Environment Design as Spatial Cinematography: Theory and Practice

[Music] thank you all very much for coming to my talk on as you can see environment design as spatial cinematography I have been asked to ask you to turn your cellphone's off and to fill in the forms at the end and yeah so before I begin the talk properly some quick introductions so my name is Maryam Ballard I am the director for visual development at Rockstar North I hold I have a background in architecture and film as well as video games so my architecture background I hold a first class honours degree in architecture from Oakland University in New Zealand so I'm a Kiwi if you can't tell by my accent with film the highlight of my film career would have been as a set designer on King Kong so that was a Peter Jackson film filmed in New Zealand and with video games the highlight of my video game Korea is most definitely my current job at Rockstar however I haven't just worked on big-budget and Triple A when I started in the film industry that was on low-budget things so short films music videos low-budget TV and low-budget features with video games I also started out at an indie studio and I returned to indie work in the middle of my career as well so this talk is not just aimed at the big budget and Triple A it should be useful to anybody who's using 3d environments and for their video games no matter what your budget level so my team so I need a team of visual development artists they come from a similar background so I have on my team architects interior designers people from film as well and also people from the 3d environment art and the kind of work that we do on my team is this kind of stuff that you see here so we do a mixture of these 3d models that you see which internally we call previous models so they're sketch models that are I suppose the first exploration of an environment we also do reference boards color palette passes and some 2d artwork as well we don't tend to work much in 2d plans we might do an initial schematic of a space but generally most stuff is worked out in 3d so I think I'll help for this talk if I also explain a little bit about how things are structured at Rockstar in terms of environment design because obviously it's different at every studio so for us the responsibility of the first path of an environment layout is actually the art department not game design so that we do it like that just because that's what suits us and that's what suits our games you know it's not you know everybody has different needs how my own team fits into that so anything that is particularly difficult or problematic and goes through the visual development artists so on exteriors that means things like there are specific problem or specific issues which are a bit more tricky and maybe specific areas as well those will go to the visual development artist on exteriors for interiors every single interior gets a visual development pass so introductions over the main talk environment design and spatial cinematography what do I mean by this so in most video games the player is controlling the camera not the developer however we can still make cinematic experiences we just need to do this through controlling the environment rather than trying to control the camera so to repeat that because that's the main point of this talk even though the player is controlling the camera we can still create cinematic experiences we just need to make those experiences by controlling the environment that is quite a big question how do we do this and like any big question it's best broken into smaller questions so how do we control the 2d screen how do we control movement and how do we control time now in film this would obviously be done either with camera or with editing all of these things but again we need to figure out how to do it with the environment so starting with the 2d screen so we need to compose the 3d for how it's going to look Toodee this is the one thing that differentiates us from other media so for painting they're obviously painting directly on the final 2d product for film and for cinematography for photography and cinematography they are obviously using the camera to control what's on screen so as we are composing the 3d for how it's going to look in 2d this is one of the things that we use our previous models for so I'm not sure if I mentioned before the sketch models that we use as part of our initial pass internally we call them previous models I haven't heard anyone else use that term but if I say previs that's what I'm talking about so we use our previous models to work out what we're getting on screen we also use them to solve lots of other problems as well I forgot to mention before that one of the goals of my department of the visual development artist is to make sure things work for everybody so we are not just solving aesthetic only problems we are making sure that our designs are going to work for the you know for game design for environment art for the art direction for narration for everybody this is not aesthetics over functionality this is aesthetics and functionality so as I was saying one of the things that we use the previous four alongside you know having giving game designs something to test with and environment artists something to feedback on we also use it to compose what's on screen so this here is the previous that we did for the undergrad the lab area and the Doomsday heist dlc so everything I'm going to be showing you in this talk is all from GTA 5 online DLC just so you know so that view was tweaked for the play it was designed around the player view that by the way is from this angle here so we are designing around choke points and through routes because we know where the player is going to be in those situations now we that this does mean that we can't get every single angle absolutely perfect but this isn't that different to film either in film you might start shopping that's beautifully composed you might end up with something that's beautifully composed but not every single frame in between that is going to be a perfect composition it's the same for us we know where the player is going to be at certain moments and where they're going to be looking at certain moments it doesn't matter or or it can't be helped if some angles aren't perfect and that again was that view there this by the way is how the previous relates to the final product