Soaring high: The making of Glyder
How Glu spreads its iPhone wings
Letting gamers play at their own pace, it sees you gliding around a fantasy world in the character of steampunk pilot Eryn, collecting glowing crystals in a traditional exploration style.
Yet behind the pastel colours and smoothing music, the 3D open world environment is technically impressive: even more so when you consider it was one of the first internal iPhone projects from mobile publisher Glu's San Mateo studio. And completed in 10 weeks.
Executive producer and studio head Daren Chencinski explains how the project was conceived and the obstacles overcome along the way.
Pocket Gamer: Where did the inspiration for Glyder come from?
Daren Chencinski: We can't exactly pinpoint the original point of inspiration. However, we do know it started in a room where we were all brainstorming on some lightweight but expansive game mechanics ranging from re-used/re-purposed to completely unique mechanics.
There were a handful of our producers and designers who were really excited about what we could do with iPhone in general - bear in mind that although we have console/PC gaming backgrounds, we've been doing mobile for a while, so this new platform was just so juicy. Anyway, we had a bunch of good ideas but we all got very excited about the hang gliding idea for a variety of reasons.
First and foremost, we were moved by the idea of a game that was peaceful and beautiful in nature while death-defying at the same time. Many of us had done a variety of aeronautical sports before, ranging from paragliding to skydiving. In these sports you are a witness to the vast beauty of the landscape, while having some degree of fear of injury or death. This is something we wanted to capture in this game. In short, we loved the idea of open exploration and visual discovery through flight and this became our goal. But I'm getting ahead of myself...
The Lost Kingdom environment
Second, we agreed that people loved tilt controls on iPhone. We kept seeing these tilt control games that were good, but not great and people loved these games and were consistently buying them. We figured that a game that used tilt controls and felt as accurate as its real world counterpart would be something we would be happy with and would be received well.
How difficult was the approval process within Glu?
It was different from usual. Alex (head of R&D), Ronald (head of Glu's studios), and I had been trying to create the opportunity to give teams creative freedom to build a game without much external influence.
The idea was to keep the development cycle short - somewhere around eight weeks - while staffing the team fully so as to have high production values. The people in that room that day were picked for a reason: they weren't only helping to come up with concepts, they were also going to approve the concept.
We all had to be excited about it or we weren't doing it. We're all gamers and we've all designed or produced successful games. Also, we're all avid iPhone users. This group came up with and approved the concept within one relatively short meeting. I pitched it quickly to Jill (head of publishing) a couple of days later just to make sure she was cool with it and she approved it too (frankly we already started it by then!). That was it - done.
The 3D nature of the game must have been discussed at length?
The game had to be 3D and we knew that from the start. Otherwise the world wouldn't have felt as rich as it did.
Naturally, it made sense to use our in-house 3D engine. It handles 3D geometry and animations efficiently, straight out of the box. From here we added the custom physics/flight dynamics, collisions, object culling, etc. We've refined these techniques and skills over many project cycles, so we could get this development cycle done pretty quickly.
Since this was our first extensive project on the iPhone, the first steps were to see what performance we could get out of it for this style of game. Ray casting for collisions was one of our biggest hurdles in regards to performance. One of the techniques we employed was to have a simplified (and invisible) version of the terrain for collision purposes while rendering a more complex one.
The biggest obstacle to the game had a relatively simple solution. We wanted the player to explore and discover, which wouldn't be compelling without a decent-sized world. The world had to be seamless too, so there couldn't be any loading while flying between worlds.
Glyder's Rift Valley environment
On startup we load all the environments into memory, with each area having a high resolution and a low resolution version. The low res versions can be much simpler and at a distance they look much the same. When we are in one area, all others can be the simple low res version, then on proximity we switch out the hi and low res, and naturally we don't need to perform collision analysis and run animations of the low res versions. The addition of fog smoothed out these transitions even further. This is also why we wanted to go with islands so they could be separated and hidden and shown fairly seamlessly.
