Tango Gameworks, founded in 2010 and located in Tokyo, is a AAA game development studio dedicated to creating exciting and creative experiences for a worldwide audience. Tango is responsible for the renowned survival horror franchise The Evil Within and supernatural action title Ghostwire: Tokyo. Most recently, Tango launched the critically acclaimed hit rhythm action game Hi-Fi RUSH on Xbox Series X|S, PC, and available through Xbox Game Pass.
Hi-Fi RUSH hit last month in a splash of retro-futuristic nostalgia and ‘90s sound, delivering a surprisingly colorful, tightly designed rhythm-action game from a studio previously known mostly for its dark, jarring horror titles.
How it came to life inside Tango Gameworks says much about studio founder Shinji Mikami’s openness to ideas that intrigue him and the hard work of the two-person team that took an idea born of perhaps one too many Shaun of the Dead viewings and turned it into a captivating prototype.
In a wide-ranging interview, John Johanas, Hi-Fi RUSH Game Director, Kosuke Tanaka, Hi-Fi RUSH Lead Graphics Programmer, and Yuji Nakamura, Hi-Fi RUSH Lead Programmer, talk to us about using Unreal Engine for that pivotal early prototyping, and how Futurama helped them lock in the game’s style–which focuses on delivering animation that is colorful, sharp, and clean.
The two also explain how an early build’s success at Bethesda earned the title more internal support, and the impact they think the game may have on the studio long term.
I read that Hi-Fi RUSH started out as an idea sparked by the feeling you get watching a hit land timed to a beat in a song or during a video game trailer. How did that turn into the first concept for this game?
John Johanas, Hi-Fi RUSH Game Director: There were multiple inspirations in that sense. The general premise was brought upon by a singular scene in the movie Shaun of the Dead where the characters fight zombies in a bar synchronized to a Queen song. However, this is normally seen in music videos and trailers where synching things to the music just makes those hits hit harder.
I had experience in high school and university playing in bands and know what a kinetic and intense feel it is to “nail” playing something. I kept coming back to this idea of somehow making an action game that allowed you to have freedom of movement, but in the end all your hits matched with the music to make each hit impact harder and provide players with the same feeling as what they might experience when playing an actual instrument live.
So, the general “action that synchronizes to the music” idea was in my head for a while, but it took a bit until I was able to clearly see a way that this could technically be achieved.
How was the idea pitched to Tango Gameworks founder Shinji Mikami and what was his initial reaction?
Johanas: I still remember pitching it internally to Mikami-san and, while the idea was far out there, I think he saw that I had a strong passion to get the idea to work. He understood the natural connection between music and action and thought it was an idea worth proceeding but knew the technical challenge of if it could be done. So, we decided to have a small team develop a prototype. In general, Mikami-san doesn’t shun even strange ideas for games if the general concept is interesting, so I’m extremely grateful for the support of such a wild idea from the beginning.
Part of prototyping the idea was also because it was so different from anything we’ve done before that. I think there was a need to show that it could work and be fun to get it off the ground, since on paper it probably wouldn’t get far.
Was there any push back on the idea of creating such a light-hearted, colorful game when the studio was so known for horror games at this point?
Johanas: Internally when we were prototyping, there wasn’t pushback, more than a worry of “could this technically be done?”
Post prototype phase, when we literally had a proof-of-concept playable version that showed the fun of the game, there was much less push back. And while the pitch document showed a colorful game as our ideal art style, the project began with specifically no art, so we could make sure the gameplay fundamentals were front and center.
We were known externally as a studio that developed horror games, but internally we knew that wasn’t a genre we wanted to be pigeonholed into. I didn’t feel there was much push back on the idea of going in a completely different direction, as long as we can make it work.
Image courtesy of Tango Gameworks and Bethesda Softworks
Why did you decide to develop the game using Unreal Engine?
Johanas: After The Evil Within 2, our studio decided to shift to Unreal Engine for our next game in production, Ghostwire: Tokyo. Due to the shift in everyone moving to that engine, we knew that using multiple engines was not realistic, so we went ahead with prototyping in Unreal Engine.