so as you can see the when the lighting team came in the final composition isn't exactly the same as it started in the previs but they have had the a good structure to work with they're not trying to affect something that's not working they've got something that they can enhance and add to all of the screen shots I'm going to show you by the way and player videos are all of just real players playing the game they are not set up by marketing or as a developer run through as just normal players playing normally okay and this is the same environment but just from a different angle so this is one of the things that we're doing with our previous so I'll explain those third markers later but what what we are doing is we are tweaking the 3d for how it looks and 2d so we are moving around the space and we are designing by moving through it so we'll start in max which is the modeling program we use and we all tweak so that we have good compositions on screen and Max will also take into it enter the T speed which is what you're seeing here so you'll notice how a lot of the lines of the building are all lining up with those third markers and again not every composition is perfect like this angle isn't quite as good as the other ones but again when we swing around it's nice again yeah and again you'll notice that the proportions of the building even the heights between the distance from floor to ceiling are all in proportion to those third markers so this is how it came out and the final product so the composition has actually been enhanced by the extra planters we added for game design and as you can see the composition from the previous has basically stood up the one but that doesn't work is where we're going right now and that was a little bit of an oversight on my part I didn't you have realized how much they were going to use these stairs so these stairs aren't composed nicely for for the 2d but you can see that it gets nice again when we get up to the top okay so that what that bit there was the only bit that we that makes us different to other media in this respect that we are composing the 3d for how it looks in 2d beyond that when it comes to controlling the 2d screen it is the same advice that you get for other for any other media so it's the kind of advice that you would get for painting for photography cinematography and so on so one of that so one piece of advice you get in those other media is to have nice to decompositions so one aspect of that is these third markers that I was that was on that were on the videos so what that is is there's a rule of thumb and is definitely a rule of thumb not an absolute rule there is a rule of thumb about dividing your screen up into thirds and having the main compositional weight of the image sitting on those third markers so you'll notice how in the first image that the weight of the composition is right at the edge of the screen and not in the center however in the final image I've pulled the walls in closer and the disk in closer and the weight of the architecture the the composition of the 2d image is sitting closer to the certain markers and it's a much nicer composition but not that much was actually changed in the model so this is something that we're doing again as we're moving through the space we are tweaking to get nice on screen compositions and again we're not only concerned with the on screen compositions we're also adjusting the space for other needs as well it's just this as the topic of this talk so that again was the view as you enter the space so again we're concerning ourselves with choke points and three roots so another useful compositional technique is this as a concept of tangents specifically avoiding tangents so you will notice in the first image that the line of the desk is lining up with the wall behind it this yes oh it's doing it in 2d from this very specific angle this makes the space incredibly hard to read it is very hard to tell in that first image how the desk relates to the wall behind it spatially however in the second image I've just tweaked the desk a little bit shorter and suddenly we can read the space much better we can see that the desk is sitting in front of the wall so and so this is what is called tangent so that tangents are when unrelated objects are lining up on a 2d image and again this is something that needs to be avoided because it's hard to read the 3d space they are very easy to fix the tricky thing with them is actually finding them in the first place another thing we want to consider is the angle of the cameras that are going to be used in the space so this here is the same space but in different our setups so the first setup the player is just exploring the space and the second setup they are in combat mode as you can see the camera is very different in the two scenarios so what we want to do as much as possible is to compose the environment so that it looks good in any camera setup that's going to be used in the space now it's not always possible to do this for every camera setup obviously but again this is an ideal to aim for so if you want to know more about compositions on 2d screen this here is a particularly good book I will also by the way be showing a list of all the books I reference right at the end so don't feel like you need to write them down as I go also if you don't want to buy books there's actually a lot of advice on compositions online so you don't need to get it as just a good one one thing that book doesn't mention however is that as tangents but if you search online for avoiding tangents and art you will find a lot of advice on it and for anyone who wants to delve into the science of why tangents are a problem there's a lot of good information in that book there so as well as creating good compositions on screen we also want to create dips on screen so a screen by its nature is 2d it's flat which means that we need to use tips and tricks to trick the eye into seeing dips one of the easiest techniques for doing this is creating layers so by layers I mean having foreground mid-ground and background elements and what happens in the scene is what you see here the foreground elements move a different speed across the screen to the mid-ground and background elements so this creates parallax animation which is obviously what old-school Disney Animation uses it's also an incredibly good way to make for very good bang for buck