do on this side of things is separate templates and instances so basically for all of our gameplay elements we have templates which have a set series of behavior that is sort of like non-negotiable it has to be that way but there's always a couple of parameters that you can mess with a bit to do something different with it and those are in the instance file so that enables the designer to actually manipulate certain things in a way that fits his current situation without the game designer pulling his hair out that you're changing everything about the enemy basically we have a sequence editor - this is sort of our cutscene tool in the UB art framework what's really interesting is sometimes when you go exploring in areas of an engine that you're not supposed to touch you find out that it actually has something to

bring to your job to and during Rayman Origins I was in charge of building a boss inside of the stomach of a dragon lucky for me there were no big animated characters that needed to be integrated with complex behavioral patterns and things like that so I could use what I had at my disposal as a level designer to build an interesting boss sequence so I took the sequence editor I took things like these fire patches and other in just in gameplay elements and I made a series of events so trigger flame trigger flame trigger the weak point hit the weak point if you don't restart if you do and go to the next sequence and I was able to prototype 10 20 different versions of this boss by myself without a programmer and in the end it's a hundred percent level design boss which is something interesting I wouldn't imagine I'd be able to do that but this tool that wasn't meant for it was powerful enough and integrated well into the engine and then other systems that allowed me to do it we also have a very interesting tool that I will not be able to explain to

you it's our animation tool in the engine it's really really cool it allows us to hybridize hand animated keyframes with bone animation and actually there was a talk last year by two of my colleagues where they explain it everything there is to know about jananam so since you all have access to the vault I suggest you watch it when you get back to your Studios it's very interesting talk and it's definitely a cool tool worth checking out especially if you're on the more artsy side of things but I'm gonna push through in this period of level design with some case studies in the production of Rayman legends that you know greatly benefited from the the workflow that I'm talking about basically because of the tools we have in the engine so one of the things that I personally was tasked to do during the production was investigate stealth in Raymond Ramon's a happy flowy platformer II game is not usually something you want to slow down in but we wanted you to feel a bit like a spy and avoid detection but we had to preserve the flow because flow for us in Raymond is paramount it's something that

we need to have all the time so how do you impose that on people we played around with systems we tested a lot we threw a lot away and we finally got to the thing that felt best what I'm going to show you now is a couple little prototype videos of the early shadow system and some interesting directions I took with developing new features so basically when I was first working with these shadows I realized that when there's something that's you know collidable like ground it would cast a shadow and when there was no ground it wouldn't cast a shadow that meant that all of my situations had sort of like an empty ville that you could fall into and that doesn't really challenge the shadow or light mechanic that just challenges avoiding falling through a pit so how am I going to use that in an interesting way to reinforce the mechanic and so I just took a template freeze I made it semi-transparent and blue and I didn't apply the shadow collision to it and that's it and all of a sudden now I invented a new type of freeze that we had to add to the kit that was integral

to making this shadow gameplay work in the final world but before I could convince them I had to convince them so I had to make more prototypes that sort of stress tested or proved that it's really an interesting addition to the Arctic kit the art kit and the tool set as a level designer so I had to make a number of different prototypes that showed the use of the glass in different ways and eventually it got approved so then there are many different uses in the final game this is just one of them that's an interesting combination of the basics entry plus the glass another interesting example is these underwater snakes they are not normally meant to be used this way they didn't have this shadow functionality to begin with they were meant for something else but we thought well if they're already there why don't we just give them the shadow collision property and that's it so I just opened

the script I changed the the material definition and suddenly it was casting shadows and it was a really interesting thing to use for the shadow game play easy peasy and here it's just a simple matter of putting a trigger that pauses and unpause is the light and that was it and it worked it really worked it was very fast and it made it into the final game one thing that was interesting for me though is I made sure to make the flicker when I was doing this because you need to have clear signs and feedback to tell the player okay the lights gonna turn on the lights gonna turn off and it was important even in the prototype to prove that so here we actually had to work on a little thing called the Wii U for a while and we have to make some prototypes for it so we had these elements you could pick up using the Wii U tablet you could drag them around and we had this shadow ding thing so I thought why not parent one to the other and see what happens and this happened and it worked so we

