The First Two Hundred

You know how they tell art students that their first two hundred paintings will be rubbish? Well I'm a game maker, making games. I haven't made two hundred games yet, and the ones I have made have been kind of rubbish. As I make my way to lucky number 200, I'm going to blog my efforts making these games; what's inspiring them, how I'm making them, and what I'm thinking as I do. It's going to be messy, embarrassing, and fun. Join me!

[thefirst200]

Game #9: Safetopia

These past couple of weeks I’ve been helping a friend to put together a proof-of-concept for a point-and-click adventure game. I haven’t really worked on a point-and-click adventure before, so I thought I’d lend her a hand to learn the process, the Adventure Creator toolset, and of course, to add to my folio, and this blog! You can play it here.

The game was called Safetopia, and it was set in the eponymous city. Safetopia is a city obsessed with rules, laws, and safety, despite being perched precariously on the edge of a cliff. It’s a kind of comic dystopia where the inhabitants of the city are oppressed by silly warning signs and strict curfews. You play as the Champion of Anarchy, a rebel and freedom fighter sent to the city to release some other rule-breakers from prison.

The section of the game we produced was the introduction, which involved trying to make your way into the city itself. It consisted of two scenes, two NPCs, and one puzzle. The art, design, animation, and writing was produced by Alice Bittisnich; you can find her work here. I handled all the work in Unity, which involved organising the scenes, scripting out the interactions, and implementing the assets. I also did a little bit of sound design and general production, so this post will be a mix of technical implementation and reflection on my thought processes.  As I mentioned, I used the Adventure Creator toolkit with Unity to put together the project. Adventure Creator is a great plugin which facilitates the creation of adventure games, both 2D and 3D. It’s quite flexible in terms of how you approach it; you can bringing in sprites, Spriter assets, or use Unity’s legacy or Mechanim animation systems to import assets. It’s also very non-programmer friendly, as the vast majority of interactions you would need in a typical adventure game can be assembled with a drag-and-drop diagram system.

I spent the first day on the project basically following the 2D tutorial on YouTube, which covered almost all of what I needed to accomplish anyway. After setting up a basic scene, and programming out some basic player movement, I could begin to place the important interactive ‘zones’ into the scene. Like the programming system, this was a very visual process, which made it very easy to see how certain elements would be positioned, and which would be interactive. This included defining the walkable area of the scene, called a NavMesh and designated by the white border in the image below, and elements which could be interacted with in some way, shown by the yellow square and lines.

You can see that the sprite for the city itself is divided into two interactive zones: ‘City’, and ‘City Entrance’. This is for a couple of reasons. First of all, by default, Adventure Creator assigns the left mouse click to using or talking to an interactive element, and the right mouse click to examining it. This is useful for PC and Mac users, but people playing on a laptop might have a hard time. So in this scene, clicking the ‘City’ section will cause the camera to zoom in it while the Champion of Anarchy makes a comment (the equivalent of an examine function), before the camera returns to focus on the player. Clicking ‘City Entrance’ will cause the player to move towards the city and into the next scene.

Of course I had to try to imitate 3D space as much as possible, and luckily Adventure Creator allows you to seperate parts of the scene into layers, which shrink the player down or up as they move between them, making it look like they are walking into the environment effectively. The next major task from here was dialogue. This was probably the most labour-intensive, as easy individual line of dialogue has to be organised in the ActionList section, which is where all the programming occurs. I also had to sort out conversation options, which involved creating dialogue trees to determine where the conversation would go depending on which option was chosen.

Once those basic interactions were mapped out, I was able to focus on doing the same to the second scene. This second environment, set at the gates of Safetopia, was a little more complicated, as it involved the player interacting with an object to pick it up, and adding it to their inventory. A lot of the infrastructure for this process was already in place, but it did require a little bit a trickery to make the character bend down, play a ‘picking up’ animation, and making the object disappear from the scene. I think my experience putting together Sai’s Quest really paid off here, as it taught me a lot about the basics of putting together 2D games in Unity.

My original idea for the two environments was to seperate them into two Unity scene files. I tried this, but I soon realised this would be inefficient, as I would need to rebuild a lot of the settings and interactions I had set up for the first scene; not to mention figure out how to keep the player’s inventory items and keep the music playing across both scenes. So instead I cheated a little, assembling both environments in the one Unity scene, and fading the camera out when the player reached the ‘City’ object. In the split second the camera was faded out, I teleported the player and active camera to the second environment, and switched out the NavMesh. This not only made programming the project a lot easier, but it also removed any loading times, and allowed me to work on both environments without having to constantly switch Unity scenes.

The above image contains most of the interactions in the game, including the an early version of the title which springs on to screen during the opening cutscene. At this point, with most of the technical work out of the way, I was able to provide a little bit of feedback on the art and animation. I didn’t really have time for research, but I know from memory that environmental animations are a really important part of making a point-and-click adventure feel alive and dynamic. You’ll notice, even in really early adventure games such as The Secret of Monkey Island, that most of the background environments have at least one animated, looping element. Over the years, this practice became more and more involved, until you have games like Kentucky Route Zero where you have swaying trees, blinking lights, running water, and ambient, staticy effects.