dynamic environments so another way to create good depth on screen is with repeating elements so if you have identical elements throughout the scene and there are different dips in the scene the player can use the relative sizes of those elements to judge the depth of the space another good way of creating depth is the use of atmospherics and so this is not part of our previous and pre production pass but because it has such a big effect it felt remiss not to mention it and so for anyone who doesn't know what atmospherics are they are when particles in the atmosphere scatter the light and they make distant objects less distinct and you get the kind of effect that you're seeing here so for more on depth on screen this here is a particularly good book so as well as creating good compositions and creating depth on screen we also want to control the shapes that we get on screen so in we can use the shapes on screen to create an emotional experience for the player so in terms of evolution different shapes have different emotional effects on us so our ancestors who like certain shapes that our nature were associated with safe things those ancestors were more likely to survive and have us our ancestors who disliked shapes in nature that were that were associated with dangerous things those ancestors were more likely to survive and have us so we have inherited the likes and dislikes of our ancestors that have that helped them to survive now we can use those shapes and the emotional effect of those shapes to create an emotional experience for the player so circular shapes and nature are associated with things which are safe and friendly so think rolling hills bushes trees that kind of thing so they were used here in this particular one because this was something that the player was going to own and so we wanted a nice welcoming experience for the player so you can see that these shapes are started in the previs and but they are very have been very much enhanced by the environment sorry there have been whoops have some water so the shapes have been very much enhanced by the interior artists who worked on it and the lighting artists as well so jagged shapes in nature these are dangerous things so think broken rocks broken sticks broken bones yet teeth claws that kind of thing so we can use those shapes to create unsettling experiences for the player so it was very appropriate for a combat space where we wanted to create a tent situation it is exacerbated in this situation and by the top-heavy nature of the architecture as well so horizontal and vertical shapes are about power and control and stability so a horizontal line that's the horizon so that's stable and unchanging vertical shapes think trees and cliffs so they are powerful things but not necessarily as unchanging so those shapes very much suited the storytelling of this situation where the player was playing as an office executive and the idea was to give them a storytelling around power and control so the shapes that end up on screen normally come from the previous but not always as you can see here the curving lines on the floor are very much overriding the previs in terms of the final shapes on screen you can also override the previs with things like well textures patterns and the lighting as well so the shadows that you're getting from the lighting can make incredibly strong shapes on screen another thing to be to notice in this particular set of slides is that even though the architecture is very square in the previs the shapes that we're getting on-screen are actually very soft diagonals and the reason for this is obviously the angle that we're looking at the space from now when I talked before about designing around chokepoints and through routes one other thing to consider is that we actually decide where those choke points and through routes are going to be so if we'd wanted to come into the space and make the first impression of it about power and control and stability we would have come in front on to create ship square shapes on screen if however we wanted a more informal introduction to the space and something a little bit more dynamic we can come in at a slight angle like you're seeing here so if you want to know more about shapes on screen this here is a particularly lovely book on the topic so to recap the controlling the 2d screen so we want to compose the 3d for how it's going to look in 2d and that is the one bit that differentiates us from other media everything else is just the advice that you're seeing for painting photography and cinematography so that's things like creating good compositions creating depth and creating nice shapes on screen so movement how do we control movement and fortunately the section is going to be quite difficult but if you'll just bear with me for it the last section gets easy again the reason for this is that controlling movement well the cameras attached to the player so if we want to control whether camera looks and goes we need to control or at least try influence where the player looks and goes so saliency is going to control where the player looks and therefore whether camera looks affordances we'll control where the player goes and therefore where the camera goes before I get into explaining what those terms actually mean are some quick myth-busting first so there is a concept in aesthetics that's been around for hundreds of years that a strong compositional line in a painting will lead the eye through the painting so these are called leading lines now modern science has shown that this is not actually what's happening what is happening is that our eyes are flicking past the compositional line and they're getting caught in it like a fish in a net we're not there is no directional element to it now if our eyes are not following those composition lines then we are then we are not obviously physically following them either there's no directional aspect to this these lines are still incredibly good for capturing the players attention and I also don't think we should throw out the concept of leading lines because it because it is very useful in terms of understanding the structure of a 2d image however it is an abstraction so it is a specialization of an abstract concept so as humans we quite often make do this we make spatialization zuv abstract concepts so for example I might refer to somebody as being in front of me or above me that again you