It sounds like prototyping was especially important for Hi-Fi RUSH because you had to prove to Mikami and others that the concept was achievable. What impact did using Unreal Engine have on that need to prove the mechanic and gameplay through early prototyping possible?
Johanas: The initial team was literally myself and our lead programmer Nakamura-san, so he can answer more technically, but while he was working on the system that I had documented about ensuring that all animation and activities would move to the music. I was using Unreal Engine to build out a playable stage that would not only educate players on how to play, but also showcase how a world moving to the music can create just a fun experience.
It was our first time using the engine, so we were essentially learning from scratch. I would Blueprint gimmicks tying into our Rhythm Synchro System to have the environment bounce and react to the music cues. Also, I did all the grey boxing, music cues, and UI design/animation to showcase their connection with the music.
The number of tools available and open to the designers can feel overwhelming at first, but when you grasp the general process, I realized I was much less reliant on the programmers to do the work for me when I could dabble in the multiple applications within Unreal Engine. This allowed a very small team to wear many hats and prototype ideas, even if they weren’t professional and clean at first.
Yuji Nakamura,. Hi-Fi RUSH Lead Programmer: At the start of development, there was only me and John, Hi-Fi RUSH’s Director, so it was impossible for us to make game-assets like specific motions or effects from other members when developing a battle system that synchronizes to the music.
We could purchase starter content or assets from the marketplace in UE4, which was quite helpful. Also controlling animations with features like “Animation Montage” and “Animation Composite” was also useful. In Hi-Fi RUSH, we implemented a system to adjust each hit timing of the player’s attack to synchronize with the beat. However, we needed to iterate a lot until we could get it to its current implementation.
An example of iteration was adjusting the hit timing of attack relative to pressing the attack button. We tested variations where it was an eighth note later, or a quarter note later, and other variants. Without this system to systematically adjust animations, it would have taken an incredible amount of time to adjust, especially since we didn't have any animators on the team at the time.
In order to synchronize everything in the game to the music, we needed to feed the music information in beats to each object in the game. We could easily handle giving the correct information easily using the “Delegate” system in UE.
This was easy to access from within the Blueprints system, so the director made props that moved to the music on his own. Thanks to that system, it was easier to visualize the direction to the rest of the development team on how everything in Hi-Fi RUSH’s world would move in synch with the beat.
Image courtesy of Tango Gameworks and Bethesda Softworks
How difficult was it to nail the feel of the game, specifically when it came to making sure that a player’s actions and the world–even in the cutscenes–sync up with the music?
Johanas: The general gameplay mechanics revolved around the idea that regardless of the input timing, we would interpolate the action to eventually execute on the beat. However, there were a few exceptions including when players jump, dash, parry and other general movement since those felt too restrictive. But any animation that had what we call a “lead in” would adjust its animation timing to make sure the finishing point of the action hit on the beat.
The next step was encouraging players to press the button on the beat. From a game design perspective, we made sure to not fall into a trap of creating a fail-state, where the player refuses to perform an action for not playing in time. The inputs on rhythm would reward the player rather than penalize them.
This concept alone required a lot of iteration, but the whole concept affected our pipeline for every aspect of the level and game design as well.
To make sure that events would always sync up to the beat, we would need to think backwards. Cutscenes and events would not only need to start on rhythm but usually at the top of the measure, so we needed to have the player perform certain actions and sometimes interpolate them to take a bit more time so a transitional sequence would always trigger at the right time. Or we’d create alternate musical transitions depending on how much time was left until the first beat, so it didn’t feel too awkward.
All cutscenes were created in musical time with a metronome starting at the first beat of a measure and always ending at the end of a measure so we can smoothly transition to either gameplay or other aspects seamlessly both visually and musically.
It required an insane amount of coordination and planning out everything, but also the assistance from systems our programmers created to ensure that everything tied into musical cues, the rhythm of the song, or user inputs.
Nakamura: Most of the team members outside of the Director and Sound team were struggling to understand basic music logic since they didn’t have much experience with music. To help, we implemented a simple UI for debugging use in-game.
