Developer Interview: Image & Form on SteamWorld Dig, 3DS Development and Working With Nintendo

"3DS gamers are generally more experienced than mobile gamers"

SteamWorld: Tower Defense arrived on DSiWare back in the summer of 2010, and represented the Nintendo debut of Swedish developer Image & Form. Its stylish approach was well received critically, and despite its brevity represented a solid offering worth consideration on the last-gen handheld's download store. Image & Form is bringing that game world back in upcoming 3DS eShop title SteamWorld Dig, but this time in the form of an action platformer, where mining is the name of the game. The developer plans to submit SteamWorld Dig to Nintendo for approval in mid-May.

Image & Form is a studio with far more behind it than a DSiWare title and upcoming 3DS release, however, with an extensive history working on PC and smartphone games. We've spoken to CEO Brjann Sigurgeirsson about the developer's history so far, as well as his views on working with Nintendo and on the 3DS eShop.


Nintendo Life: First of all, can you give us a brief history of Image & Form?

Brjann Sigurgeirsson: Image & Form today consists of 11 people working full-time: 1 CEO/marketing guy (me), 1 project manager, 4 graphic artists and 5 programmers. Wonderful, creative, hard-working people, the lot of them.

I founded Image & Form in 1997, and we started out as a 2-person multimedia and web company. The reason for founding the company was that I had developed multimedia experiences while living in Tokyo (90-95) and San Francisco (95-97), and when I got back to Gothenburg I applied for a few jobs - and got them all. So I decided there must be a market for the things we did, and started Image & Form instead. We went from multimedia, web and small games to web development on a bigger scale. In 2002 we were asked to "save" a capsizing edutainment game project for a Norwegian publisher, and then got asked to do games from scratch for them. It still felt like a side business until 2007, when we were asked to increase the production speed from one game in 18 months to 8 games per year... then we didn't have time for anything else!

In late 2011 I sold half of the company to Bergsala, a game-industry behemoth by Scandinavian standards. Amongst other things, Bergsala is - and has been for more than 30 years - the Scandinavian agent for all Nintendo products. Image & Form is the Bergsala Group's development arm.

NL: You've previously worked on a number of PC titles, so how have you come to work in the handheld space?

Sigurgeirsson: Simply put, it was a mix of creative starvation, necessity and opportunity knocking. We created some 30 (!) unique full-length edutainment titles in the same franchise from 2007 until early 2010, and then both our publisher and we decided that enough was enough. We had a lot of "grownup" game ideas and were creatively starved, and we had also slid to the rear of the pack by using the same old framework for the same old platforms (CD-ROM games for Mac/PC). So we thought that in order to survive, we would have to come up with good game ideas, and develop own IPs for new platforms. A huge leap out of the comfort zone, going from the safe-yet-fragile role as a work-for-hire developer to a self-publisher. Would we make it in this new role?

Late 2009 we decided to pull ahead, and started developing our first iOS title, Gyro the Sheepdog. It's a nice little game, but we really were rookies on the iOS scene - as were many others at the time. We didn't know at all how to spread the word about the game, what we could expect in terms of ROI, or even how we should limit the scope of the game. We decided to release it just to learn the mechanics of the App Store, reviews etc, but generally it was an experiment. (Or a failure, depending on your mood.)

By the end of 2010 we were awarded a grant from the Nordic Game Program for a game idea that we called Anthill. Since we got the money we felt obliged to produce a game, and to be honest our initial thoughts - for which we got a grant - were sketchy, to say the least. We really had no idea how Anthill would work, and reading the game pitch now is a great source for laughs. They must have seen something that definitely wasn't there.