know those are ways of specializing something abstract that is all that's happening here by talking about leading lines we are specializing something abstract and so it is still helpful to use but we just need to remember that it's not literal that the player is not look their players eyes are not literally travelling in a line and therefore we can't lead the player with a strong compositional line and the thing that doesn't directly lead the player and in terms of like direct one-to-one although it has an influence is lighting so what you're looking at here is an actor is an academic study on visual attention and 3d games now the at this moment in the study the players were asked to exit the space the correct exit was behind the lit better wall that you see there the players had a lot of trouble finding their exit and they tried things like doors and fireplaces in spite of the fact that they were in darkness now this is because we don't automatically walk towards the light if you think again about our evolutionary heritage most of our evolving happened with sunlight if you had a break in the clouds and a bit of light spilling through the clouds onto the ground there's absolutely no reason that our ancestors should be walking towards that bit of ground so we have not evolved to work in that way what like can do however as it can help us see things that we might not otherwise see and it can give spatial information that we might not otherwise have so it is still very useful in leading the player but it's just not a direct one-to-one so it's something else in this here is I just want to give it a bit of nuance on this one so there is a concept called wienies which a lot of you will be familiar with and because every time I've given a test of this talk people have asked me to spell out that's W EE in IES it is the German word for sausage so the term was coined by Walt Disney and the idea is that our weenie leads the guests in the same way that a sausage leads a dog weenies absolutely work like there's no doubt about it but the reason they work is because of the mystery element so weenies have two aspects to them they are a landmark and they have mystery a landmark by itself is incredibly useful and incredibly important in terms of the wayfinding for the player but if you want to actually lead the player to the landmark you need to add mystery I'm going to be talking about mystery later so we will come back to this one so myth busting over how do we lead the play and therefore let the camera so salience II will control where the player looks and therefore where the camera looks affordances will control where the player goes and therefore where the camera goes so that's starting with saliency so saliency is a scientific term which means a tension so something that is highly salient as something that is a tension grabbing there are two types of salience II we have bottom-up and top-down so bottom-up salience II is things is all about contrast so it's contrast of color its contrast of size contrast of shape contrast of orientation contrast of texture and that's the bottom left-hand corner that one's a bit more subtle and contrast of movement top-down saliency on the other hand it's about things which are meaningful so human faces are always highly salient to us because they are meaningful to us humans in general are always highly salient because again they are meaningful and things which are associated with the current task or goals so if I am playing a ball game then a ball shape is going to be very salient to me if I'm getting ready for work then shoe shapes will be salient if I'm thirsty then shapes associated with drinks will be salient and of course there are some things which are always salient to us because they affect things like safety so the final salient seaso where our attention is going is a mix of both bottom-up and top-down how much comes from bottom up and how much comes from top down depends on what mode that we're in so the more we are in a task orientated or goal orientated mode the more top-down saliency is dominating now this is quite important for video games because obviously our players are usually in a very task orientated or goal orientated mode this means that compared to other media the top-down saliency is going to be much stronger so in film where people are just sitting there and watching a film the bottom-up saliency will be stronger because people are just relaxed in a relaxed mode for our games the top-down saliency will be stronger so what you're looking at right here this is an eye tracking heat map they were that was done for an architectural study so they were comparing to photoshopped versions of the same image now an eye tracking heat map for our intents and purposes is an identical to a salient see map where people are looking as where their attention is now this is not where they are consciously looking this is where their eyes are subconsciously flicking to now what is interesting here is that the people looking at these images were would not have been a very task orientated or goal orientated modes they were just asked to look at a picture however the top-down saliency is incredibly strong if you look at the image at the eaves on the top of the image try get that one as well and see where there contrasted against the sky in the background that is the strongest bit of contrast in the image so if this was about bottom-up saliency everybody should be looking at the eaves of the building but nobody's looking at the eaves of the building they're looking at windows and doors and the signs above the door that is because these things are meaningful so of so with a window we can see through it is also associated with people the same and doors as well are associated with people and the sign with information okay so if you want to know more about salience E and attention unfortunately I don't have any easy reference with this one so this here is a lecture which I'm pretty sure is aimed at first-year neuroscience students so if you're anything like me you will not recognize a lot of the terms in the video but a few lesson throughs you do start to pick things up and this is a similar one as well again I'm pretty sure it's a first-year lecture aimed at euro science students and this was the academic study I mentioned on visual attention and video games have not seen very much on this topic I mean it might be out