For objects like the crystals and thermals, we kept a bank of objects and recycled these by placing them in the positions closest to the camera. In other words, there may be seemingly hundreds of crystals, but in reality there may only be a handful of actual 3D objects.
How did the fantasy setting and the character of Eryn come about?
The fantasy setting was an obvious one for us and was agreed upon in that first meeting. We wanted the world to be colourful and interesting. We wanted the player to experience a world that was familiar but extraordinary. Also, let's face it, you're dealing with a bunch of dorks so fantasy flies really easily with us.
The player character wasn't as obvious. When we started talking about the character, it was clear to us that s/he couldn't be flying around with a regular hang glider. We quickly got into the idea of this character fashioning his/her own flying device. This naturally snowballed into a steampunk and Leonardo da Vinci-inspired character that would be a cool inventor type. We got into this idea of an old inventor guy who was going to be so dope with his inventor goggles and gadgetry on his belt.
Progressions of Eryn's character
But then we shifted our thinking to go with a strong, independent female. She had to be smart, accomplished, resourceful, and just plain intelligent to create these cool wings. She had to have purpose and determination, and she would be willing to mix it up if push came to shove.
We also felt that a female character allowed us to take a unique, softer approach to steampunk. It's a relatively subtle but vitally important point: it just made our character fit in the world we were envisioning.
As the game progressed, we evaluated our worlds. Brian [Collins] had created some promising first drafts for six of them. He was just rapidly getting first passes out of the way. They were great, but they had to evolve. At that time, Ambrosia and The Mushrooms didn't even exist. There was a cloud world, which was cool but ended up looking like a giant marshmallow, so that ended up becoming The Mushrooms' gear satellite. For all the other worlds, we wanted to take Brian's initial concepts and make them even more extraordinary.
As an example, Rift Valley always was below sea level. That was the extraordinary thing we planned from the start. We wanted the player to start below the walls and then discover there was a bigger world to explore. We made Rift Valley more fantastic by making the walls even higher and adding some of the semi-believable rock formations (specifically the rock ring and balancing rock platform). Also, the volcano was increased in size and made more dramatic.
Revamp work on the Dark Pinnacle environment
We took similar steps with Dark Pinnacle and the other worlds. Dark Pinnacle underwent the biggest changes. We doubled it in size and added the tunnels along with the spiky area on the backside. Along with all these changes we undertook a massive colour revamp which we found to be a lot more interesting and appealing.
The game has an 'at your own pace' feel to it, which is in contrast to the competitive nature of many games. How much do you think this was a strength and/or a weakness?
This was definitely as much of a strength as it was a weakness. Even internally, people had very strong and divided reactions to the game's relatively passive approach. The App Store reviews are still all over the map partially because some people aren't getting what they want, which is fine - everyone has his or her preference.
Some people wanted enemies to fight or avoid. To us, the obstacle was death itself.
Other people wanted more direction. To us, it was fun to explore the world and find stuff to do. We thought it was fun to find all the crystals of one colour and see what happened as a result.
Then there was the other side of the spectrum. Some people wanted it even more open-ended. They wanted to be able to land anywhere in the world. To us, the game was really about flying and exploring, not necessarily adding in an elaborate landing scheme.
In any case there were clear advantages to our design approach, which were intentional. We wanted to let the player decide how they wanted to play the game and at what pace. Fans of the game really enjoyed that. Lots of people found all the crystals in an area before moving onto the next. Others loved that they discovered the gear satellite in the first five minutes of the game.
In retrospect, we agree we could have engrossed the player a bit more quickly had we included a fourth set of crystals that popped open Dark Pinnacle, Ambrosia, and The Mushrooms.
This crystal set would have all been contained within Rift Valley and there would have been very few of them so every player would have gotten them quickly. This would have illustrated the point of the game a lot earlier and we probably wouldn't have confused as many people as we did. In any case, some people don't like these types of games and we're okay with entertaining those who do. At the same time, we know there are ways to capture those who are on the fence and will aim to do better at snagging their attention.