ended up making a different version of the shy of the light casting thing we integrated the components of being able to carry it on the screen and we had a new gameplay element to use in different situations and it was sort of just like a happy experiment I act happy accident actually I just I thought why can't I combine these two things so I tried and they worked and that's really the biggest thing in this engine for me I think is it just like taking two things that don't really belong together and sort of thinking like well why not they could right and this is another example of that we have this thing in the game digging where basically on the touch screen you can touch these cakes and dig tunnels for your friends we wanted to reuse this mechanic for the rest of the the world to make it consistent but not every level needed to have the touchscreen gameplay we didn't want to force that on the player for every level so what we tried to do was separate the functionality and combine it with other things to give the same feeling but in a different way so I use the parenting

system of course that we have we had a separate actor for the digging and allowed us to have flexibility and prototyping because we just said I think that this thing should be able to dig through cakes so you parent the object to it and you see what happens okay that doesn't work you try something else you move on if something starts to work okay cool you move on from there and do a number of different prototypes and eventually integrate the component into the gameplay element so every snake now has this digging component and when you put it in these sort of digging cakes they start to dig tunnels so here for example you've got this lovely dancing skeleton and this is how the digging works basically with the touch screen but what I'm going to grab over here is the actor that's separate and you can see it has the exact same functionality but it's not attached to the touchscreen gameplay so what I'm going to do is use our parenting system and you can actually see all the bones in the

character and you can parent anything to any of the bones so I could parent something too of the guy and it'll be attached to his hand and when he moves it up and down it'll move with him it's really really strangely powerful and now you can see that as he walks around he's destroying the cake and that's what I did with the snakes this takes were more interesting than this guy so I stuck with the snakes but there are some other interesting cases too for example there's this really big luchador and one thing that's interesting about this case is the original version of this there was a flat ground he jumped he landed and he sat there and you had to kill him but because of his animation and the way that he was it just felt a bit weird so what they did was the level designer just dropped some digging at the bottom and put the digging component into the guy all of a sudden he jumps off falls into the abyss it solves one problem and it creates an extra danger behind the player which pushes him forward

problem solved I just a matter of replacing one little functionality for another and seeing what happens and here's this next I was talking about it's really interesting because you can make all kinds of interesting tunnels configurations with different snakes digging through and visually it's really interesting - it's very different from something that you would normally get with a player controlling it usually they'll just wipe the screen clean done but here you've actually got a nice tailored level design that's created and generated by these enemies in the game this is an example that I'm particularly proud of because I did it it was basically just we were between projects after Rayman Origins was finished and we were playing around some adversarial gameplay you know it's a co-op game but we like to smack each other can we make that a thing we wanted slight randomness because otherwise one guy dominates and everybody feels bad and we don't want that we're a family game everyone should feel good and it needed to be a

convincing prototype because if it's not a convincing prototype you're not convincing anyone are you so ultimately this is what it is it's just a soccer ball and you can kick it around it's fun in for players people can smack each other it's two versus two three versus one all that kind of good stuff and what it basically started was was I just tossed a projectile from an enemy in Rayman Origins he threw boulders and they bounced once or twice the ironic thing is that after they bounced once or twice they died it exploded so I had one shot to get it into a goal and otherwise I'd had to restart the map so it wasn't really easy to prototype that one but it proved a point it felt good to try to kick a ball into a goal so then I made a trigger that detected the ball exclusively and when I triggered something for example I had numbers and I would just say when the ball enters shift the numbers up so it says one now it's 2 now it's 3 it was like a really rudimentary messy prototype but on the outside when you're playing it you don't feel it you don't know that it's a messy

prototype it feels like what it's supposed to be so then we developed it that was the final version this is the first version this is actually something internal we did this for our Rayman Origins release party actually in the final version you've got some nice little enemies dancing in the foreground when you kick the ball in for this version we have our lovely creative directors face but it was good fun and actually what's interesting to note here we had moving goals multi-ball at the top bouncing surfaces for the floor and we moved all of them for the final version it was fun we could have done more but we felt that the simple version was clear enough and we had more things to focus on but it was a lot of fun and the reason that we kept it was because we we made the prototype we brought it to our Rayman Origins release party and everybody started playing it and they couldn't stop for the entire evening can you imagine an evening you're supposed to be drinking and celebrating that you're done with that freaking game you not to look at it anymore yeah party hearty and