there I might have just missed it but this is all I've personally found so if salience E or attention is controlling where the player and therefore the camera looks affordances are controlling whether player and therefore the camera goes so for anybody who doesn't know what affordances are a quick recap so affordances the action that an object allows so a chair afford sitting a car affords driving a door affords opening floor affords walking window affords looking through you get the idea now there is such a thing as miss perceived affordances this is when something looks like it does something but it actually doesn't so you might have a chair in your videogame that looks like it affords sitting but you can't sit on it and so obviously in videogames this happens quite a lot because we pull shapes from the real world but we don't copy all the affordances of the real world so one solution to this is what I call sik called signifiers so signifiers are arbitrary signals that the player needs to learn and that communicate affordances so a traffic light and the real world affords the fact that you can go in videogames we often use you know yellow lines and things like and strong colors to highlight things which are interactable and signs and symbols are of course of sacrifice as well now one thing to be aware of with signifiers is that they do to a certain extent break the fourth wall because it means we are talking directly to the player now this suit some games are not others this is entirely context dependent but it is something to be aware of and there is quite a bit online about video games and affordances so these are good places to start and that there was the original book on the topic this was where product design not games but the same principles apply so recap over how do affordances lead the player so we're going to start with the motor with the simplest type of affordances because there are many types you will see that the player is walking straight to the vending machine it's pretty simple so this is object affordances and the player will simply move towards an object to use it it's really is that simple now it doesn't just have to be something that's very very real-world affordance like this it can also be affordances like a plank affords walking a rope affords climbing a window of woods climbing through a fence affords breaking these are things which the player will move to Ward's if it is the only affordance and the scene as well then the player is definitely very likely to use it to go there but of course we need to make sure that we communicate that affordance to the player and that the player can see it and understand it so as well as simple object affordances we also have surface affordances so when I mentioned to you before the floor affords walking this is a surface affordance this is also something that people will walk towards just like the object so what you're looking at here this is two different academic studies the one on your right the white one that was a real-world study of the tape gallery where they were tracking human movement through the tape gallery the study on your left on the black image that was a digital recreation of that study so what the people were trying to do in the digital recreation as they were trying to prove this idea that surface affordances attract people that we walk towards a walkable surface so what they did is they set up some very simple rules for the box in the study those bots were told that every so often they would stop and change direction the direction they were most likely to walk towards was the longest sight line now in the context of the study and the Tate Gallery a long sight line correlates with a larger walkable surface area or the greater affordance of a walkable surface area so as you can if you look closely at the way at the at the bits where they're walking in the centre so you'll notice that the shapes made by the real humans and the shapes made by the bots are very similar then neither of them are walking right at the edge of the space and we're also not getting a very narrow channel down the center we're getting very similar shape of movement this shows that we as humans do walk towards a walkable surface area that the bots did that the rules the bots had dead correlate with human movement so it is worth remembering that in the context of the state gallery study that there was not much else going on in terms of affordances obviously when noir most real environments have more types of affordances which would complicate this so as well as object affordances and surface affordances we also have complex affordances and again these are attracting the player were concerned with these attracting the player because the cameras attached to the player yes so as I saying so we have complex affordances here so I sent out a tweet with this particular view asking people if this was a video game where would you go I had just been reading the Tate Gallery study at the time so in my head everyone was going off to the right towards the greatest walkable surface affordance the responses I got her I'm sure what every single one of you are thinking right now if this is a video game I'm going up the alleyway to the side yes so after a bit of rethinking or recalibrating on that one what I eventually figured out was what is going on here is that we have created a complex affordance that we have taught the player the signifier of this is a small alleyway or a side path and the affordance is some kind of gameplay reward so that might be a treasure chest or it might be a way to flank enemies or some new areas to discover so yes in a video game the greatest affordance with this is the small alleyway or safe path so people are going in that direction in real life it is still the road to the right because we don't have treasure chests in real life it's a shame so as well as object affordances surface affordances and complex affordances we also have spatial affordances so there is a concept in architecture called prospect and refuge prospect is a view and refuge is safety so this particular image I would describe it as being very strong prospect but maybe medium refuge so this here this is also prospect and refuge the red the prospect isn't quite as strong there is a you but it's not as epic as the previous one but the refuge is much stronger this space feels much safer and much more protected this