Repository navigation
Journal
The first (gameObject) stone has been put, and I now have a shiny TANKS! game in front of me. The tutorial went well, but i was still surprised by the amount of time I had to dedicate to only complete it, which was around 10-12 hours. I went nonetheless through it in good faith, and managed to discover a few good tricks. That's the pro of doing tutorial from the real pros (read: the ones making the engine), you get to do things the right way, so in this frame of mind I don't mind redoing basic stuff like settings up a scene or tinkering with prefabs. You always get better (at least at my intermediate level). The coding part wall well-orchestrated too; leaving the boring bits (variable declaration, etc) already done so that they could delve right into the real subject matter. I did felt like a typing monkey most of the time, but stopped at many parts to make sure I got them right (which might be the reason it took me so long to get through it).
Having spent my WHOLE work week on that, my intervention needed to be simple, with no turning back (because I was running out of time). Literal World of Tanks has been born under these condition. It's basically the same game with the same rule, but every gameObject is now a tank. The buildings are tanks, the rocks are tanks, the bullets are tanks and the tanks are... tanks. I was surprise how readable the game still is, which underlines the importance of rigorous color-coding, which separates the game in three caterogies; blue/red for the players, white/brown for the obstacles and yellow for the navigable space. There also something satisfying in seeing tank shapes among otherwise abstract larger shapes. For instance, the silhouette of a rock is still roughly the same but, as you squint (not so much) your eyes, your recognize a cannon, and wheels, etc. Pattern recognition kind of become another element of this version, where the player constantly looks in the environment to see these familiar shapes.
When brainstorming for ideas, I usually down down in my notepad and scribble some nonsense in connected bubbles and indented lists. But sometimes, I just get in the editor and move things. I fly the camera around, turn off some gameObjects, scale up or down some others... I'd stop as soon as I see something that pique my interest, or make me think of something. It's almost like pattern recognition; while my hands kind of automatically mess with things, my head is empty, or at least receptive, and observant. Two interesting ideas came out of that process.
Came up with this when I made the camera face the tank. The closed-up version of its cannon instantly made it feel more intimidating. To accentuate the feeling, I scaled the tank up and pushed the camera down, as if it was from a human's perspective. At that moment, I instantly associated it to the Tiananmen Square Tank Man picture, where a single man is blocking off a tank column.
How can I make this playable? Do you play as a first person controller who walks up to the tank column? i suppose you should start out of the way so that you actually have to make the decision to intervene... or let the column drive past you. The tanks would stop as they face you. And would keep on marching if you stepped aside.
I also made a night version with headlights on the tank, as they would accentuate the intimidating aspect of the tank. I suppose that sets the tone for the rest of my esperiments: I want the player to look at the tank differently. Not as this cutesy toy but as a machine of destruction. The next ideas is also based on the perception we have of the little tanksy guy.
I think the instigator for this scene is Pippin's "white light at the end of the tunnel" idea. More precisely, I was really interested in adding some sort of narrative around the minigame. Like, in his example, what happens after. This is a take on what happens before. I started parking tanks next to one another, colourless, and it already felt like some pre-match lobby, where players are waiting for the fight. I added walls around, and used a spotlight as the only source of lighting. The spotlight would be flickering as explosions are heard from the outside. After an explosion, the player would have control over a parked tank, and would need to drive it out, and then controls would be locked, until the next explosion, and so on and so forth... This scene also made me think of Portal: on the outside, the immaculate and soulless test chamber but, behind-the-scene, a bleak and desolate environment, with tons of equipment being seemingly wasted for nothing.
"Loneliness" was the objective, here. I suspect a tank surrounded in darkness might be a popular idea in the class, but I was still interested in the execution of it. In this scene, the camera starts centered on the tank. As the tank moves, the camera zooms out and reveal more of the scene. Well, it would reveal the scene if it wasn't unlit like that. However, as the camera zooms out, the tank shrinks, and shrinks, to the point where it is only a blimp on a pitch black canvas. This frenetic exploration, coupled with a slow zoom (and potentially a slow sad piano song) could have a great effect. The small tank is desperate for its opponent, and is looking everywhere, but the world is large and the night is dark and full of----- yeah. I'm especially attracted with how the camera is pacing the narrative (vast world, hopeless search) and that pacing is determined by the player's movement. I should be paying more attention to these potential "player-mapped pacing" moments.