I think it was a priority for Safetopia to include some similar animations, especially since they would be so simple and fast to produce- only a few frames each. Alice decided to focus on the following areas: the moon’s reflection in the ocean, the lights of the city, the neon sign above the gate to the city, and the trash cans outside the gate. The flies hovering above those trash cans were particularily productive I think, as they also drew attention to the fact that the player could, and should, interact with them. Overall, I think these little environmental animations really add to the scenes and make them more engaging and lively- similar to the little bits of motion in Take That, Ya Lousy Dimension! It’s good to see that I’m able to use knowledge from making these other works in my current projects already.

I spent the last few days on the project tweaking text sizes and adjusting animation speeds. Admittedly this was a tedious and somewhat frustrating process, but that stage of any project is hard to avoid. I learnt a lot in the short time spent putting together this project; particularly about the Adventure Creator toolkit, and adventure game design principles. It was quite refreshing to be able to just put my head down and work on a project without having to worry about figuring out a piece of programming, or being bogged down trying to produce art assets. While the genre isn’t really my creative focus, I had a lot of fun putting together Safetopia, and would love an opportunity to work on another point-and-click adventure at some time in the future.

[thefirst200]

LoveMaze: Writing the Text

Once I had the basic structure roughed up in Unity, I knew the next step would be to put the text together which would be covering the walls of the maze. This involved going back through old letters, texts, IM messages, and memories, and actually turned out to be quite an emotionally painful process; negative memories were re-lived, and positive ones turned into thoughts of “what if I had said or done this...”. Hence the delay in getting this part done, but at least no-one can say I didn’t suffer for my art!

I began in Photoshop, with five different groupings of text, one for each relationship. 

These were completely raw pieces of text; either directly copied and pasted from text conversations or letters if avaliable, or quoted as accurately as I could remember from conversations. My aim was very much to represent the positive and negative in each relationship. Even if they soured later, or ended badly, I didn’t want to naively represent them as purely good or bad experiences; as with most experiences, there was good and bad, they were joyful and painful.

I gave each relationship roughly four sentences each, and arranged them in lines in an attempt to fill the page. As you can see above I got pretty close, but there are still quite a few gaps. It was also proving difficult to manage all of the different text layers, even if they were grouped appropriately. 

I decided at this point to just use one text layer, and to place all of the quotes in the same place to make them more manageable. This made it much easier to edit and arrange them, so I spent some time rewording, cutting down, and formalising many of the quotes so they would all form one nice, solid chunk of text I could place on the walls of the maze. I replaced all of the gender-specific pronouns with gender-neutral ones, to make reference to all relationships, rather than just the ones I experienced as a straight male. I also tried to generalise the quotes as much as possible, to respect the privacy of the people they concerned. 

It was then a matter of blacking out the appropriate lines of text, so that I could save out five different texture files, each with a new group of quotes added. My plan is to fade between each successive texture file in Unity when a new pickup is collected, creating the illusion that the new pieces of text are appearing in the overall group. Whether it’s as simple as that remains to be seen, but for now, this part is done!

[thefirst200]

LoveMaze: Modelling the Structure

This post will just be a quick documentation of how I modelled the structure of the maze itself in SketchUp. It was quite simple, and it certainly seems that all that pre-planning paid off, which is good! I started just by opening SketchUp, and importing a screenshot of the maze layout I had produced in Photoshop.

From here, my plan was to use the rectangle tool to trace over the walls I had defined in this level design. This was pretty simple, and worked fairly well, but I did notice that the thickness of the walls was a bit all over the place as I drew them out. To remedy this, I started again, this time making sure to draw out a perfect square whenever the form rounded a corner. I then drew out the rest of the wall from that point, making sure that, even if the walls weren’t in the exact right position, at least their thickness would be consistent.

Using this process, I was able to draw out the entirety of the maze in about 20 minutes or so. This was where my choice to use SketchUp really paid off, as to achieve this same model in Maya would take a lot longer, and probably require some extensive UV work. With SketchUp, it was simply a matter of drawing out rectangles, and making sure that they were nice and consistent.

At this point, I just had to go around and use the eraser tool to get rid of the leftover edges on the corners of the maze. There were also a couple of not-so-straight lines, but these were also easy to eliminate with the eraser. Then, it was just a matter of pulling the form up (what would be called extruding in other 3D programs) and I had a maze!

Now I can drop this mesh In Unity, add some colliders, and move on to applying some textures!

[thefirst200]

LoveMaze: Finalising the Structure

Picking up from where I left it yesterday, I jumped straight into evening up the planned structure of the maze, so it wasn’t as asymmetrical. 

This mostly meant re-shaping some of the more spiral-shaped pathways on the left, so that they better matched the love heart overlay I was using. This was better, but now the focus was unnecessarily on the room in the centre-right of the maze. Luckily this was easily fixed during the process of the next step: difficulty.

I used a pretty simple rule to begin to design out an increase in difficulty: after a pickup is collected, the amount of wrong pathways increases. So after you pick up the first one, there’s only one wrong turn you could take (with the exception of going back toward the start of the maze) , and after you collect the second pickup there are two potential wrong turns, and so forth. This became difficult the further along due to space limitations, but I found other ways to increase the difficulty, including a loop near the end, and multiple pathways from a single room. Whether or not any of this will actually be difficult, engaging, or fun remains to be seen, as I’ll have to have to actually get in there and play test to see for myself.