is also both prospect and refuge but from where we are further back of it the prospect is quite low there's not that much we can see but the refuge is incredibly strong obviously for the soldier who is right in the window they have both very strong prospect and very strong refuge okay so there are two types of prospect and refuge in the architectural literature there is primary prospect and refuge and secondary prospect to refuge so this here is secondary prospect to refuge and I'm just going to jump back some slides because I shouldn't have moved quite yet these here are all primary prospector refuge so primary prospect and refuge is when we have a view where we are right now and we have safety where we are right now so these are all primary prospect and refuge secondary prospect and refuge is when we can see a spot in front of us that would afford prospect or afford refuge but we don't have it right now so we know that if we went up to that castle we would have a view from that castle and we would have safety in that castle that is secondary prospector refuge as humans we like prospector refuge and so this is something that we will walk towards now this by the way is again about our evolutionary heritage the reason we like prospector refuge is because this kept our ancestors safe the ancestors who like prospect spaces and refuge spaces were the ancestors who are more likely to survive and have us so this has a very interesting effect on the movement in a space so you can see this particular interior here there is very weak or there's no primary prospector refuge but there is a lot of very small implied secondary prospector refuge the player is constantly being promised there might be a view around the corner maybe the next pillar is going to often going to offer say 50 what this does is make the player move so again we're concerned with the player moving because the cameras attached to our player so you will get a lot of very fast fluid movement through the space this is in contrast to very strong primary prospector refuge so here we have a very strong view and a very strong safety and please ignore that plant by the way this was the best example I could find just wish I didn't have the plant but yeah you can see that they're not moving like they have very strong primary prospect and refuge there is no promise of secondary prospect or refuge anywhere there's nowhere that it looks like you would have a better view if you move there and nowhere that it looks like you would have better safety if you moved there so the player is just not moving when they do eventually move like they have now they're moving slower now this is neither good nor bad this is simply about controlling the type of player movement and therefore camera movement that you're after another thing to be aware of between the two is actually the amount of walkable surface area in view and this one here where we have the much faster more fluid camera movement we are we have a lot more walkable surface area and view as compared to this one here where there is a very little walkable surface area in view so the difference between the movement in these two spaces is both the prospect and refuge as well as the walkable surface area so if you want to know more about prospector refuge this here is the book to go to melissa is an architecture book so and as I said before this is about our evolutionary heritage so this is not just tactics for shooter games and this is how we have evolved so this these principles can be used in nonviolent games and they can also be used in non-violent moments of action games so as I as we talked about before mystery also tracks the player how this is working in terms of affordances is that the option is that the affordance is that there is information for us to learn as humans we very much learning information so what is going on here is that the castle in this case is promising us that if we move closer there will be more for us to learn now how it is doing it in this particular situation is with the very interesting profile and with the extra detail on the architecture that tells us that there are stuff to discover if we get closer the secondary prospect that I was talking about before and that is also about mystery if we move closer to that spot we will get a view and of course a view is information so as well as simple object affordances surface affordances complex affordances and special affordances we also have wayfinding elements so a quick recap here and for anyone who doesn't know so there are five types of wayfinding elements we have landmarks these are individual objects which stand out so think a church spire or a mountain we have districts districts are areas with their own flavor so think Chinatown the wharf central business district that kind of thing linear elements are or paths re linear elements which afford movement so think of path obviously but also a road if you're in a boat then a river would be a path if you're in a car than a motorway so edges are linear elements which prevent movement so think of a wharf also think again of the river but this time if you're traveling in in the car and think again of a motorway but this time if you're traveling on foot so these are very context dependent in terms of how you're traveling and then we have nodes nodes are moments of decision or choice so think of an intersection a town square and also think of things like a airport or a train station so the this as I mentioned before wayfinding elements don't directly lead the player but they do affect the mental map the players have and that mental map affair where the players will go so if to give you an example if you were to give someone directions somewhere pre Google map days you may you might say something like this follow this road till you get to the tree at the tree turn left once you cross the river it's the red door on your right there I just used wayfinding elements so the road that we were on was a path the tree was a landmark the river that we crossed was an edge and the red door was another landmark at a different scale so what we're doing is we are structuring our understanding of a space around these different elements and it is our understanding of a space that will affect how we move through a space now this is absolutely something that could easily be a whole talkin its own right so I'm going to leave it there but if you