I remember when I first discovered Line Rider, that Flash game from 2006 with the simple premise of drawing a track for a sledder to ride on. I was immediately sucked into it, doing things like devising elaborate tracks for the rider to overcome, building worlds for the rider to explore, and manipulating the rider to perform stunts.
It had an odd universal appeal, quickly propagating through the internet and reaching many other teens who were similarly captivated by this toy. We gathered into a community and the Line Rider subculture was formed. We were young, creative, and imaginative, but we also had something to prove. We wanted to make impressive tracks, whether it be with highly detailed illustrations or by exerting fine control over the rider’s movement.
In 2008, I set off to create the best track of all time, where I would demonstrate proficiency in every style of movement, create elaborate illustrations, and introduce new Line Rider ideas to the community. Of course, I was too ambitious and settled with releasing an unfinished version of the track. While it was widely praised, my vision wasn’t complete, and I continued working on it sporadically.
Eleven years later, after I reversed engineered and recreated Line Rider, after I developed as an artist and explored all types of creative mediums, I finally completed the project and even went beyond my original vision, reclaiming the project to tell a new story.
I present to you, Omniverse II.
(If you prefer to watch a video explaining the background and significance of this project, my colleague Bevibel Harvey (aka Rabid Squirrel) created a great video essay about Omniverse II)
A lot has happened throughout the process of creating this track. I break down The Making Of into four periods:
- Olympic Puppetry
- World Building
- Rethinking Line Rider
- Execution and Reclamation
Following that, I share:
- Lessons on the Creative Process
- Lessons for Line Rider Artists
Estimated Hours Breakdown
- June-July 2008: 150 hours
- 2009–2010: 30 hours
- July 2017: 90 hours
- March 2018: 1 hour (thanks Olio)
- December 2018: 10 hours
- December 2019: 120 hours
The Making Of, Part 1: Olympic Puppetry
The Line Rider community was most active in 2007–2009, what I call the Golden Age because of our collective nostalgia. We initially admired elaborately scened tracks, like that of TechDawg’s, but as we discovered bugs in the physics engine and learned to harness them in developing new tricks and styles, we slowly shifted focus onto technical proficiency of movement. We created a fun challenge for ourselves and got really into outdoing each other. It’s like if puppetry were an olympic sport. Except, instead of pulling on strings, we drew carefully placed lines. And abused physics glitches.
Before beginning the Omniverse II project, I created Omniverse in 2007, where I showcased all known techniques and styles. This began a trend of tracks featuring a variety of styles. We called it “Omnistyle”. To expand upon the olympic puppetry analogy, it was like a triathlon of Line Rider, since it also functioned to prove the skills of a track maker.
The track ends with a glitch where the rider stalls on top of invisible line extensions. Almost as an afterthought, I tacked on a story: The rider attempts to escape, but the glitches in the world prevent him from doing so, blocking him from exiting through the escape tunnel.
Omniverse II beta
In 2008, we had a big competition with 96 contestants. I fully intended on winning. Since Omniverse had good reception, I decided to make a sequel, with more intensity, more style, impressive illustrations, and music synchronization! And the rider actually escapes this time.
I made the bare track using styles like below, each section synchronized with each section of the soundtrack.
I created a flatsled recycling section, referencing the overuse section in Omniverse.
I planned out two “wow” moments featuring “monuments”, impressive standalone illustrations drawn in Line Rider. These two moments were synchronized with the music with the rider in free fall for added dramatic effect. I wanted to impress people, but I wanted to do it with style!
I was too ambitious and had to rush to finish before the deadline. I pulled an all-nighter. The soundtrack didn’t make sense with the rushed track so I had to replace the music. I didn’t get to do any of the scenery ideas. It wasn’t finished! We considered tracks to be whole when they had scenery! I called this release Omniverse II beta because I was definitely going to finish it later.
Despite all these issues, it ended up winning first place. KING (also an unscened track) won second place. Somehow, both tracks beat multiple impressive scened tracks, like Transcendental. This marked the moment we realized we shifted focus from scenery to olympic puppetry.