everyone wants to play the prototype it just doesn't make any sense but it told us something they knew that it was actually quite special so we pushed it forward and it made it into the final game so through all of this I mean what went right obviously a lot like right right although I wouldn't be here talking about it like this or I'm just hopelessly optimistic well there's the rapid iterative failure thing we did fail a lot we failed hard sometimes less hard other times but we failed consistently and that was by design we knew that the better off we were we were better off by building a lot of things throwing them away building more throwing them away then putting all of our eggs into one basket and realizing when it was too late that it wasn't working our actors our templates the free system which I'm sure impressed a number of you all of these different systems are obviously going right because they allow us to do our job so fast unimaginable until you try it and we're a small team for ease

of communication because with the tool like this as well you don't need so many people you know you build systems like it's procedural which is everyone's doing these days kind of a hot topic it is relatively procedural the way that it's set up these freezes and everything it saves a lot of people a lot of time and also having a small team is great because you can actually go up to somebody who's right next to your programmer and smack him for something or tell them this or that there's not this sort of like hierarchy or you have to report to is lead and this and that we're a small team we work together and little strike team sometimes other times we're in our big level design group programming group but we're all in the same little villa in France it doesn't really make a difference when I have to go upstairs the water cooler and talk to somebody or two down the hall and of course there's the simplicity of workflow and the flexible systems which I've hopefully demonstrated to you guys already but they're there and they work very very well of course everyone makes mistakes

when you say that failure is sort of by designed in your workflow you're gonna fail a lot but we didn't really do design documentation you can imagine that when you're working extremely quickly when do you find the time to write things down and when you do write things down aren't they obsolete that's pretty much the way that it went for us but the bad part about this is the fact that when someone new joins a team they don't really have a resource to go to they can't just like Oh read the Bible and then there you go you go through it and you figure out everything and then you start working we have a help menu in the editor which is good it's a good base but for example the design of the game no no we're not ready yet we're still experimenting we have two months left we're still experimenting we'll see sometimes we had a lack of tools for specific tasks I mean we have some great systems in there but there are some other processes that are a little bit more tedious less flexible basically and sometimes we could have communicated better on what we needed so we could actually work more efficiently and there

was a lack of communication of engine workflow issues because the thing is when you're used to working so fast and trying to pump things out very quickly when you realize that you have to do a very tedious work around for something just to get it to work you take the extra ten seconds to do the work around rather than telling somebody the parenting system isn't working you have to parent three times instead of one but it's like oh whatever I'm sure they know and I work with it and then like six months later I hope someone knows well you keep working with it instead of just going and if I just went to the guy and told him it's not working it would be fixed in the next build but it's just this mentality that we get into of just deal with it just push forward you're gonna fail anyway so you might as well just keep going and when you hit something that's good then you improve it but the engine is a different story and sometimes we have to realize that and there of course was sometimes a lack of communication of new features and additions because we worked so informally sometimes going straight to a programmer and asking for a new feature that meant that there were times that a

new feature was created and one level designer knew about it and he didn't tell the rest of us and we lost a lot of time and then some guy comes up to me a year-and-a-half later and says that we had the thing all that time and I've been doing it the wrong way people get annoyed and it's definitely something we need to work on but it's cool when you get new features just tell people when they come out in the beginning so it's not just Rayman in the VR framework it started with Raymond but it sort of evolved since then we actually do mobile games in newbie art as well Rayman Fiesta run Rayman Jungle Run actually can I have a show of hands who's played one of the new Rayman games from origins to legends to mobile not bad good job guys [Music] thanks for helping me survive eat everyday so we do mobile but we're also expert well not me personally but

Ubisoft is experimenting with other genres in the UV art of course so in Montreal they're working on a Japanese RPG child of lights using the same framework is still a 2-d game is still using a lot of the similar systems but they're adding some really cool features that we didn't necessarily add for our projects but it's cool for us to see how other teams adapt and impact the development of the engine and of course we also have another one valiant hearts which is a world war 1 adventure game that's being developed actually by the guys who did Rayman legends I not on the team personally I have other things to do but a lot of my friends are working on it you should buy it it's gonna be great and it's it's a really cool game it's it's different to have something that's a bit slower paced less action II more emotional and everything and having an adventure game in an engine that was built originally for a platformer but it's working it's an interesting take on the genre and we're trying to push forward and see what we can do with the engine from there I actually write

sometimes for the Ubisoft VB blog I've written a couple columns about working in the engine working on the games and working with the artists and trying not to fight with them and all that kind of stuff and it could be quite interesting for you to check in especially if you found this presentation interesting despite the blackout you might want to check this out and have some further reading just to give me some hits anyway you know so if I'm gonna give you a couple things to take away apart from the fact that have a good engine that empowers your artists and designers which is a given fail often don't be scared of it you know plan for it or don't plan I mean to plan failure let's just build stuff and see what happens try to make the most efficient fast process that you can for getting to that sort of proof of concept stage get there see if it's working or move on but that's it just push to fail often and you'll eventually stumble on the right thing embrace artistic constraints a lot