We continued to do work for hire, among others a game called Mariachi Hero, which was quite cool but got terrible reviews in the US. The game was set in the Mexican desert, where the protagonist was either a Mexican-looking man or woman. The game graphics were very cartoonish, and so were "our heroes". We had no other intentions than to make a fun, cartoonish-looking game, but some US reviewers decided that we were poking fun at Mexicans, and labeled the game "incredibly racist". The shock and horror...! I guess there must be a collective bad conscience swimming about in the US, and we ended up paying for stereotypes like Speedy Gonzales and the "ugly" guy from "The good, the bad and the ugly"! It ended up being pulled off the App Store, but we may redraw and re-release sometime.

After that we concentrated fiercely on Anthill, and a while into the process we realized it would be a stunner of a game. We were still marketing rookies, almost completely unknown, didn't really know where to turn to get noticed, and in desperation - a month prior to launch - decided to contact Apple and really flaunt the game. We were surprised and lucky: they thought Anthill was awesome, and decided to feature it as iPad Game of the Week - which was a huge break for us when we released it in October 2011.

Since then we've developed an iOS game called Hugo Troll Race, a 3D horizontal runner in the Temple Run vein. That was probably the last work-for-hire game we'll ever do. And now we're focusing all our efforts on SteamWorld Dig for the Nintendo 3DS.

To summarize the departure from PC: there's nothing wrong with making PC games. There is a constant stream of great games, both big and small, being released for the PC platform - and with Steam, GOG and others, there are now simple ways to distribute and sell these games. But at the time we were making work-for-hire games for a publisher, and they were only distributed physically and locally in Scandinavia. We may well port our own games to PC in the future. I just need to find a relevant e-mail address to someone in charge at Steam!

NL: You've been working on iOS and DS, yet SteamWorld Dig is currently only listed for 3DS. Why did you choose Nintendo's handheld for this title?

Sigurgeirsson: For a number of reasons. In 2010 we developed SteamWorld Tower Defense for Nintendo DSiWare. In comparison to the congestion and thunderous white noise of the App Store, the third-party DS-title ocean seemed very blue - and so we believed it possible to make money off a game on the DSiWare store. The DSiWare storefront itself was terrible, but we had a couple of good reviews, most notably from IGN who decided that the game was "Great", and the game paid for itself pretty soon after launch. It was a great experience, and later we "knew" that we could try developing for the 3DS and the eShop when it came along.

Today, we have a pipeline of around 15 great game ideas that we want to realize at some point. SteamWorld Dig, which we call a "platform mining adventure", is one of them, and we really want to make a game for the 3DS while it's Nintendo's state-of-the-art handheld platform. The ocean is still blue, and the communities are active. 3DS gamers are dedicated to the platform, easier to reach and - at least we think so - less finicky than mobile gamers. They'll pick up a good game if they see one.

We think that SteamWorld Dig suits the 3DS very well. For one, it has the controls of a platform game. As such it's definitely a D-pad game, which more or less rules out touch devices, such as mobiles and tablets - I have yet to see a really good implementation of a virtual joystick. Also, the no-clutter split screen is perfect for the game: the game action shows on the upper screen, and the stats/map/buttons are displayed on the lower. And finally, it's a gamer's game: the player decides how long the game sessions will be, since the gameplay is continuous from beginning to end, not divided up in mobile-like, "coin-op" levels.

So the 3DS was the obvious choice. If you're a mobile-only developer, switching to the 3DS may seem daunting and you decide not to investigate the opportunities. But we've done it before, so it's not a hurdle.

NL: You describe the game as a "hardcore platforming adventure" in some marketing materials, but can you tell us more about some of its key gameplay mechanics?

Sigurgeirsson: By "platform mining adventure", we mean that the controls are like a platform game, the objective is mining, and the emergent gameplay makes it an adventure. We believe that mixing game genres makes for a deeper game experience - a lesson learned from the success of Anthill, which was a mix of castle defense, real-time strategy, resource management and line drawing. In SteamWorld Dig, we aim to mix platform play with puzzles and emergent elements. We've tossed in "hardcore" in the description, because we're certain that the game will appeal more to experienced gamers than casual players like, say, my mother.