Finally, I roughed in the locations of the pickups, and marked in blue the locations of locked doors. I tried to utilise this mechanic in further defining the difficulty; you can see that the first door is right in front of the first pickup, but as you get further along, the unlocking process becomes more and more complicated; you’ll have to actively seek out the next door for example, or even walk past it before you find the appropriate pickup (which I’m thinking will act as keys), causing you to have to walk back to find the door again.  Hopefully this work will pay off later on and I won’t have to make as many edits in-engine down the track. For now though I think it’s a pretty solid plan, and next I’ll be using it to put together a basic mockup in SketchUp.

[thefirst200]

LoveMaze: Designing the Structure

Now that the pickups are well on their way, I want to now focus on the actual structure of the maze itself. There’s a few things going into this idea, so I’m going to break up this post with a few in-progress screenshots to give you an idea of how it evolved over time. First up, this is the very basic structure I currently have in Unity to test out scripts and ideas:

It’s made up of cubes that have been stretched out to form walls, all arranged in this love heart shape. I like this shape, and wanted to do it from the very beginning, obviously because of it’s visual connotations, and because I think it’s well suited as both a logo for the game, and a nice way to construct the labyrinth around a central theme or idea. I put together a grid in Photoshop, then printed that out and using a pencil blocked in a first attempt at a more complicated maze.

Already I think this is quite visually interesting which is a good sign, but there are a few issues which you can see I’ve noted in the bottom right. The biggest issue is that I originally wanted to have the final pickup open three doors at once, overwhelming the player with choice, but this didn’t end up being feasible with the door-unlocking system I later devised (which I will elaborate on when I get to programming it). The rest of the issues I slowly tried to resolve over the next few iterations. 

As you can see my first change was to jump into Photoshop to further develop the design. On paper is always useful as a starting point but it’s much faster to iterate digitally here, especially when I can just treat the design as pixel-art, since I was already basing it on a grid. The above image is pretty much a re-creation of the sketched version, just to get an idea of how to all might look on-screen. 

The first issue I tackled was the closeness of the pathways, which I felt was much too tight in the original design. I wanted there to be some ‘breathing room’ between them, both visually from a top-down perspective, and as you’re in the maze exploring it. At the very least, more white space inside the heart shape could be filled with other pathways to follow. My solution was quite simple for this; I just made the pathways (in beige) twice the size of the walls (in pink). Still not perfect, but a good start.

The heart was quite skinny still, so I wanted to stretch it out a bit into a more traditional shape, closer to the original design in Unity. To begin this I grabbed a love heart image from Wikipedia Commons and laid it over my design in Photoshop. I then broke it apart and spread a few sections out to better match the overlay. I then spent a bit of time connecting these seperate parts in a logical way.

I then spent a bit of time separating out paths, re-connecting them, and creating new ones to fill in areas of excessive white space. Using the template proved to be quite useful in replicating the love heart shape, even if the below shape is a little too asymmetrical for my tastes, so I’ll work on that.

Next up I’m going to continue to work on this design, and start to think about it from a level design perspective as well, in terms of how difficult I want certain sections to be, and how the structure will work with the pickup system.

[thefirst200]

Organisation

This will be a very brief post, just quickly outlining my current approach to organisation and why I think it might be useful.

It begins with this cork board, which I picked up a couple of months ago but decided to put to good use only recently. I’ve blocked out the general projects I have to or want to complete, and then below those some more specific categories for general tasks involved with those projects. The idea is I can use this to structure the next step of organisation, and take a note off the board when I’ve completed it as a motivator. I took the ‘pickups’ note from the LoveMaze column just before taking the above photo for example.

I then use Wunderlist to organise each of those tasks in a little more detail. First, I basically just put in those general headings from the board, as in the above image.

Then, I put down whichever sub-task I happen to be working on at the time. As I’ve just finished the pickups, that’s what’s still under the LoveMaze to-do list. Once tasks are ticked off as completed, they are automatically hidden, which is useful.

Finally, under each of those sub-tasks, I specify the specific tasks needed to finish that task. For each pickup, that usually meant modelling, texturing, and lighting, but some required a little bit more, such as tweaking shaders or implementing particle effects.

Finally, once that’s all been sorted out in Wunderlist, I add each task as an event in Google Calendar, where I can simply click and drag them around to structure the upcoming week. It’s a little tough right now because I’m still sorting out solid commitments such as regular classes and meetings, but that will hopefully become more concrete, and thus easier to schedule around, in the next couple of weeks. It’s also a little hard to determine how long certain tasks will take to finish (on average I’ve found I’m able to complete around three tasks a day), so hopefully I’ll improve at that moving forward as well! At the very least, this has been useful in figuring out a more reasonable and productive schedule. 

[thefirst200]

LoveMaze: Pickups complete

Alrighty! So the pickups for LoveMaze are complete, or at least have a couple of decent passes done on them, to the point where only a little bit of tweaking should be required here and there. The watercolour style worked well for the most part, although I did have to make a few adjustments to a few models to accomodate. The first pickup model I’ve already posted about, so I’ll just briefly go through the other four one by one and comment on their production.

Although this pickup ended up being pretty straight forward to put together, a lot of time was spent here experimenting and developing techniques which later became useful for the rest of the models. Originally I had two seperate models for the wine and the glass, but this resulted in some weird glitchy stuff happening when I put it all together in Unity. Instead I made it all part of the same model, and just reversed the normals inside the wine glass so they would appear to be part of the liquid when the outside polygons were rendered transparent to look like glass.