of people think all constraints are there so constraining but they're not they they really push you to be more creative and work outside of your comfort zone there been so many times that I'm given a new art kit in a new world and I'm like what am I supposed to do with this come on guys and then I think oh no but actually I can do something I've never done before and that's the point if they just gave us the same kit re skinned every time we'd be making the same levels every time we wouldn't move anywhere and now we have worlds that are very unique different gameplay elements different direction and it's because of these artistic constraints that we sort of have to live with and embrace and my personal model is waste your home time not someone else's I'm very happy to spend a lot of time making my prototypes trying to prove people that it's gonna work and I don't mind if I have to throw it away because I know I'm not impacting the rest of the production I can make up for my own lost time but I don't want to feel guilty for somebody else's and so

you have to find a way to really clearly express yourself in your idea so that everybody understands and knows where you want to go with it and we can make it together so that's it check us out at the UB lounge upstairs we're doing demos of the engine and I hope you appreciated it and thanks for being here [Music] as far as I know I think there's some time for questions if anyone has it you have to step to the mic and otherwise I'll be in the wrap-up room after the question time can you talk about the process for choosing the concepts and our tickets for the levels and who's who's involved in those discussions yes so the people that choose the art kits that we use basically the artists through the concept of course and there's a always a higher level artist or creative person that oversees it and stuff like that and they just sort of decide okay this one's interesting not clean enough they have to be able to

visualize the gameplay as well if they can't really see how this concept is gonna translate to the level or be clear enough then they go back to the drawing board and you know revamp it a little bit or throw it away and try something new there are worlds we didn't include in the game because they didn't work as well as others and we didn't have the time to make them work but so basically just comes down to people working hard and somebody above them telling them it's good enough for it's not business as usual you know sure hey so you talked a lot about the art impacting the design and I was wondering if there are any instances where are impacted things on the programming side yeah I mean there's I'm guessing there's a lot of times when especially in a game like this that's cartoony that the animations sort of guide the way that you program certain functionality right for example the character controls of Rayman and the other main characters of the game they're very very based on the animations the cartoony sort of ease-in ease-out floppy animations that's why it

feels that way but it works because it's sort of linked to the animation that was originally given to the programmer to work with so yes of course in cases like that it comes down to taking the artwork understanding the intention of it and programming the functionality to make sure it feels the way it's supposed to of course so anything like new features that were added well of course I mean the the artists they did tests for the lighting system the new 2d lighting system basically they did them in things like 3ds Max and Maya just to sort of get a feel for what kind of system they wanted and then they told the programmers this is what we want this is how wanted at least to work this way and they said okay boss and they did it and it turned out cool in the end but it's because they had a clear direction they knew what they wanted and they had to convey it to the programmers okay thanks yeah hi I was really curious how did you get your job at Ubisoft magic no no I was I was a I was a student three years

ago and I made a portfolio for level design because I wanted to one of my teachers used to work at Ubisoft and he was like hey looks pretty good and so he sort of like sent it under the table I was actually busy an entire weekend putting together like stacks of applications for other companies like I'm gonna do this and then on Monday it's like ready to send group want to work at Ubisoft well whatever so I went and worked at Ubisoft and then I stuck around and I have a contract now so I've been there for three years but it was really that but I guess they liked my portfolio or something I think it - you know my question concerns I know you talked a lot about just creating levels just by doing and prototyping and testing and failing whenever you got those gem moments and you had success because the iteration process is so fascinating awesome how did you know when to like just stop you know did you ever get sucked into just

being in that level and enjoying it and like how would you move past that to the next one how did you know to move on well a game has never done everyone knows that so it can be hard sometimes to say done and move on there have been times that I actually cut out an entire portion of my level after it's been approved after it's in beta and rebuild it sort of like a ninja thing and somebody noticed isn't they asked what the hell do and like no trust me it's better and I'll make sure it's clean don't worry and it usually is but honestly it's just a matter of saying like did I fulfill what I needed to they want a boss level with three arenas and creative ideas I built all three of them I could iterate on top of it more and all this kind of stuff but it's like scope wise it's there functionality wise it's there it's cool people appreciate it it works in the play test let's leave it unless I get tired and I fix later thank you yep hello this might be a bit big question for here but if