The alternate-reality (and largely untold) background is this: Humanity made a great technology leap around the turn of the 20th century, and among others managed to create steam-driven robots who could replace the working class. Not long after, humankind self-destructed in a great war, and the remaining quarrelsome, desperate low-lives - the very bottom of the barrel - live and huddle underground. These people are neither role models nor heroes. In fact, the robots are the good guys - hard-working, everyday "people" who strive to make things work.

Our protagonist Rusty, a lone robot (or "cowbot", as we call him) comes to a small Western mining town called Tumbleton, after having received a deed for a mine from his long-forgotten uncle. Although the inhabitants are friendly and helpful, there's something very wrong with the town. And Rusty will find out what it is - towards the end of the game.

Rusty is your token lone Western cowbot hero, who doesn't really like to talk too much, and definitely doesn't complain about the workload. He's fearless, handsome in a robot-ish way, and has an adequately-sized coal furnace for a belly. He's definitely one of the good guys, and perhaps the robotic equivalent of Clint Eastwood's character in Unforgiven. His objective is to get the town back on its feet, and helping the townfolk in the process. Rusty isn't prone to violence unless it seeks him out. He really has nothing against humans, but the despicable remnants of humanity hold a grudge against robots. Bad mistake and their loss (we hope). He's not that curious, but here is a real mystery. And someone's got to solve it.

The objective of the game is to mine, sell, upgrade equipment, watch the town recover - and dig ever deeper to find what evil lurks below. There are a lot of quirky mechanics that help you navigate the underground, things that may seem obvious to a robot but which require some mastery for humans. All treasure hidden underground needs to be converted into cold, hard cash before you can buy things with them. A steam-driven robot has use for other things than humans - coal for light and energy, and water to produce steam. Luckily, this can be had - as long as one has something to trade for it.

3DS gamers are generally more experienced, "better" gamers than mobile gamers, and they don't deserve a torrent of semi-crappy games on the eShop. Quality before quantity, please.

NL: Can you tell us more about this game's structure? Is it linear, are there side-quests to take on?

Sigurgeirsson: SteamWorld Dig is linear in the sense that it has a beginning, main objective and an end, but there are side-quests and areas that you may miss, and want to replay in order to complete.

NL: Would you say it's a fairly lengthy campaign, or is it designed with replayability in mind?

Sigurgeirsson: Yes, it's fairly lengthy. We're looking at 6-8 hours of continuous gameplay from beginning to end, but I believe people will enjoy it in smaller chunks, shorter sessions here and there. Or maybe that's just me - damaged goods from too much mobile gaming.

But the maps are generated dynamically, which means that no two game sessions will be the same. Locations of objects, exits, shortcuts, exceptional treasure etc are randomized within parameters, so the game can be replayed - should be replayed - in order to be fully explored. It will also be fun to replay SteamWorld Dig to simply try different upgrade tactics and digging strategies.

NL: The screenshots (though at a higher-than-native resolution) look impressive for a 3DS download game. Have you been using a custom engine for this, or a specific pre-existing engine?

Sigurgeirsson: It's our own engine, and the production pipeline is quite elaborate. For graphics, we draw everything in Adobe Illustrator, then we export the graphics to After Effects for sprite animation. All frames are rendered and then ported to Flash for sequencing. We like to work in vector format, since it makes it easier to port to all kinds of resolutions. Besides, it's vital for our typical cartoonish style. Programming-wise, SteamWorld Dig is obviously programmed in C++, but then again we use our own C++ engine for all our game projects. For example, for iOS games we simply use an Objective C wrapper.

NL: What features of the 3DS were the most important factors in you choosing the system, and which are most important to the game's experience?

Sigurgeirsson: We chose the 3DS because of the portability, the D-pad and the split screen. As I mentioned above, the game mechanics are very well suited to the 3DS, so it all works together. We have basically made the game FOR the 3DS, although it can ported to other D-pad-driven platforms later.