You’ll notice that the liquid inside the glass is titled; this is so that when I rotated the glasses inside Unity as part of the final composition, the wine would be level. I played around with the texture of the wine for a while, and settled for a watery filter over the watercolour one. I’m not 100% sold on it, and I think I’ll go back later and just use a flat, deep red. The P-Plates I also played around with a little, and their texture went through a few iterations. I didn’t want to just use a straight up photo, so I tried sketching out a ‘P’ with irregular edges, but it just looked too messy. I ended up using a photo, but sketched over it using an off-white and the brush tool in Photoshop, then running that through a watercolour filter. This technique would prove useful for depicting glass and other ‘plain’ surfaces later on.

This pickup was initially going to be a jar of lolly snakes with that floating ‘u’, but that ended up being too much effort to model, texture, arrange, and get it to display properly in Unity. So instead I grabbed some hard candy models I’d used in a previous project (fittingly, one that also referenced this particular relationship), and dug out a birthday card this gal had given me. Obviously I didn’t want any identifying information on there, so I removed any names and smudged out most of the text on the outside of the model. The text on the inside is still pretty clear, but it’s too difficult to read with the pickup constantly rotating.

This pickup was relatively simple, but took some precise texture work. Those two white sticks are chopsticks I was given after a trip to Japan, and the drink is a bubble tea thing we have here in Melbourne called Chatime. The model for the drink was initially a lot more complicated; it involved the cup, the straw, individual ice cubes, individual models for the fruits (those black things at the bottom), and a model for the tea. In the end it was much easier to just texture the entire cup as if were filled with tea, and just focus on the detailing on the lid. This worked out well, as the tea and black fruits fit well with the watercolour style.

The final pickup turned out to be the simplest, even with some added particle effects.

It’s a couple of cigarettes (featuring rising smoke, just the default smoke particle effect provided with Unity and a little tweaking) and a couple of sheets of paper. The papers are actually a letter I wrote this lady but never sent, and I used scans of the actual letter as the texture from the model, again with the text itself obscured so it can’t be read properly. You’ll notice there’s no spotlight below this one, which is just a weird glitch in Unity I’ll have to sort out later. When you move to one position it appears, move to another, and it’s gone.

So that’s about it for the pickups. I still want to add a few elements of polish before they’re totally finished, including some kind of visual effect and sound when you pick them up. I’m also thinking a thin light fog, similar to the one I used in Looking Back, and some subtle particles around the pickups to highlight their importance even more. I also had some ideas for integrating the pickups into the gameplay a bit more but I’ll expand on that when I get to scripting this thing up properly.

Finally, I wanted to include a little paragraph in this post on some organisational tools and techniques I’ve been using to balance the various commitments I have with the projects I want to work on, but I’d kinda rather watch a movie right now. I’ll throw it up tomorrow, along with an update on a small little experiment I’m working on for my PhD.

[thefirst200]

LoveMaze: Pickups modelling, texturing, lighting

Finally able to get back into it this week after sorting out leaving my job and having a cold for most of last week. I decided to focus on LoveMaze for the next little while because as I mentioned in the last post, it’s probably the furthest along of my current projects.  Over the last couple of days I’ve just been figuring out how to present the various pickups that are scattered throughout the labyrinth. I started with the first one, and used a photograph of an old CD-R I had lying around to play around with filters in Photoshop.

I opted for the watercolour filter because I pretty much just like how it looks. I’m not going to say why it’s labelled ‘Wayne’s World 2′, but that was added just by typing it out and using the Smudge tool to distort it a little. I’m particularily happy with how the base of the CD turned out- from a distance it looks almost like a photograph, but when you get up close you can see the detail of the brushstrokes. In a 3D environment this is quite interesting visually, as the player is constantly moving towards, away from, and around the object. I’ll think I’ll adopt this watercolour look as the environment’s visual theme, doing what I can by editing photographs, and painting in details to give the pickups a bit of a more stylistic look. 

Once the objects were in Unity, I was able to position them together and add the proper materials. I also put together a simple script to cause the objects the float up and down and rotate, the first step in presenting them as an important object in the environment.

The script essentially ensures that the object is constantly rotating at a speed of 50 degrees per second counter-clockwise. It also moves it up in the world 0.5 units per second for 1.5 seconds, then moves it down. The logic dictates that when it’s not moving up, it’s moving down, and vice versa. With all of that in place, I wanted to further emphasize the object’s importance with some lighting. I started with a fairly strong spotlight illuminating the object from above, which gave it a solid base and directed the player’s eye towards it in the otherwise dark environment. Unfortunately this meant that the base of the CD model, an important element, was in shadow, so I placed another, less powerful spotlight below the object, and disabled shadows. Finally, I added a third light in front of the object on the ground, facing up at an angle, to create a bit of a glint in the lenses of the sunglasses. 

This is how it looks in Unity. There’s still some tweaking to do, but as a basic layout I’m happy with it. I’m guessing that each pickup will need to have it’s own individual lighting, but as a general purpose setup I think this works pretty well, and hopefully the others won’t require too much tweaking. I’m going to focus on texturing, positioning, and lighting the rest of the pickups over the next few days, so hopefully that will be done relatively quickly. There are a few issues I’ll have to resolve (Unity renders transparent objects in a really weird way, as you can see by the shadow of the lenses in the below GIF), but I’ll cross those bridges when I get to them.