there's some information on the block or could be I'd be very interested so basically you talk about these freezes for creating for example ground or pipes similar and assume creating a completely new type of freeze requires some programming but I'm curious about the process if you have a freeze already and you want to adapt it to a new environment with completing your art what are the tools for doing that it's a good question as far as I know what it comes down to is there's like it's like an atlas and you can see all of the pieces like the top the side and all that stuff this things at tile and it's a matter of making a new atlas and then using a tool to sort of specify this is the top this is the side this is the fill and things like that the technical actually at the UB lounge there's a technical artist who uses UB art at the moment and he can actually open everything and show you all those tools and how it works so you can get some real details because I'm just a level designer but it's more or less like that you I think yeah I had a question about Rayman legends about the musical levels

so you mentioned working with the artists to create a look at the canoe tell me who was there a different process making a musical I was considering they're very synced to the the beats that's an interesting one actually it was it was just sort of like a happy accident that we started making musical levels we were just making some challenges and we made a video of it to show internally and put some really cool rock music synchronized with the action in the video and after they showed that it was kind of like can we play with the music I mean is this real or is this just a video and it's like well no but maybe we can make it and so then they started building a system for synchronizing the gameplay to the music the the composer Christoph Farrell he made the original version of the song and then the level designer built a level based on it and then it was a process of back-and-forth between the two of them to make the final product and then of course the animators came in and did all the fun animations background art things like that yep how much do you think the editor evolved

during this dream and I mean it sounds like you create the levels the graphics got created graphics but they were edit or just was there somehow how many new features came up all had to be done by the programmers lots changed I mean if you look at for example just a touchscreen functionality alone or the three-dimensional entities in the game they just weren't there and those are a big chunk of what makes rayman legends' different from a man origins right we also had a lighting system that wasn't there at all it changes everything about the way the artists decorate their levels if you actually press a button to turn off the lights the levels look really bad like in war remand or adjourns we worked with it and we made the levels look like they were maybe lit but it was a different style but here you turn it off wolf and so it's important that it was developed properly and so there are big big features like that that really impacted the way that we started to produce things and there were just so many countless little features I can't even remember because I'm just used to the current version but

it was a very different tool when we did Rayman Origins and it's improved a lot since you you mentioned working on a small team how small was your team I think at its peak it was about maybe a little bit over a hundred people I'm not sure but usually we were about 6070 then we ramped up a little bit more near the end to get things done yeah anyone else you how did the decisions get made about what levels are selected and what aren't lead level designer creative director okay I mean I'm in era three much things that you thought were particularly a good level that they weren't into and you had to make your case for more or of course yeah but I mean that's what it comes down to right it's about convincing people that's what I've been talking about with the prototyping process and everything it needs to be clear to convince somebody that this is

a good idea and I've seen a lot of people that go in with a half-baked proposal they say oh look it's a red square on the floor it's a button it's like oh okay I guess I trust you and they jump and it doesn't do anything except to open a door and stuff it doesn't make sense but when you make it look like what it's supposed to be feel like what it's supposed to be and then you show it in a multitude different situations then it's very clear for them whether or not they just hate the idea or they can't deny that it's actually quite cool so when it comes to reviewing levels and things like that it's just that the level design lead come sits down with you and says this is good this is not try to go more in this direction and if you really disagree you can take one more chance to try to prove it to him but honestly that's the part about failing and moving on nothing is sacred just trash it and move on you can make something better in five minutes anyway thank you I'm a little curious about the development of the engine itself or the framework what led up to it and how long was it in development begin before it

was used and when you add key features and stuff how long does that typically take well the engine was originally created by the creative director Michel Ancel and a small team of I believe five senior programmers the initial version of course I mean then we ramped up to a more a few more engineers as the process went on we were building the game itself but it was a small team developing the tools for that and then the team started to ramp up maybe a year later I believe and we started to actually go into production on Rayman Origins and it's just been in constant update status ever since but when it came to something like the lighting engine the new update for the lighting system it took quite a while like I believe if I remember correctly maybe a month or two to wait to get that thing actually working but it's a heavier system so it takes more time but some engine improvements they don't take any time at all they take like maybe one to three days tops depending on what it is thank you

yep I think I can't take any more because they held up a yellow sign or something so thanks for being here I really appreciate it guys [Applause]