want to know more about wayfinding elements and this here is the book that those originally come from and there's quite a lot of interesting modern science on how we do a spatial mapping so that there is a particularly good article and there was also a talk on Monday which unfortunately you've missed if you didn't go to it but I'm sure it'll be on the vault but yeah that was also a good talk on spatial mapping so to recap the movement section so salience he will control salience your attention will control where the player looks and therefore where the camera looks we had two types of salience II we had bottom-up and we had top-down and and again the top-down saliency is stronger when people are in a task orientated or goal or goal orientated mode affordances are controlling whether where the player goes and therefore where the camera goes so we had an simple object affordances surface affordances complex affordances spatial affordances and wayfinding elements so moving on to the final section now time so time in film would normally be controlled with the camera always editing for us we obviously aren't directly controlling the camera apart from what control we what guidance we can give the player and we don't have editing outside of cutscenes luckily for us spaces experienced overtime so when I take in this room I don't take it all in in one instant in one moment I take it in because my eye is traveling over it or I'm physically walking through it or in a video game my avatar might be walking through it this means that what is laid out over space is actually experienced over time so we can set things up in space that we want to be experienced at a time one major element of this is repetition and rhythm so in film they would often do editing and to get different types of rhythm so you might have much faster editing to create an intense moment in the film so we can do so we can set up rhythm with architectural repetition so repeating elements architectural II will be experienced as a rhythm over time so if you look at these columns here you don't again take them in all in one instant in one moment you take them in over time as your eye travels over them we can change the layout of that of that rhythm of those of the architecture to change the change the type of rhythm we're getting now that is easiest explained on a little mock-up so you can see that the columns on the left are much closer together than the columns on the right and create a much more intense rhythm than the columns on the right so we can change the density over space to affect the intensity over time so for more on the architectural side of this this here is a particularly good book now as with anything that involves intensity this is something that we want to change over time so to be constantly at high intensity can get a little bit overwhelming from the player and they can also become numb to it and to be at a low intensity intensity constantly can become a bit boring so we want to change things up and again changing things up in terms of intensity over time means changing things up in terms of density over space so we don't want to have the same kind of layouts and patterns throughout our spaces we want to vary them so as well as rhythm and repetition we also want to think about transitions so in film this would be how you go from one shot to the next so that might be a cut or it might be a dissolve or a few other things as well but those are the main ones so again for us what is happening over spaces experienced over time so we can do this through our spatial layouts so the classic example here would be a Disneyland so what they do when you go between spaces in Disneyland as they create stylistic transitions so what you're looking at the red arrows are elements of one style it is a little bit hard to see I'm afraid but you can always go to Disneyland see ya so the red elements are the red arrows are pointing to elements of one style the yellow two elements and the other style so they're things like lampposts and handrails and stuff like that that it's pointing to and the orange one doesn't fit either as you can see there is an abrupt transition and they slowly stop one and introduce another so there are moments right in the middle there where you actually get lampposts of one style that are sitting over the top of handrails of another style so they are creating and stylistic dissolves between area and again they're laying about they're laying them out over space but they are experienced over time so we can do this architectural II as well so what you're looking at here so this was done for the after-hours DLC so these two very different architectural styles that you're looking at I don't mean they were supposed to belong to the same space so the first one we the reason we use the post and beam industrial cell architecture was because this was supposed to be a club and this this was the kind of spaces that you saw in those types of clubs however the final one with the curving architectural style so anything water here yeah so the final one so this curving architectural cell the reason for that was that it was a storage space and it would just be incredibly boring to have a post-and-beam architectural style for us for a storage space and we try very much not to do boring so what the architect who worked on this what she did was she created transitional a transitional area between the two so the office area which sat between these two spaces we had well she had created and she had both elements from both styles so there were the curves and there were the posts and beam elements and now suddenly the spaces actually feel like they belong together so for more on different types of ways you can transition architectural II this here is the book to go to this is definitely a book to dip into rather than to try read cover to cover luckily for us with architecture we actually have a lot more options here compared to film that to give you another example if you were to say walk into a building straight off the street so there's just a door directly on the street that would be the equivalent of a cut if you added some elements like a veranda or similar between the door and the street you would create more of a dissolve there are lots and I do mean a lot of different ways we can move between spaces architectural II and yes that's the book to go to so to recap the whole talk now so we