The Making Of, Part 2: World Building
Following the competition, I pushed forward with adding scenery. I wanted to make something completely new in the Line Rider world. So, instead of borrowing from existing Line Rider scenery styles, I searched for external sources of inspiration and looked towards the aesthetic interests I had since childhood.
When I was a child, I had access to Cartoon Network, and the intro to The Real Adventures of Jonny Quest left a lasting impression on me. I was so obsessed with those parallel lines! Inspired by that effect, I drew many doodles of parallel lines.
Along the same vein, I was a big fan of the Tron-esque cyberspace grid. In elementary school, I made a pop-up book based on it.
I spent lots of time looking at math and science diagrams, some featuring wireframes of surfaces. Being a kid, I didn’t fully understand them, but they looked cool. Yes, I understand them now.
Other math diagrams I enjoyed looking at were polyhedra. One of my favorite pastimes was creating polyhedra out of modular origami.
In Line Rider, we had the trope of XY scenery. I thought it tends to feel crammed, so I tried a more spacious approach. I’ve also enjoyed the aesthetics of printed circuit boards. In university, I designed a t-shirt based on it.
I took lots of inspiration from video games. I didn’t actually play these games, I just watched my older brother play them. One game featured this memorable scene where cyberspace mixed with real life.
I liked the idea of representing “gravity wells” (Line Rider physics bug used as a technique) as visible perturbations in the fabric (grid) of spacetime. I also wanted to subvert the trope of XY scenery by transitioning from XY to non-XY. As for the rest of the chaos, some of the blobby curves reference the popular Line Rider style of smooth cartoon-like scenery, but otherwise I wanted to create the impression of glitched out reality, like a messed up dream/video game with objects intersecting in bizarre ways.
The Making Of, Part 3: Rethinking Line Rider
Around 2010, I lost motivation to work on the project. The tools at the time were difficult and tedious to use. The community was dwindling away. From then until 2017, I focused on developing as an artist and programmer, eventually leading me to recreate Line Rider and have a fresh perspective on what’s possible.
Recreating Line Rider (several times)
I went to university for engineering. In my first programming class, for my final project, I recreated Line Rider in Python. During a summer break, I collaborated with a Line Rider friend on a new version. It didn’t get very far. I took a webapps course and created a proof-of-concept of a Line Rider track gallery.
After university, I immediately got to work on recreating Line Rider using modern web technologies. After years of development, after some more collaboration with that Line Rider friend, and after connecting with the owners of Line Rider, it wound up becoming the official version on linerider.com. This version resolved many of the issues I had with the original version, and I always kept in mind that I would use it to finish Omniverse II.
When I was a high schooler, I didn’t think much about art. Until I fell in love with this Art School Girl. Infatuated, I wanted to understand what she did, how she thought about things, how her mind works. Having the mind of an engineer, I had trouble understanding the mind of an artist. It was so weird. But there was something compelling about the subversiveness. I loved that attitude.
I eventually went to art school myself. Specifically, in university, I took classes on conceptual art and electronic media and frequented the maker spaces. Those classes greatly expanded my perspective on what electronic media could be and how it tends to be perceived. My professor asked me a question that I still think about when I see an art and tech project: “Is this an art project or a tech demo?”
In 2017, I traveled the world for a few months and ended up in Berlin taking an art and tech workshop where I learned a bit more about electronic media and the creative process.
When I returned home that summer, I resumed working on Omniverse II.
The Resurrection of Line Rider
In 2016, the remaining Line Rider community had a 48 hour long competition. I made a subversive music synchronized track, Ragdoll. To everyone’s surprise, it got a lot of attention on Reddit and was the first track to go viral in a long time. Something seemed to work about music synchronization.
In 2017, Rabid Squirrel, a long time Line Rider artist, made a long and meditative music sync track, This Will Destroy You, which eventually also got a lot of attention. Later that year, it inspired DoodleChaos, an outsider of the community, to create the well known classical music sync track, Mountain King. The viral explosion from that track was basically the singular cause of bringing Line Rider back into cultural consciousness, even inspiring knock off videos and apps.