So I think I am going for the "searching for red tank" prototype. Right now it consists of:
- a lone green tank, with added headlights
- lonely green tank goes from left to right, top to bottom, nothing is there
- as the tank move, camera zooms out
- as the camera zooms out, a light source is revealed
- upon investigation, the lonely green tank discovers another tank carcass
The "lonely green tank stumbling on a dead old friend" moment is one that I randomly stumbled on, when I was moving the green tank around in the editor view. I think the tank headlights highlighting the dead tank, coupled with the "searching for red tank" title creates a small melancholy bubble, as if the green tank was somehow mourning.
One issue that came up during runtime was that the camera was too far to appreciate the moment. Zomming it in before zooming it out would break the pacing, so I am thinking of juxtaposing two viewports together. I came up with two ideas today:
#1. Transluscent
The two viewports are comletely superposed. I think the visual works since the quasi-entirety of the zoomed out cam is black. It does have some kind of cheesy connotation to it, since theses faded-out images are usually seen in corny memes.
One thing I should try would be to have a different camera angle, so that the zoomed-in image doesn't just feel like an englarded version of the zoomed-out cam. Maybe try lookAt cam too?
#2. Inset
This somewhat less intrusive version shows a
Moving forward
First, some things are missing from the initial idea
- music (was thinking of a slow piano song)
- soundscape (wind)
- music to soundscape transition (as the camera zooms out, music fades out and wind fades in)
- other tank carcasses
- randomly spawned tank carcasses
Something else I should think about is when the second viewport is displayed. Sam suggested the other day that the image could be display as soon as the scene starts. This could be confusing (depending on the second cam config) but could have an interesting moment where the player eventually connect the dots. Downside to that idea is that it is not scalable to multiple tank carcasses. Or maybe, but would lose its surprise aspect. Gah.
First prototype is done. It obviously need some polish and debug, but all the elements are there. I decided to let go of the idea of a randomized spawn location for the carcass, and allocate my time somewhere else.
I think I've settled on the transluscent cam. Changing its angle got rid of my hesitation. The superposed image is less centered, and gives a better composition (close-up below, larger scene on top). A fadein/out effect for the raw image gameobject would really help also.
I integrated some some to go along with the experience. Again, everything is mapped to whether the tank is moving or now, and I really like the slow crossfade between white noise and melancholic music.
Right now, I don't know what to add. Or what to remove. I like that. Comfy state of mind. I'll leave this on ice for the next few days, try not to think about it, and playtest it on Monday to see where I might want to go next.
The build is in, the playtests are done. So far the (little) feedback that I got was positive. Keywords that came out the most were "emotional", "sad", and "like in real life", which was prettttty much my target aesthetics. So that's nice.
The final result is quite polished, visually and conceptually, but I know it could be pushed even further. For the Final version, here's a small todolist to remember:
- add an alpha crossfade to the second screen
- add a flickering behaviour to the emitter sphere of red tank
- add a slow flicker to green tank's headlights
- add a random timer on red tank's destroy script (so that it's dissapearance feels more natural)
This is the first entry of the second intervention on the TANKS! game. I should have written this one a week ago but I basically got sick, recovered, flew to Europe, gave a conference, got sick again, so NOW I have enough energy and headspace to write.
When thinking about the intervention, I was first interested in replacing the whole sound folder with contrasting sounds. This was my main reference. I first found an interesting library of death metal sounds, for instance. I was interested in flipping the mod of the game into something very rough and noisy, though I don't think this is leading me anywhere. Even if the aesthetics would change, the dynamic of the game would roughly stay the same. I don't think our perception of the game would change neither.