[thefirst200]

Quick Update

Just briefly: I have finally had a chance to reorganise this blog a little, so if you go to the main blog site,  you’ll notice a new, clearer layout, and a link at the top of the page to all of the games. Each item on the list collects all of the posts for a particular project for easy viewing. Right now of course only the latest project has a significant amount of posts, but that will change as I complete more games. Also! I am starting a PhD in a few weeks on virtual environments. This means a couple of things: I’ve started another blog for posts pertaining to my research, from responses to readings, to general thoughts, to development updates.  It also means the amount of time I have to dedicate to these projects has been drastically reduced, as I will essentially be treating the PhD as a full-time job. I won’t stop work on the 200 games or this blog, but posts may be a bit sparser, and their content a bit lighter :) I finish up at my job at the end of this week, so hopefully I’ll be able to post a bit more, but until then, feel free to jump over to the PhD blog if you’re interested in hearing about what I’ll be researching!

[thefirst200]

Re-structure

As I mentioned a few times over the development of my last project, I felt that forcing myself to work on one idea at a time isn’t really a productive way for me to make stuff. So for the next little while I thought I would try a different method, where I essentially work on what I feel like working on, and don’t worry too much about grinding through one project at a time. I’m hoping this will make the process more enjoyable and productive, as when I get sick of one project, I can work on another for a while. I still intend to finish everything I start, but I obviously won’t be able to predict when something will be finished for the most part. What that means for this blog is that posts will be a bit all over the place, most likely I’ll post about one project for a while, then another for a little bit, and so forth. I’ll try to tag posts with a certain title and see if I can set up some links on the page so you can view only posts which pertain to a certain project, but that might take a week or two cause I suck at Tumblr. Just for now though, I thought I’d give a little introduction to each project, so you have an idea of what each is and what stage of completion it’s at, then from now on I’ll start with the new structure.

Darklands

Darklands doesn’t actually have a title right now, it’s just what the folder is called on my desktop so I guess I’ll go with that. It’s an exploration-based game where you wander a large structure in the middle of an ancient crater. I’m working on it with a friend and artist, who is responsible for the above concept art. It’s a little difficult to get the ball rolling with our schedules, but it’s definitely something that’s in the back of my mind constantly, and definitely an idea which still excites me. We’re still producing concept sketches and spitballing ideas right now, and thinking about how to present the space, whether that be first-person, third-person, fixed-camera, etc.

The Test

The Test is a first-person, virtual adaptation of a radio play of the same name by Scott Carrier. I first heard it on an episode of This American Life almost a year ago, and it quite honestly blew me away. I adapted it into a Twine game a few months after hearing it, and decided to expand that idea to a series of 3D virtual environments. The idea is the player will walk through a series of continuous spaces which connect in abstract and novel ways, like opening a door and finding yourself in the middle of the ocean, or getting off an elevator into a living room. As they move through these spaces. text and voice-over from the radio play would be heard, leading the player through and acting as the narrative. As you can see I have a few of these areas blocked out and a basic visual style down, but I haven’t really thought of every single environment or how they connect together, so right now I’m just producing areas as I think of them, and hopefully as I do more and more I can start to form them into a coherent whole.

 Heikegani

Heikegani is a single-location virtual environment based on the Japanese myth of the heikegani crabs, which are said to be the physical manifestations of the souls of samurai killed during a battle in 1185 A.D. It consists of a small island which the player can explore at their leisure, and features some old ruins, and the crabs themselves. I’m trying to experiment with visual style with this one, by painting all of the textures with ink and paper, and scanning them in to apply to the 3D models. I have a basic prototype which can you see above, with the essential elements included and textured with a similar technique of fine-liner and crumpled paper. I’m working on this with a programmer, and we’re trying to produce some unpredictable and emergent behaviour from the crabs when they interact with each other and the player by giving them a set of A.I. ‘rules’ they have to follow. This is also a little difficult to work on due to scheduling, but I think we’re both keen and work on it when we can. Luckily there’s a lot of work I can do just by myself on this one, collaboration isn’t as necessary most of the time like it is in Darklands.

Advice

Advice (will have to think of a better title) is a relatively simple app for Android/iOS which gives you a piece of advice one per day. The idea though is that each piece of advice is slightly abstract and contradictory. I have a list of advice for this ready to go, which all up would last a couple months if someone used the app daily. Some of the advice is positive, some is negative, some is genuine, some is completely against what I believe, but it’s these kind of conflicts I’m interested in, to see which pieces of advice people follow or reject. The core idea here is to present information or ideas to people, and have them start to think about them and whether they themselves believe it, and what that says about them. More of an art piece than a practical app really.  I have the text all written up, so hopefully this one won’t take too long to finish off. I just had to settle on a font and layout, figure out how to dispense advice only once a day, and get it up on the App Store and Google Play. 

The Tower We All Must Climb

This is an environment which came about from just mucking around with some ideas, and had slowly formed into something more interesting, and still is. I used it as an excuse to learn a bit about SketchUp, so right now there’s about 15 houses dotted around this giant white-washed landscape. I also really like the idea of climbing a massive tower, so I included that, also made with SketchUp, and now I’m shifting towards focusing on the tower and integrating the houses into the climb somehow. I’m thinking the tower could form as you climb it, and the houses could appear that way, but we’ll see! This one is still very much in flux, so I have no idea when it will be finished.