Many new people started joining the community. Many new tracks were being created, mostly inspired by new-school music sync and old-school olympic puppetry. Some were pushing the envelope. But it wasn’t enough for me. I had so many unexecuted ideas, both old and new, that no one seemed to be exploring. I had to finish Omniverse II to get those ideas out there and demonstrate what’s possible in Line Rider. So I resumed work and eventually set myself the deadline of end of 2019, partly motivated by having this track be considered for the top ten tracks of 2019.
The Making Of, Part 4: Execution and Reclamation
I wanted to build a cohesive world, so I reused and adapted many elements from Omniverse and the first draft of Omniverse II with a lot more intentionality. With the help of a hacked-in curve tool, I expanded the grid to the entire track to create a sense of large scale structure.
I originally planned to draw intricately designed ribbons for this section, but decided it would be better to be as minimal as possible. Part of the inspiration was the iconic twisting DNA scene in Manual Mayhem X, where I wondered “Can’t we have the rider RIDE the DNA?” So I did it.
There were certain sections I couldn’t draw a ribbon over existing track since I couldn’t delete track lines. But I found a clever workaround where I aligned the twist on top of the track line.
Writing on the Wall
While finishing Omniverse II, I realized I could reclaim this feat of olympic puppetry and, instead of having the “wow” moments of monuments, I could direct focus towards the narrative. So I replaced the monuments with the writing on the wall (and the portal).
“YOU MUST ESCAPE” comes from Omniverse’s escape story and a glitch art piece of mine. A college friend sent me a photo of her bedroom with “YOU MUST ESCAPE” painted on the walls and it inspired me to add the glitch effects.
“DANGER” comes from a video by DoodleChaos. There wasn’t any real danger, but it felt like something, and I wanted people to feel something. After feedback from Rabid Squirrel, I added in “danger spikes” for real danger.
“GET BACK UP” comes from Omniverse. That and “TURN BACK”, I added because I wanted ominous directives that the rider visits twice and obeys the second time.
“THE OMNIVERSE IS A GIANT QUIRK IN THE SIMULATION” has two meanings:
- This world, “The Omniverse”, is a large glitch in the universe simulator
- This track is a very large “quirk”, the track style that heavily uses physics glitches, in the physics simulator that is Line Rider.
When I realized the final zoom-out needed something more, like a grand revelation, this idea spontaneously came to mind.
“EMBRACE THE VOID” was another spontaneous idea. It has three meanings:
- The text is fading away, embracing the void
- The rider escapes by entering the void, thus “embracing” it
- A call for more thoughtful use of emptiness and negative space (elaborated under Lessons for Line Rider Artists).
Redoing the Ending
I rushed the ending of Omniverse II beta in an effort to get the rider back up. I wasn’t happy with it, so I redid the last quarter of the track (the ribbon and portal sections).
In its place, I revisited more of the existing track, intending on making more of a visually interesting spaghetti of this area where we would see the rider follow ribbons in multiple paths. This approach was formerly known in the community as “space recycling”.
Exiting that section, I tried to create a new passage upwards. But the vertical tunnel was right there and it just seemed to make sense to reuse it. It would have been extremely difficult to do with the old tools. This technique of recycling old track was another feat of olympic puppetry. But I had new tools, and the story demanded it.
Entering the portal section, the music felt like something needed to be impactful, and I came up with the rider suddenly spiraling into a pillar of energy, accelerating faster and faster upwards.
The most impressive part of the track, this technique of overlapping frame animation has been in development in the community, starting with Bug Thief that featured these transforming grids.
For Omniverse II, I used the same principles, except coupled with a rudimentary 3d renderer to generate the polyhedra and twisting grid lines. The renderer had no z-culling and I had to manually make solid shapes correctly cover each other.