Controls are clearly a big feature for a game like Glyder so how did these develop?
At first pass, the controls took us about a week for concept and implementation; Chris [Hosking] and Brian were able to collaborate perfectly to get it working. One of the tricky parts was deciding on how much drag to apply to Eryn. We also wanted to make sure the player could effectively dolphin (climb without thermals). And finally, we definitely spent a solid day or two on getting the camera right. The first camera was downright odd but that was gone quickly.
Over time, we refined the controls and camera further as the game world and rules evolved. We do remember, in particular, that getting the thermals to be effective was a bit tricky. At first, in order to obtain the maximum lift in a thermal, Eryn's wings had to be perpendicular to the updraft. However, as the player tilted upwards in a thermal, the lift and speed would decrease. This was simply confusing and unfriendly to players. We had to add some rules like 'Eryn can't lose speed in thermals' to get thermals to be fun and rewarding.
Did you ever think about a touch control option?
We never considered an alternate control scheme. It didn't seem necessary to us. We were more interested in tilt calibration and sensitivity options.
We did very seriously consider this idea of a power drop. The player would tap the screen to activate it and then tap again to turn it off (or just hold and release - whatever). The wings would fold in and Eryn would shoot straight down, gaining insane speed. She could use this to drop into otherwise inaccessible areas. Also, this would make dolphining more effective.
Can you give a rough breakdown of the development time and any significant tools you used?
Total development time was ten weeks. This included pre-production which was a week.
The only reason this short timeline was possible was because we used our existing, proprietary 3D tools that operate within 3ds Max. This allows us to build the environments and see them exactly as they will look on the device. This gives us the ability to develop the art and integrate it extremely quickly.
Redesign of the Rift Valley environment
The game was designed in a pretty logical manner. All areas of the world have objects such as thermals, crystal groups, achievement triggers, specific collision regions (e.g. water), and animations named the same. Once the logic was set up and refined for the initial areas, new areas became relatively simple for Chris and Brian to add on.
What were the biggest obstacles you had to overcome?
From the start, we were concerned we wouldn't have time to create full detailed worlds given our short development cycle.
To solve this, we chose a simple, sometimes abstract, art style that would lend itself to rapid development. We did reach a point however, where we thought the environments looked good, but we still weren't totally happy with their quality. The iPhone is a powerful device, but it's still not a console or gaming PC.
With the addition of some simple rendering effects, like distance fog and the high speed effect, we were able to imply detail and make the scenes look less like "my first 3D app" and more like a game.
Maybe it's obvious, but we really wanted Eryn to have a shadow to give the player another tool to judge her altitude, but what we ended with was a shadow that you only see moments before you crash. Querying the terrain to generate an accurate shadow is actually pretty computationally expensive, so more time may have allowed us to come up with a more ingenious solution, but we had bigger fish to fry. It turns out that not everyone cared about this anyway. Still, that doesn't change the fact that we think the experience would be improved with a proper shadow.
At what point did you realise you had come up with something that exceeded your initial inspiration?
Brian and Chris passed around a build after the first week when they had the first pass of the controls in. We knew we had something cool going on at that time since it was just fun to fly around in a single world (which was the first pass of Rift Valley) with nothing else to do.
There was certainly a point at which we weren't thrilled, which is when we sat down and figured out what we were missing.
The Mushrooms environment
That's how we went from eight to ten weeks. We used those extra two weeks to get in the final portal, revise the worlds (including crystal placements, achievement additions, and trip modifications), add The Mushrooms (while trashing the marshmallow cloud!), and optimising the game so our full screen effects worked well. We also added the boost crystals, which we all agree was a great addition to the product.
Were there any features or elements that didn't quite work out or make it into the game?
Sure, there were tons of things we didn't get in that we wanted to! Aren't there always?
We've already discussed the power drop, cloud world, and shadows. Really though, we're not sure the power drop is really something we're even interested in any longer. The cloud world was a cool idea but wasn't going to work with our artistic approach. The shadow is definitely something we wish we had gotten in.