I was then tempted into making some sort of tank piano. Mashing keys on the keyboard, and have coloured tanks light up on different places of the map. It would have resulted in a very static piano game using the TANKS! visuals. I ended up scrapping that idea as well, this time because it changed the nature of the game too much. It was not about a TANKS! game anymore; just a funny way to display the tank meshes in a radically different context.
As I was writing my references, I then thought about the game Bastion, and its voiceover of the game. The more I think about it, the more I am settling on it. I'm attracted to the idea of making the player feeling self-conscious about every move, and every state of the game.
I could either go e-sport commenting, and being overly enthusiastic about everything, but I think it would be more interesting to have a calmer kind of narration, with every other sound in the mix toned down to focus on my voice. I'd describe the game as it is first, then maybe play with repetition, or underlining the boring/insipid gameplay of the game, then move on to other existential questions, or having some sort of one-way discussion with the player.
I would leave the game intact: i might only switch it to one player only with a couple of AIs, to make things easier to track. So yeah, it would be an interesting occasion to reflect on the TANKS! basic game AND an excuse to get more coding done (so that I can track everything that's happening).
Flu is out, one very good thing behind me. Got back to work yesterday, and spent the day flipping my concept up and down. One constrain was central: _the game should stay intact. I want to alter the sound to alter our perception and approach to the game, and see how the aesthetics change.
First idea was to narrate what was happening, with everything else silent. The narration would be as if this was fiction, rather than sport. I still think it's a cool idea, but I'm pushing this aside because this implies too much production time.
I'm now rather interested in a variety of changes throughout the same shitty TANKS! game. Play the game, but through different soundscapes that alter the personality of the game. The goal is (I was gonna write "as always) discomfort: make the player ponders about the motive about this game. We're having fun there, but really what is this mimicking? Are we clapping for the wrong reasons? Is this fucking disrespectful to some? Why is it always coming back to violence?
Four soundscapes incoming:
- Esports. GEE GEEEEEEEEE!!!
- Warfare
- Original sounds
- Spoken word track
Prototype is now out. For each round, I change the background music of the game. The result is rather simple, but there is something there. I'd say that, for now in my head, the only things missing would be to replace some of the tanks sounds like, for example, the tank's cannon sound in the "real war" soundscape. But that's it, not sure where to go after that.
But this thing revealed a deeper concern.
This prototype differs in the first one in that it is an actual intervention. Like, I modify the sound aspect of an already existing game that stays pristine in its original form. But unsurprisingly, as I was submitting my thing, I could see all the other games who were all crazy departures from the originals. Some, if not most, are basically games started from scratch, using some tank assets, that focus on sound. I don't know maybe this is just me being insecure, but the output of both approach differs quite a lot. Or am I just comparing apples and oranges? If it's a question of effort, then there no comparison to be made.
So this is it. Back in Montreal, and the sprinting to get up to date begins. So I think my rough prototype defenitely was a step in an interesting direction, with the notable issue that it was incomplete and I didn't knew where to go next.
The problem with this rough prototype is also one thing that I like very much about it: the game is untouched; only the soundtrack changes. In the final version, I will try to keep it as it is, but add other visuals. My solution would be to go inception style, and to make a game within a game. I would thus not only use sound to change the context, but use different environnements in which the game will be displayed to look at the different ways we can look at it.
I am now going in. The plan is:
- Esports: a low-income bedroom
- Warfare: a projected image onto a classroom's wall
- Original sounds: a corporate meeting room
- Spoken word track: an art gallery
Wish me luck.
It's in. The integration went quite well, but was much more time-consuming than i thought. The final result somehow has the same vision of the rough prototype; it cycles the same tanks game through a series of different contexts. A seemingly lonesome person watching a tanks championship match on Twitch. A classroom watching a war documentary. A dev meeting doing a build review of the game. A war museum displaying gruesome statistics. What I'm hoping for is that all the themes tie back to violence, either its exposition, or its banalisation.