In the last part of the ending sequence, we see the lines and shapes phasing in and out, alternating between staying in place and jumping around. This phenomenon we associate with strobe lights is what I call “rational oscillations”, when you have the period of a repeating interval sweeping through rational numbers. You can see this effect in the beginning of BLOOMS and in my computer art piece RATIONALIZATION. This effect allows us to see how simpler rational numbers like 2/5 are “in phase” and more complex rational numbers like 13/43 are “out of phase”.
That was a meme from 2006. I planned that from the very beginning.
Lessons on the Creative Process
Blank Canvas Syndrome vs Laying Out All Your Cards
It’s hard to get started with a blank canvas. I struggled with this when getting back to work on Omniverse II. Yes, there were some existing pieces, but it was still largely empty.
On the other hand, when all the pieces are laid out, it’s very easy to do work and make creative decisions. It’s like having a lot of lego bricks in front of you and being free to mix and match. Many decisions felt spontaneously made but that’s because of this opportunity. A lot of the text and lore was made on the spot when I was nearly done because I saw all the pieces right there and my intuition connected the dots in interesting ways.
Blob Syndrome vs Breaking it Down
Blob syndrome: Being overwhelmed with large scope and many undefined tasks. This project had many parts to it. I spent a lot of time procrastinating because I would feel overwhelmed by how much there was to do. I couldn’t just suck it up and tackle it head on. I didn’t even know where to start.
Effective project management accommodates for the limitations of our brains. We can’t fit too much in our heads, so we have to break down the project and focus on one thing at a time. So I would tell myself things like “you will only work on this section” and “you will only work on this piece” and deliberately focus on one specific task while avoiding the thought of future tasks. And, bit by bit, the project would get done.
Perfectionism vs Iteration
We tend to want to get it right the first time. But we also tend to fear we won’t get it right. Thus, we avoid doing it at all and endlessly procrastinate. But it’s okay to make mistakes. In fact, we must make mistakes in order to find out what works best.
The effective creative process involves iterating through rough drafts since that iteration allows us to repeatedly make ideas concrete, evaluate them, combine them, and make new ideas. Refinement comes towards the end after the major creative decisions have been made.
I was stuck on building the extended grid because I wasn’t sure if I would draw the curves correctly, but overcame it by drawing rough sketches of where I wanted them to be, then refining from there. It may sound obvious, but I frequently forget that this is something I should do.
Rules vs Intuition
We create rules to give ourselves constraints. Otherwise, we’re paralyzed by the paradox of choice. Having constraints is one way of resolving the Blank Canvas Syndrome. But, by design, it can get limiting.
For certain effects, we need to resort to intuition and freestyle doodling where we can’t be bogged down by rules, explicit or hidden. Having the mind of an engineer, rules are very natural to me and I have to put in effort to use intuition instead. (for those who read The Making Of Part 3: I have to be that Art School Girl)
I was stuck on scening the chaos/warped space section. Every other section had clear rules, e.g. I could only use certain elements, whereas this section is supposed to have no rules. So I had to employ a different mindset to move forward: “Don’t think too much about it”. I didn’t think too much and doodled away and a beautiful mess came out.
Automation vs Manual Work
Programmers like me frequently have this dilemma: Should I manually do this tedious thing, or create automation to do it for me? In my case, I’m building tools that could be useful for everyone, so I do have incentive to automate as much as I can. But I had a quickly approaching deadline. I determined that it would be faster to manually draw ribbons than to figure out how to extend the curve tool to create them for me.
Then, after I finished drawing the ribbons, I found out it took less time than expected to extend the curve tool to make ribbons. I’m not sure what lesson I learned here. Maybe I tend to err on the side of pessimism as a reaction from being too optimistic in the past? Predicting the required amount of work is a generally hard problem in software engineering.
Getting Attached vs Redoing it
There are times we get too attached to what we made and are unwilling to iterate upon it. There are times we keep redoing something without making progress. There are times we accidentally lose progress, but after redoing it, we realize we did it better the second time around. If you accidentally lose progress, reassure yourself you’ll do it better the second time.