It was a simple UI like sheet music, constantly showing a 4/4 beat where we could see when a button was pressed or when a gimmick was activated. Using this debug display, we could identify things like where we needed to adjust animation timing, or when things didn’t sync up with the music.
Were there any particular Unreal Engine tools that made this process easier?
Nakamura: The animation editor made creating all the player and enemy animations very convenient.
The programmers could adjust certain playback speeds, immediately check them in-game and then give concrete directions on how to adjust the motion data. This was an efficient workflow which reduced the back and forth between requesting animation fixes.
When creating the environments, we used the Editor Utility feature, which allowed us to change the parameters of all objects in the level with a single command when there would be changes to the design.
What made you decide to go with cel-shading and how was that look achieved?
Johanas: The cel-shading look was a direction from the very beginning. While the style hadn’t been determined in the pitch, we wanted something that visually was experimental but also nostalgic of a late 1990s and early 2000s era where games took a turn to creatively push art styles in games. While there have been multiple cel-shaded games, I pushed for something that felt simple but polished. I asked for no color gradations, and nothing that would feel 3D.
Our original concept art was a 2D image of the first stage with its blue sky and exaggerated architecture. I took it to the team and requested that we make the visuals indistinguishable from that 2D art.
Internally, we focused on three keywords that helped define our style, which were “Colorful,” “Sharp,” and “Clean.” They were simple enough to make a clear identity without too much detail. Over time, we layered on different techniques, but surprisingly, it was always adding to the initial look we established without redoing or reapproaching our goal.
Kosuke Tanaka, Hi-Fi RUSH Lead Graphics Programmer: Achieving our cel-shading look was a challenge and involved a lot of experimental back and forth between the graphics team and the various art teams. Graphic techniques have evolved since the late 1990s and early 2000s, and in order to give a newer take on cel-shading, we felt it was important to try to incorporate modern 3D lighting techniques such as deferred lights, volumetric fog, ambient occlusion, improved shadows, global illumination, real-time reflection, bloom, and shader lens flares in the expression of our own colorful 2D style.
Unreal Engine provided us with a great base to work from. Our ambient occlusion, real-time reflection, dynamic shadows, static global illumination and bloom are stylized extensions of the base Unreal Engine functionality. Engine modifications were necessary to achieve our look, but we were careful to make changes to the base engine code. A lot of our modifications are contained in original additional passes, making engine merge conflicts less troublesome.
A clean, sharp image was an early art goal, and this meant aiming for maximum native resolution on our target platforms. For consoles, this means 1440p on Xbox Series S and 4K on Xbox Series X. A goal set with a clear warning that since our game is a rhythm action game, a solid 60 FPS takes priority over graphics. We were disciplined in our core tech selection and always added features with performance in mind. The graphics and all art teams are happy and proud that we were able to achieve a stylized look while making sure the game runs great.
Image courtesy of Tango Gameworks and Bethesda Softworks
The look of the game is very retro-futuristic. Where did you draw inspiration for that aesthetic and what were you hoping to achieve?
Johanas: The art style was deliberately meant to be that sort of retro-futuristic as part of that throwback to the late 1990s and 2000s that we kept coming back to. Things needed to feel not sleek and simple to match the visual aesthetic we were going with our cel-shading look.
One of the basic inspirations was the television show Futurama, which was far future but also the robotics were very retro.
I do think there is a tendency to make robotics and those elements as sleek as possible, but just like the characters in the game, I wanted to have this sort of “work in progress” feel to everything, including the technology in the game.
I think we found a great balance that gives the technology its own character while keeping the visuals simple and readable for the player.
I understand that the game was quite a hit internally while still under development. How did the internal virality of the game impact your development?
Johanas: That internal virality was essential for getting the game greenlit. As mentioned before, it is kind of a tough sell coming from us and even Bethesda. But when we made our prototype and it was passed around the office, it immediately struck a chord (sorry...) with other developers and people within the company. It was different than anything we’ve done before, but it was a lot of fun and unique.