Playing the tanks game through another screen is easier than I thought. Maybe I'm way too used to play it, but there is something very interesting to have a game inside another game. Have I had more time, I would have upgraded the controls and make the player able to move the camera around, and maybe interact differently with the game (turn it off? change mode?).
Here are the four final rooms.
I struggled for SO LONG to find a concept for the 3D object assignment. Wow. This is what happen when you have to be creative while fatigued. I also lost a lot of time trying to embed brutalist architecture onto a game that has nothing to do with it.
This rough final version is rather, like the sound prototype, a shift in perspective. One of the main things that is pissing me off about this tanks game is not its shallow gameplay, but its cartoony depiction of violence.
It's such an innofensive game, but at the same time it underlines so well our unsensitiveness to violence. And that it's an official tutorial. It reminds me of the Unreal Engine first person game template: you start with a placeholder gun in your hands.
So the point of this intervention is, again, taking the same things that are packages in a cutesy way, and subvert it into something less funny.
This current version is basically a first person controller in the untouched map, with added particles / lighting that give another ambience. I added some still tanks around with a lookat script. I don't know yet why I did this but for now it really gives them a chilling effect.
I'll leave it ambient for now. And playtest it, and see where I want to go next.
After making the first version of the first person prototype, I had a long empty moment of me not knowing where to go next. A good tactic of mine in these mentally blank moments is to spawn in the map, and navigate around it.
After a good hour of running around, I started sketching a few other things based on the following rules:
- layout needs to be as untouched as possible
- tanks will be used as static objects
- and as always, focus on the melancholia around the tanks
The approach of working on an already existing and "set" landscape is very interesting to me. Traditionally, level designers shape the terrain / objects according to their needs, but in that case it's the opposite: I need to design around something already "existing", as if I was building something on a landscape in real life. This means that composition will never be perfect; the boundaries will be more difficult to show (unless I put actual barriers, which I won't).
Dioramas are one possible path. Since it's a 3D object assignment, I think having a few setups of re-used objects will be interesting. Something like Braid's books could also be interesting. I'd be sort of a walking trip from a tank to another, each one saying something different. The tanks following the player like stray dogs could also be interesting. Anyway, FLOP is in an hour and I'll get some more fuel for this.
So before trying everything else mentionned above, I'll just go for the main constraint first : stay as relevant to the original game as possible. Now that I captured the aesthetic that I was looking for, I'll integrate the tanks AI and see what happens.
Game idea: a 3-player networked game in which two players play the original TANKS! game, while another player is playing in an adjacent room the first-person version. I can't code this but I like this. Very much.
How used to violence are we, as gamedevs?
Unreal Engine's first person template gives you the most basic ammenities to create a first person game. They took care of adding a default gun, too.
One of Unity's most popular and complete tutorials is a game about tanks destroying each other
You can find it in the most unnocuous spaces, barely there, but at the same time all-emcompassing.
Violence do have a place in games. Answering #notallgames to the WH's "violence in games" video is hurting videogames. Violence has a cathartic potential in games.
I have beef over us being gradually desensitized from it. Or using it sometimes in a sexy and attractive way, like the cool-ass operators in Rainbow Six with fans cosplaying as special force soldiers. Or the default use of "vague middle-eastern" setting, like the tanks game.
I'm tired of this. I'm tired of indirectly representing this world.
Final proto is on the way. i think at this point it's all about cranking up the immersion. It'll be about:
- adding more realistic sounds
- adding a camera shake
- add headlights to tanks (from desert map)
- better explosions
- more tanks?
Also I need better pacing:
- empty with distant fighting sounds at first
- OMG WAR IS HORRIBLE
- ???
- Emptiness
So I'm done. I spent a good amount of the day fighting against the horrible AI, which gave way to way too many comedic situations. That's the issue when you want to make a serious point: any bug fucks your rhetoric up.
I also found out that the calmer the scene was, the better atmosphere it had. I could not make combat scary, so I need to dial down on the bleakness of the environment. I still added one patrolling tank in the map with agressive lights to correlate back to the 3D objects theme, and how the cutesy tank isn't always cutesy.