Working Outside Your Comfort Zone
When you work outside your comfort zone, you become a lot more aware of your creative process. I’m not an illustrator or story teller, but I forced myself to work in those mediums and became hyperaware of the nature of those mediums and my own processes. This is how I’m bringing you all these lessons I’ve learned. This experience will help me with my future ambitions.
When we get engrossed in a project, it’s very easy to zoom in on details and lose sight of the bigger picture and we tend to get desensitized to other details. If you’re looking to achieve a specific effect with your project, or just want to understand how others perceive it in general, the best thing you can do is to ask for critical feedback from other people who work in the same or similar medium. Their perspective is uncolored by how much you’ve already stared at your project, and their different experiences, backgrounds, and tastes can bring you really valuable insight that you may not have been able to see otherwise.
I already planned out most of Omniverse II, but right when I was finishing up the project, I knew I should verify if what I planned actually had the effect I wanted. So I solicited feedback from Rabid Squirrel, and they gave me really helpful suggestions like tweaking the camera work and adding the “danger spikes”.
Lessons for Line Rider Artists
Note: As I wrote this out, I realized these may be more suitable as standalone pieces with potential for way more depth. Consider these rough drafts.
The biggest lesson is how to tell a story. From that follows world building, lore, set creation (spatial structure), pacing, and generally being critical of everything with respect to fitting into the narrative. You may have a lot of cool ideas, but if you want to tell a coherent story, you need to make it cohesive and you’ll probably have to throw away the irrelevant parts, even if they are cool.
Unique to Line Rider is structural cohesion, how the track is spatially arranged. Consider the structure of the world you build and how Bosh’s traversal drives the narrative. Is Bosh entering a new area? Is he returning to a previous area? Did he fall and need to get back up?
Olympic Puppetry as Narrative
I wanted to demonstrate how we can use movement techniques as a means towards something greater rather than for its own sake. And the clearest way to do that is to reclaim such a feat of olympic puppetry as a compelling story, retrofitting a narrative in its place where the movement seems to emerge from how Bosh interacts with the environment he is in.
Recycling/Revisiting as Narrative
Recycling was another one of those movement techniques done for the sake of overcoming the challenge. But it can be used for narrative purposes, like being stuck in a loop or traveling through a past part of the track in the opposite direction to “turn back”. I think there’s more narrative depth that could be done with revisiting, much more than in Omniverse II, perhaps in a track featuring a more intricate story.
Embrace the Void: Negative Space
In a composition, negative space is the absence of content, contrasting with the content that’s there. While negative space is already commonly used in Line Rider tracks, I think it’s still worth discussing. Negative space in Line Rider can be in the form of the white void (absence of lines) or as airtime (absence of movement). There are obvious uses like dramatic moments in the music, but we should also consider more subtle “less is more” cases, like bringing attention to an object by removing details around the object.
Embrace the Void: Fade Out
In the first draft of Omniverse II, I encountered the issue of scenery extending beyond its designated area. I could have just done a hard cut off, but I came up with the more interesting approach of having the scenery fade out, like it’s receding into a white fog. I later employed the fade out to transition between areas overlaid on top of each other and to create the effect of scenery “behind” something. The fade out and white fog lends itself well to the aesthetics of Line Rider and I’d like to see more uses of it.
Embrace the Void: Sparsity vs Density
In the past, track makers would have the goal of making impressive scenery by making it really dense with details. Yes, it’s impressive, but is it tasteful? Removing the details would make the scenery less impressive, technically, but sparser scenery lends itself to the aesthetics of Line Rider, brings more focus to the rider in relation to the environment (it’s hard to see the rider in dense scenery!), and also involves a lot less needless grind. And, when there’s high speeds or large zooms involved, so much of the details get lost anyway! There are cases where density is warranted, but, again, less is more, and be judicious about if it’s really necessary.
Camera work fulfills several functions. While we understand that the camera’s primary function is to keep the rider in view, we can also interpret it as moving the environment around the screen. In a way, the camera frames both the rider and the environment.