LoveMaze 

Finally, LoveMaze is an extension of a small game I made a year and a half ago for university, where you explore a maze surrounded by text. It’s kind of a tribute to/reflection on various romantic relationships I’ve had in my life, whether they be girlfriends, crushes, rejections, or whatever. The text on the walls will be various quotes from such relationships, both positive and negative, and at five points throughout the maze there are floating pickups that symbolise certain relationships that I think of as key. As you pick up one of these objects, a part of the maze opens up even further, so I’m hoping for this sense that each of these relationships, whether positive or negative, is a chance for growth and development.  There’s also reference to the idea of love more generally, where you think you’ve found what you’re looking for, but eventually, especially when you’re young, you find out you have so much more to learn and explore. This project is about halfway done, I really just need to add the text, finalise the model of the maze, and polish up the textures and materials. No idea what I’m going to do about music though, because I think that will be an essential component. So those are the projects I’m working on now, and thus the shape of this blog for the immediate future. If I start any new projects I’ll be sure to add them to this rotation, and I’ll try to only post about work I actually plan on finishing. I’m back to work this week for about three weeks so posts might be a little slow, but hopefully I get at least one of these finished before I start my PhD in late February! 

[thefirst200]

Game #8- Take That, Ya Lousy Dimension!

Take That, Ya Lousy Dimension! is done. You can play it here. You can watch some gameplay here. This will just be a wrap-up post to reflect on the project, and see if I can’t pick out some lessons going forward. First though, some good news! I posted the game on itch.io roughly 24 hours ago, and and as of writing it’s been viewed 790 times, downloaded 266 times, and 11 kindly folks have donated, which is great! This is largely thanks to the fan website Four Finger Discount, who posted an article about it overnight. I know these figures are exclusively because I remade a scene from an enormously popular TV show, but it’s still nice to see those kind of numbers on something you created. It’s certainly much bigger than anything I’ve done in the past; here’s some comparisons: Sai’s Quest Views: 138 Downloads: 0 Broken Space Views: 71 Downloads: 3 Looking Back Views: 94 Downloads: 8 This also shows how much word-of-mouth and promotion can affect the amount of attention receive. Itch.io keeps track of where the traffic is coming from, and right now 647 of the visits are from Four Finger Discount, with the remaining figures a mixture of general browsing on itch, my website, and a reddit post I made. I’ll do a proper social media sweep (Facebook, Twitter, my main Tumblr, Instagram) after posting this, but I think I’ll also send it to a few other Simpsons fansites, given how effective just one article has been. Anyway, onto the project itself! I originally wanted to spend a few weeks on it, and it ended up taking roughly two and a half months. Not too bad, but still over time. This included a brief period where I lost interest in the project and drifted around between a few other works. I think this is a key event that I should learn from going forward. Whether it was because it wasn’t a super original project, or because I simply can’t stay focused on one idea for too long, I did get sick of this project about halfway through. Working on my own stuff, I’ve found that I tend to get excited by an idea and work on it solidly for a month or so, then my enthusiasm kind of dips down as the work becomes more about grunt work such as producing models, textures, and sounds. This has always been a weakness of mine (and I think, a lot of creators), which is part of the reason I started this blog, just to finish stuff, no matter how awful it is. As I’ll lay out in my next post though, I think there’s a way, or at least a possible way, to both finish projects without having to dig down and just grind through weeks of boredom. On a more technical level, there’s plenty I learned about Unity and environment design itself through this project. As well as recognizing the value of keeping asset file sizes low during production, I also discovered plenty of neat, cheap effects which added some much-needed polish to the scene, including Global Fog, water effects, and Glow.

A lot of these effects were extremely simple to implement and edit, which meant you could get the effect you were looking for quickly and easily. Most also had an ‘optimised’ version, so they didn’t take up too much processing power either. Compared to Looking Back, adding the final details to this project was pretty painless. I also learned a few simple tricks that ended up being quite effective, like rotating the skydome to give the illusion of drifting clouds, and adding the small sparkle to appear periodically on certain objects. Overall, it’s these little details that I think were the most important take-away from this project. Just adding such simple little details to a virtual environment- the lasers shooting across the floor, the clouds drifting, the sound of the water, the bouncing cone -all these incidental animations and sound add so much to the scene, and make it so much livelier. There’s such a culture in game development and in particular at University to add only what is absolutely necessary, not to exceed poly counts or textures sizes, but it’s these additions that don’t technically have to be there that really bring a game to life.

And that’s about it! I’ll see if I can tag these posts and try to organise them a bit so people can view the development of each project under different categories. It is pretty interesting going from the first post to the last and seeing how the game has developed and changed over time. In a way, those posts together are more valuable than this one in terms of reflection.  I’ll be back tomorrow with another post detailing the other projects I’m working on, and how I’ll handling this blog moving forward. Until then, thanks for following!

[thefirst200]

Final Touches

Alrighty! So it’s taken a little longer than anticipated, but the good news is, I’m almost done! I will definitely be releasing tomorrow, so anything that doesn’t get done tonight, won’t be getting done. Luckily, the environment is perfectly playable right now, so if I don’t get a chance to finish those last things off, it won’t be a big deal. This will be the last ‘update’ post, as tomorrow I’ll just put up a final wrap-up post, detailing the environment, what I learned from the process, and so on.  The last week or so has been mostly focused on visual polish, sound, and some final details. To finish up the visual effects I added a title to the game, viewable in another post below, which moves towards the camera when the player clicks. Another click puts them in the environment proper, where they start in a small black room, a doorway to the virtual landscape in front of them.