It definitely is a challenge to design a map in which you cannot control all the parameters, the same way a landscape photographer must deal with the landscape he chose, and use the parameters in his or her control to the best (position, time of day, etc). In my case, lighting, scale, particles and camera were the only things I added / modified, and it gave way to a radically new game.
note for Pippin: I'll condense the final four entries into this one
So the experiment is done, the final builds are uploaded, the second year of my BFA is over and I am now done with Concordia University for at least a long time. This game marks the end of that era. I'd say it feels fuzzy on the inside but I just feel pain in my back right now.
While this tanks project got gruesome at many points this winter, I still managed to get a clearer understanding my of vision(tm) and got several intense conversations out of this thing. So for one last time, let's list the things that came out of these three game-essays, and call it a day. Or a year.
This first experiment is, I believe, my most successful. I think one of the reasons why is that it feels complete. It has a clear beginning (o shit it's nighttime), a clear -- yet monotone -- progression (this world is huge, this world is empty, a clear breaking point (omg red tank nooooo) and a clear ending (lights off). Well actually, maybe the ending isn't that clear. Is the green tank dead? Did it went off-bounds? Or is this just a cut to black? I think I've made it clear by then that I like games that have ambiguous endings, and having the sond continuing to play while not being able to do anything falls into that category: this kind of conclusion has the potential to make the player wait, and maybe get lost in the poetry / evocativeness of the moment. And that is great.
on a more practical level, the juxtaposed camera has a great effect. I don't think I'd be able to make it work in a lot of situation, but this scene has so much negative space, it was the perfect occasion to try this technique. Mapping the camera zomming out and the audio crossface to the movement of the tank was also a great effect, since it amplified the feeling of being lost; instead of letting the player go in a straight line at first, it forces him/her to run in circles, which is the kind of movement that epitomizes the feeling of being, you know, lost in the middle of the night.
This project started on a wieeeeeerd note, as I was horribly sick in Amsterdam and past the deadline. I rushed this prototype will little thought beyond it, and ended up cutting 74% of it in the final version. There is however a few interesting thing to take out of that. The title, for once, has a humorous effect that also acknowledge (maybe ironically) that the experience is quite short, to-the-point yet a bit cryptic, and somewhat pretentious (games! art! what are games?.
The feeling of playing TANKS is also changed with the absence of sound (that has been replaced by this monotone voice) and the emptiness of the gallery. I think it does push the violence forward, which is maybe trigged by the tank carcasses? Or the voice enumerating deaths? It's still hard for me to tell, but it's much harder to play the game with the lightness of the original one.
This exercise was motivated by its initial constraints: DO NOT TOUCH A THING. It took a while to get it right but, in the end, I only added a few FXs, tweaked the lights, scaled-up the world, added some post-process and that was it, you were walking into an entirely different world. It is a very short experience that can be stretched into a long exploration, depending on the attraction of the player to that emptiness, and bleakness. I always like to see the huge tank slowly crawling from far away, there is something very sad and melancholic about it.
This is the most incomplete game of the whole bunch, I think. And I have this idea for a more complete version: Grounded is a 3-player networked game in which the first two players are playing the original TANKS! game on the same screen. The third player is playing on a different screen (and in a different room?) and plays with a first-person perspective inside the TANKS world.
This project is, in the end, two things: a formal experiment that not only build a better method of game-making, but also make specific reflection on specific tools of game development (camera, lights, sounds, 3D objects) as well as made me learn more about remixing things, and making games out of already existing stuff.
It also turned out to be a very interesting conversation on everything that orbits around violence: death, grief, isolation, loneliness, memory and melancholy. I've whined to much about violence in games, and making games that pokes that really does feel good, and actually add something meaningful to the conversation.
Looking forward, I think I'd actually like to use the TANKS! template elsewhere. Maybe in jams, maybe in workshops, maybe in other projects. This game still has a huge remixing potential, and I feel like the potential commentaries about its violence have yet to be discovered.