One obvious way camera work is used in Omniverse II are the dramatic zoom-outs showing the large scale structure of the world, which also frames the narrative. The other obvious way is the music-synchronized zooming. More subtle ways camera work was used was in framing the scenery, like the constrained movement in the XY section and the relative stillness of the camera in the chaos section (also relating to the above point of how density should involve less camera movement).
The camera currently only has two parameters: zoom level and bounding box. There’s also the higher level parameter of smoothness, used to add easing to changes in parameters. You generally want to default to using easing curves. At some point, I will add in controls to directly control camera position irrespective of the rider, but these parameters were enough to achieve the effects I wanted in Omniverse II.
The camera is relatively simple compared to what goes into designing the movement and scenery, but it still has strong effects on how the track is literally framed, and it’s something we should put some thought into. In the case of Omniverse II, the mindset I adopted was “how do I make this into an action movie”. From there, the creative decisions followed.
Illusion of 3D
I wanted to explore ways of faking 3D using geometric effects (vs textural effects based on dense shading). Creating the illusion of 3d on 2d canvases is a centuries old problem with many creative solutions. As of now, Line Rider is limited to an unchanging flat 2d canvas, which is similar to traditional media, except the camera pans and zooms around the canvas, which complicates things.
With a traditional canvas, like a painting on a wall, we can see all of the image at once, and we effectively see it from only one perspective. If the image uses realistic perspective projection, it can be considered equivalent to a camera in one position (perspective) taking a photo of a scene, rendering it onto a 2d photograph. When the perspective is mixed up and “improperly done”, perhaps in a child’s doodle as naive perspective, or intentionally broken in cubist art.
The problem with perspective in Line Rider is that the camera position changes all the time (usually), so using that technique requires carefully mixing up the perspective and designing what the viewer will pay attention to and what they will ignore, kind of like the way magicians employ misdirection to keep the audience’s attention away from how they’re actually achieving their magic.
Images can also use orthographic projection (isometric for 45˚ viewing angle), where 3d objects are uniformly flattened onto a 2d plane. The benefit is that perspective remains the same no matter where the camera is panned. This is commonly used for RTS or RPG games using an overview camera, since the camera needs to move around the world.
Orthographic projection suits Line Rider very well as well, though there are only a few examples like Urban Run. However, orthographic projection is prone to confusing optical illusions, and care needs to be taken to avoid situations where the rider appears to break perspective, unless that’s what you’re going for.
Depth management, depth and visibility: Creating the illusion of depth is simpler: Just have objects cover each other. When objects are kept near each other in the Z direction (they’re all similarly far away from the camera), the parallax effect can be factored out and there’s more freedom to use whatever perspective you want, provided it’s similar enough to orthographic. With objects close to the camera, it’s a challenge with no good solutions right now, since we expect objects near us to move out of view quickly. With objects far from the camera, we can employ effects of distance, like reducing size, reducing detail, or using the fade-out effect, whatever it takes to make it appear to be in the background.
The most effective way of making objects appear 3d is through 3d movement. This was achieved with the animation at the end. In a way, that tells an alternate story of Bosh going through attempts at 3d, going through a portal of 3d, and perhaps ultimately entering the third dimension.
Omniverse II 3d details: Spheres are orthographic (easier to draw, acceptable to our eyes), polyhedrons and flatsled recycling manifold are perspective (these shapes’ 3d qualities are enhanced by perspective projection). Shapes in perspective far away enough appear orthogonal. The larger scale environment has two camera positions: Top (so we’re looking up towards the ceiling and down towards the tunnel), and bottom (we’re looking up towards the tunnel).
I will never spend so much time on a Line Rider track ever again.
But I will continue doing everything I can to keep pushing the expressive possibilities of Line Rider, sometimes in the form of tracks, but mostly in building and maintaining Line Rider. I want to build actual UI for controlling the camera. Maybe I’ll add in parallax layers for true 3d. Whatever the case, you can help by supporting Line Rider!
Follow me on Twitter for more updates on Line Rider: https://twitter.com/d_e_lu