I wanted to emphasise the idea of stepping through to another space, and thought this would be a nice way to do it. When the player steps through, the walls of the room break apart and shoot out in all directions, leaving just the doorway standing by itself. I thought this would be a cool effect, and a way to encourage the player to explore away from the doorway, which is now an obvious dead-end. 

A similar idea finishes the scene- as the player approaches the edge of a particular part of the map, the walls of the room move in and form again. Both instances were achieved with a simple trigger and animation. The room at the end of the environment also features my website, a tactic I used in my last project, Looking Back. I think it’s a nice way to state definitely that you’ve reached the end of the experience, and is a good way to promote my other work, if people want to see more.  I also put the last touches on the pieces of floating text around the environment by adding a neat glow effect to them. This was another image effect hidden away in Unity 5, and I think is pretty effective. It had the added bonus of giving the lines on the ground a little bit of a glow as well, which looks great with the dark greys of the sky and fog.

Once that was done, it was on to sound. Most of the incidental sound effects, like the lasers moving past, the cone bouncing, or the water in the pool, were already in the scene in some form, so this stage was mostly just tweaking those and getting the timing right. I also replaced the default footsteps sounds with something a little more Homer-ey. For the music, my basic plan was to imitate the fantastic score from the original episode without completely ripping it off. Shhh, ye want to get sued?

I set up the scene so there was a basic ambient synth playing at all times, and different instruments would fade in and out as you get near certain landmarks, like a piano for the temple, or violins for the pool. This was relatively easy to implement, though unfortunatley I’m an awful musician so my attempts at a compelling soundtrack feel pretty flat. Instead, I stuck to the synth theme and simply added different layers of looping synth in the place of the instruments. It’s not a mind-blowing score by any means, but at least it’s not actively distracting. You can see the areas where the different pieces of music fade in and out below.

At this point, I was ready to export the project and begin optimising, having learnt my lesson from the last project. Fortunately I had already taken on a lot of those lessons during production this time around, so the environment was well under itch.io’s 200MB limit. Instead, I spent the extra time on the download page for the environment, preparing some simple logos, banners, and GIFS for screenshots. I haven’t completely mastered the art of the GIF, but they’ll do for this project.

So I think that’s about it! There’s probably a few things i forgot to mention, but I guess I’ll leave that for folks to discover when they actually play through the thing. Before launch tomorrow, I’m hoping to implement a simple pause screen that blurs the scene when you hit the ESC key, and to put in a bit of a visual glow around the edges of the screen when you walk through the doorway at the beginning. I’m not sure how likely either is at this point, but I guess we’ll see!

[thefirst200]

Almost there...

[thefirst200]

Visual Effects

This is just quick post wrapping up the visual effects for this environment. As I mentioned last time I’d hoped it would take one night, but it ended up taking two, so I can live with that. I began by finishing up the materials and textures for the primitives; thankfully a simple task. I created the marble surfaces for the giant spheres by blending some textures in Photoshop, and emphasised the matte materials with some individual spotlights for certain objects.

Once that was done, I wanted to get the distorted doorway nailed down, and I tried a few approaches, including duplicated the water from the pool, and trying to mock up a foggy, transparent texture in Photoshop. In the end I used a shader provided by Unity, Stained Glass, took out the diffuse texture, and replaced the normal map with my own design. This created a kind of fish-eye effect when looking through the plane, as can be seen below.

I’ll probably work this into the introduction to the scene (i.e: the player will have to walk through it at the beginning or something), but for now it’s a nice, cheap, visually interesting little effect. I then turned my attention to the sky, and played around with the distance settings to make sure I had a dark fog on the ground but could still see the star texture map I had also put together in Photoshop. 

To finish off the sky, I edited the clouds to become more thin and wispy, giving them a blue tint to match the clouds in the episode. I also played around with the speed at which they drift; I wanted them to move fast enough so that if you stood still you could notice them moving, but not so fast that they would be distracting during moment-to-moment exploration. You may also notice I lightened the colour of the ground, to bring out the shadows of the objects in the scene a bit more.

Before moving on to the larger task of finishing up the floating text elements, I put aside some time to get one last particle effect in there. It’s kind of subtle, but in the original episode, certain objects in the scene sparkle briefly as Homer walks among them.

I replicated this effect by creating a simple sparkle texture in Photoshop, then setting the plane to fade in, scale down and rotate, and fade out, in Unity. Like the lasers on the floor, they’re set to spawn at random times in various positions around the scene. Another simple little addition that improves the liveliness of the environment immensely. 

Finally, I finished up the floating pieces of text which dot the scene by jumping in Maya and creating some pieces of 3D text, and editing them as needed to match the font used in the episode as best I could. Then I just used a simple script to get them to float up and down, and animated their materials to fade in and out.

There’s still a couple of adjustments to be made, but for the most part that means the visuals are done! Hopefully I can get most-to-all of the sound done tomorrow, then it will just be a matter of putting together a simple menu screen, and working through the last bits of optimisation. I might miss my NYE deadline, but definitely in the home stretch now!