The trips ended up being simple point-to-point goals that were ultimately too simplistic to capture the feel we were looking for. Don't get me wrong, we like a bunch of them that are in there. However, the system we originally designed was intended to include waypoints to make the trips more interesting and challenging. The lack of waypoints is one of the reasons that we sorely wanted to include the boost crystals so we could use them as semi-waypoints and build climbing trips.
We even wanted to build trips that spanned multiple worlds. We think big fans of the game would have really enjoyed these, particularly with the inclusion of the waypoint system.
What part of Glyder are you most happy with?
We're happy that people were engaged enough to explore the worlds. Like we discussed, the game didn't have a ton of specific direction. We are glad that the simple thrill of flight coupled with compelling worlds to explore was enough to motivate players to engage with the game.
To that end, we do feel that we captured the experience of flying pretty well. This turned some players off, of course, because all they wanted to do was point in a direction and go there. However, those who enjoyed the challenge of navigating a semi-realistic hang gliding experience really appreciated the thrill of gliding. It felt good to reach a thermal in the nick of time just before crashing to the ground or flying close enough to a crystal to just barely get it to attract to Eryn before maneuvering out of the way of impending death. Players felt rewarded when they climbed high enough to dive and hit terminal velocity or try to run the lava tube achievement.
Flying insider the Dark Pinnacles environment
We're also glad that Glyder fans seem to care about Eryn even though she's fully clothed and you're mainly looking at the soles of her boots. We could have highlighted her a bit more in hindsight, but we are glad that players like our character even though she isn't scantily clad.
Were there any particular surprises you had in terms of player feedback?
The 1.0 version of Glyder didn't have calibration, which we wanted to put in the game but didn't have time. However, we were surprised at how visceral the reaction was by some fans about not having it, particularly since we had two axis of accelerometer controls. We wanted to take the time to do the calibration right and did so with our 1.1 release, so now you can play it lying on your back if you want.
Significantly more surprising is that many people didn't realise you could leave Rift Valley at any point. We felt a bit bad at first when we'd get emails asking, "How do I get to the other worlds?" We would respond by basically saying, "If you can see it, fly there!" We felt a lot better when those same people responded and said, "Oh wow, I can't believe that - that's cool!" The open world was a tough feature to keep in given the added development time to support it, but we think it paid off for the majority of players.
On the bright side, we were pleasantly surprised by the volume of positive feedback emails we received. It was the first time we were able to really see the extent that players were engaging with our games, even asking us for tips on how to find the one or two remaining crystals to help them finish the game.
How did you deal with the update in terms of what features to add/tweak?
We paid close attention to the user reviews, emails, and boards that talked about Glyder. This made it an easy decision to add calibration first. The following updates have focused on crystals, which provide core Glyder users new experiences in the same world. We also added the Facebook Connect component so players could tell their friends about their latest Glyder adventure - this will be released shortly.
Are there any future plans for Eryn and Glyder?
Yes! We are very excited to announce that Eryn will be starring in the sequel to Glyder, Glyder 2, which will be available later this year. We're looking for some Glyder fans to help us with the development process. If you'd like to be involved, please send an email to iheartglyder [at] glu.com.
Thanks to Daren for his time.
The full Glyder team consisted of;
producer: Daren Chencinski
designer and artist: Brian Collins
engineer: Chris Hosking
artist: Phil Barrows
QA: Dan Clukey and Marlan Smith
music and sound: Yichel Chan and Qi Yi
and special thanks to: Dan Sciarrino, Chad Caruso, Corey Annis, Chris Handloser, SeungWon Park, Charles Barnard, Michael Steliaros, James Wysynski, Jay Brown, David Brittain, Tor Hagen, and Stephen Ardron.
Some of the Glyder team, from top left clockwise:
Phil Barrows, Marlan Smith, Dan Clukey, Chris Hosking, Brian Collins
Glyder is available from the App Store priced $1.99 or £1.19
And check out our other Making of interviews such as
Space Invaders Infinity Gene
Toy Bot Diaries