were looking at the fact that even though we can't control the camera we can control the environment that the player moves through this means that we can create cinematic experiences by controlling the environment so with the 2d screen we are composing the 3d for how it looks in 2d that was the main thing that was thank you that was the main thing that was differentiating us from other media and beyond that everything else was just the normal advice that you get in other media so it was creating good compositions it was creating depth and was creating nice shape on screen with the 2d with the movement section that was about saliency and affordances so salience he is controlling where the player is looking and therefore where the camera is looking and we had bottom-up sanity and top-down saliency affordances are controlling where the player is going and therefore where the camera is going and for that we had object affordances surface affordances special affordances and we're finding elements with time spaces experienced over time so we can set up spatially what we want to be experienced over time so this was rhythm and repetition was one of the main elements but also controlling our transitions yeah so this is the end of the talk and that's the promise list of references [Applause] and we've got about eight minutes questions thank you thank you excellent talk okay I have a question like if if you apply the same principles to something like the are like a virtual reality you know so would you just throw away the 2d screen yeah I think everyone else would work but definitely the stuff on the 2d screen wouldn't be as relevant oh sorry so did would this apply to virtual reality as well so most of it would apply just not the 2d screen section yeah any other questions okay yeah over everything yes a wonderful talk I was just curious about choke points as we were beginning and I was trying to think of other examples aside from tight hallways and then I saw that last slide of yours about uh transitions and I was curious is it possible to have a sort of overlap in that can contrast between rooms and transitions between those rooms could that also be considered a choke point absolutely a choke point it's right would how do I describe that question so is there an overlap between the transitions between spaces and choke points yeah so absolutely so a choke point is simply a spot where you know the player is going to be and so transitions can help you know where that spot will be you can also do it with props and stuff as well it doesn't just need to be architecture but the answer though yes very well thank you thank you hi hi first great talk also my question is on the footage you showed are those are gathered for artistic meetings or carry out by QA teams that was just the so the question was the footage I showed where is it from and that's no more players playing the game so it's not QA and it's yeah and how are those are used for evaluation and how do you evaluate whether it's you back effective or not sorry can you repeat how do you evaluate the effect the effectiveness of the conversation based on the footage you gathered I think so how do we evaluate the footage is that the basic question so my question is kind of how to integrate this into the pipeline and based on different iteration so how do we integrate this into the pipeline so this is often about analyzing what we've done in the previous pack and trying to take lessons from it and then incorporate that in the next design so it's very much a learning process sometimes if there are stuff that we're seeing say in the play test footage from the game design that we noticed something off we might be able to change it in the current pack but normally it's about learning from past experience and applying it next time thank you magic thank you oh you think that's already done cool oh thank you all very much okay oh okay okay yeah so the question was how do we deal with the designer how do we deal with the designer side of it so we talk to them a lot so yeah it's as simple as that so the briefs that we'll get can be anything from something as simple as this needs to be a place where the player defends to a very detailed list of everything they want so we will when it's a brief brief we might flesh it out and take it back to them and so when we feel like we've got a brief we can work off and we'll communicate that brief to them and we'll normally communicate as well like initial design ideas like I'll show them reference this is the kind of space that I think are doing will that work for you and they'll say yes or no and then yeah we show them as we work like we show them more when it's something difficult or complicated if it's straightforward we might just do it but yeah and then when they get at their playtesting and there are for any changes they need so let's say yeah it works yeah yes thank you for the talk would you have some advice for more open world I'm thinking about for example red attention where checkpoints sometimes when on the player has a lot of opportunity to wander around in the wilderness and to come up in a location from any direction do you are some you know advice or a way to read environment or to analyze it in a way that we can apply this yeah so the question was if we have any advice for open-world games where the player might be moving in any direction and coming from any direction so what I was talking before about the wayfinding elements so if there is a path on the scene the player is more likely to be following that path even if it's not even if it just is say the vegetation being less distinct in that particular area might be an animal path so we're more likely to walk in those kind of routes so you can just walk around the space and see it and you can also again see play tests to see where people are actually moving but yeah it's really as just about experimenting taking stuff into the testbed where does it feel natural to move how do things look when you move that way did that that's your question thank you thank you okay any more are you done well thank you all very much for coming [Applause]

Channel: GDC

Views: 53,253

Rating: 4.9629631 out of 5

Keywords: gdc, talk, panel, game, games, gaming, development, hd, design, read dead redemption 2, rockstar games, GTA V, grand theft auto

Id: L27Qb20AYmc

Length: 58min 9sec (3489 seconds)

Published: Mon Nov 18 2019