[thefirst200]

Visual effects about halfway done!

Tried to post a GIF yesterday but Tumblr didn’t like that for some reason. I’ll try again in a visual effect wrap-up post, but you can also see it on Twitter here.

[thefirst200]

Finishing the Pool + Fog

I actually finished these tasks about a week ago, but didn’t end up getting around to posting about them until now. Taking a quick break before I try to finish up with the misc. visual effects tonight! So it took a bit of tinkering, but I’m done with the pool, and fish within it. I started with some basic models for the fish, which I dig a rough texture job on. I then went back and refined both a little at a later date, and ended up with this:

Yeah so I’m not the greatest texture out there, and especially not with organic forms. But it’s close enough, and with the distortion of the water and a bit of bump-mapping, the detail doesn’t end up mattering too much. After this was done I focused back on the model, and gave it a basic three-bone skeleton for animation. I tried to animate within Maya but wasn’t 100% on getting that information into Unity, so I just ended up animating the bones within the engine. After about ten minutes I had a relatively convincing swimming motion, at which point I animated the fish moving to see how it looked. Rather than animating the fish itself, I animated an empty, invisible object moving slowly around the pool for each fish, and set each fish to follow their ‘leader’ object. Unfortunatley, this meant that the fish went through the walls of the pool and each other, so I ended up just adapting a script to make each fish swim in a circle; fast enough to notice movement, slow enough to not draw attention to the loop of the animation.

So the fish were all done, and I focused back on the pool itself. I just had some placeholder water in there in lieu of a proper shader, but as I went on, I decided just to edit that placeholder until I had some water that suited my goals. This was a bit of a struggle with the advanced water that comes with Unity 5, as that water is designed for a rough ocean with large waves, rather than a calm reflecting pool. 

I played around with different settings a prefabs for a while, until I eventually stumbled across the older water object, which came with the Pro version of Unity 4. It was perfect! All it took was a little playing around with the settings and refractions to get an effect that looked great, and surprisingly faithful to the original episode. As a finishing touch, I changed the sections of the texture map which would be under the water to a more appropriate blue, to imitate how real-world swimming pools are designed, and again to match the pool from the Treehouse of Horror episode as much as possible.

Finally, briefly, I also played around with the fog settings. By habit I used the settings avaliable through the Lighting menu, which weren’t as helpful as I had hoped. They created a bizarre effect where if I looked in a certain direction within the environment the scene was well-lit, but if I looked in the opposite direction, the entire scene would darken. Again, the solution lay in one of Unitys older Pro features- Global Fog. With this camera effect I was able to create, for the most part, a convincing fade into the distance, with no weird effects. The only drawback so far is that the skybox seems to disappear. The clouds remain, so it may be an issue of distance, which I’ll play around with soon. 

As I mentioned above I’m going to try and finish up the textures and visual effects tonight, with the final sound effects, music, and movement settings to be done over the next couple of days. I’m hoping to finish this one before the new year, freeing myself up for other projects in the two months I’ll have between now and going back to Uni!

[thefirst200]

Details

Okay! So obviously I haven’t updated this for a little bit, so here’s a few of the things I’ve been working on in the past couple of days. In all honesty I lost interest in the project for a couple of weeks there, either due to the fact it’s not my own original work, or because I just generally don’t work well if I have to focus on something for a long period of time. This may mean I’ll restructure this blog a little but I’ll get to that after I finish this project. So I’ve kind of been drifting between about five or six ideas in the past few weeks, which means work on this one has been sporadic. So in an effort to reignite my interest in this environment enough to finish it up I’ve just been working on a few details which I find quite a bit more interesting than just slogging through the texturing. The first thing I did was a floating kind of doorway to signify the entry point of the level, similar to the original scene in the episode.

I experimented with a few variations of this, including some perspective lines inside the doorway to imply some kind of walkway, but it ended up looking muddled and only really worked if you were looking at the object front-on, so just stuck with the original design. I want to add a kind of hazy/rippling effect to the door itself to give the idea it’s a gateway between dimensions, and add a little humming sound effect, but I’ll do both a bit later. You may also notice in the screenshot, that I have added an initial skybox.

This proved to be a lot easier than I anticipated. I played around with the default skybox tools provided by Unity and couldn’t really find one that suited. I decided to try out the classic skydome approach (create a sphere, cut it in half, add a general night sky texture to the UVs) and it was pretty successful, especially when paired with a layer or two of clouds which are set to very slowly rotate, giving the illusion of drifting. I’ll have to polish them properly, as there are a few quirks (see below), but overall, a pretty useful approach! Definitely have to remember it.

Finally, a bit of a fun one. The bouncing cone was a pretty big part of the episode, so I thought I’d start on some animating ahead of schedule and play around with that. It also turned out to be pretty simple, just a little time consuming. Right now the cone bounces alternately on its tip and base in a square pathway.

I still have to add a proper metal texture, as well as time the sound effects of the bounce correctly, but right now I’m pretty happy with it. I added a few frames of brief animation whenever the cone bounces, so it appears to settle for a split second or two, exaggerating the bounce enough for it to register while you’re watching it move around.

I’m thinking I’ll try to get the fish done today or tomorrow (it’ll be rough, 100% quality is no longer my goal) and then I can hopefully just focus on the last few details so I can get this thing out of the door.