With that said, the stereoscopic effect - a defining trait of the 3DS - is of little or no importance to the gameplay. However, background layers are revealed as you chip away, which adds beauty to your mining experience. A few weeks into development we looked at 2D platformer Mighty Switch Force and thought that WayForward had put the 3D effect to good use in a similar manner. Just like MSF, SteamWorld Dig looks just as good in 2D. We've put a lot of love into the sprites and graphics, and don't want to rely on an effect to get the game across or to "score extra points".

We've been asked about stereoscopic 3D a lot, since we got cocky somewhere and said that SteamWorld Dig is "not another 3D effect showcase". We meant that gameplay and experience are key, not platform feature exploration: we don't add extra depth or microphone-blowing just because we can. If it's not vital to the gameplay, then leave it out before it gets annoying. Unnecessary feature exploration is comparable to product placement in movies - it's there for a reason, but is the reason right enough?

NL: How have you found the experience of working with Nintendo so far?

Sigurgeirsson: Working with Nintendo on SteamWorld Dig has been very nice so far. They promptly answer our questions, and are both professional and personal. I do think they've evolved, and their tools and processes are easier to work with these days. Obviously we are also better at asking for their help than before. When making SteamWorld Tower Defense, we were very self-conscious and felt there was a huge difference between tiny, insignificant Image & Form and enormous, all-important Nintendo. But naturally they're friendly people, who do all they can to help. And we have grown more confident.

NL: Can you share your thoughts on the eShop, in terms of working on it as a developer and your opinion on the library itself?

Since we've developed both for the App Store and the DSiWare Store before, I'd say that the eShop both looks and feels good. First, it's a world of difference from the DSiWare Store, which felt inaccessible, hard to navigate and "alphabetically unfair". (We actually considered calling SteamWorld Tower Defense something silly just to appear on the first page, such as "AAAAAAA Tower Defense".)

The eShop is a mature marketplace, with good opportunities to showcase your games and set the price points dynamically. Also, you can show a video clip directly in the store, which you can't on the App Store - a great feature. It gives the audience a better chance to understand a game before buying it.

The library itself is relatively small, which is both good and bad for developers. Your game becomes quite visible, but you may hesitate: Why are there so few titles? Will my game stick out? Will it be scrutinized harder, will it hold up?

Of course I want more titles on the eShop, but not for the sake of numbers. It's good that developers think twice before making 3DS games. 3DS gamers are generally more experienced, "better" gamers than mobile gamers, and they don't deserve a torrent of semi-crappy games on the eShop. Quality before quantity, please.

NL: Do you have any future plans for work on 3DS?

Sigurgeirsson: Yes, we're planning to expand the SteamWorld universe in titles to come. When we made SteamWorld TD, we felt that we only scratched the surface on this alternate-history, twisted world where humans are the thieving, dynamite-wielding bad guys, where the robots needed to defend their mines. It was enough for a decent background story, but it was never told. However, the game mechanics are obviously totally different in SteamWorld Dig, so it's not a real sequel, but it's a continuation of the same "what-if" idea. At the same time, we don't want to reveal the entire story and universe in SteamWorld Dig. We've got more plans for SteamWorld in the future. New games, with new and very different game mechanics.

NL: What do you think of the Wii U, and do you have plans to work on the platform?

The Wii U is a lot more powerful than the general public thinks, but it hasn't been communicated successfully. It's a hard message to compose, since the console looks just as playful as the Wii, and the cute appearance of the Miiverse adds to this notion. However, for a developer it's interesting. The ocean is extremely blue there, and there are lots of opportunities. SteamWorld Dig would make for a fairly straightforward port, since the controls are the same.


We'd like to thank Brjann Sigurgeirsson for his time. What do you think of his comments, and are you interested in SteamWorld Dig? You can see the most recent beta footage below.