That virality helped us know we were on to something special, but also helped us get the support we needed to get internal support since everyone enjoyed it even though it was just a grey-box.
What inspired your selections for the game’s soundtrack?
Johanas: The soundtrack was also curated from that late 90s/early 2000s feel. It’s personal for me, since the licensed tracks were from artists that I listened to and reminded me of that time period. Additionally, there is a certain sound and production quality that has gradually shifted over time, as trends tend to do, but that time felt a bit more unrefined and had a bit of a roughness to it.
Abstract as that maybe, it just felt right, which is more of a feeling than anything. We played around with BPM ranges that worked within our combat to have a general pace that we thought played the best, but also knew we can adjust the BPM faster to increase difficulty, create variation, and our interpolation system would make everything works as well.
The non-licensed tracks were produced with me giving a playlist of rock genre songs to the sound team for an overall direction on different vibes that matched with the levels and their visuals and events. But all those tracks were mostly rooted in that throwback vibe.
The streamer mode is an interesting addition. How important do you think it is to provide a streamable soundtrack to a game in terms of its long-term success?
Johanas: I think the streamer mode was invaluable to our success. Our whole release and even early development revolved around the virality of the game and spreading the joy and fun experience we created. We wanted to have these licensed tracks since that was an integral part of the identity of the game, but also knew that streaming is a constantly changing legal issue with how takedown issues are handled. For that, talking with our marketing team, we knew that the streaming aspect was something we wanted to allow players to do without fear, as to showcase the game and hopefully have the same successful virality as it had internally.
Making it work, however, was a huge challenge in itself. All the licensed music sections were custom created to the songs and included things such as boss attacks to specific guitar riffs and cutscenes with subtle animations to parts of the song. We had to reverse engineer these songs to create new songs that matched the flow of the original ones and include aspects that mirrored what the originals had so nothing of the experience was lost. On paper, we didn’t think it would work out, but with the help of our outsourced partners (who named themselves The Glass Pyramids for their work on the project), we made some amazing original tracks that don’t detract from the experience.
Image courtesy of Tango Gameworks and Bethesda Softworks
What made you decide to release the game the same day it was announced? Do you think that ultimately was part of the game’s success?
Johanas: The plan for the immediate announcement then release was Bethesda’s marketing strategy for the game. We knew we had something new, novel and we were all very confident in the quality of the game. But one of the biggest factors that led to it gaining popularity internally was when people played the game, they felt how good the experience was.
We knew we wanted a short campaign, but it slowly became harder to find a good announcement point to lead into our release date. When the Xbox Developer_Direct idea came up, the proposal of launching it right after we announced it was decided.
While we can attribute the surprise of the announcement of the experience to be its own story. I think the fact that the game was suddenly available gave players an outlet to immediately play the game, share their experience and talk about it, without room to wonder about what the game is or isn’t.
In the end I think it wound up working out better than any of us could have expected.
Hi-Fi RUSH appears to be doing very well so far, with 98 percent of critics recommending it on OpenCritic and it hitting the top ten for sales by revenue on Steam. What do you think of the overwhelmingly positive response and high sales for the game?
Johanas: As a developer, we are blown away by the positive reception we’ve received. Taking a leap of faith on a new IP and going so far away from what our studio was known for has its risks, but to see the almost unanimous sentiment be positive of the game is amazing.
But one of the greatest things is seeing players comment on the pillars of the game resonating so well with them. The accessibility of the game for non-rhythm fans, the polish in the product, the aesthetic, and overall “fun” and throwback feeling were all things that we wanted users to feel when they played the game. So, to see them literally use the same words as our pitch document, confirmed we hit our mark.
Are you able to play Hi-Fi RUSH without tapping your foot?
Johanas: When you work on a game for this long, just like Chai, you can internally feel the beat, so I no longer need to tap my foot at all. Instead, I see everything around me pulsing to the beat, which may mean I need to take a break from playing.
Nakamura: I don’t think I need to since I’ve been playing the game so much, but sometimes I realize I unconsciously AM moving my body!
Tanaka: Answering this with a sheepish look, but throughout dev, I used our debug mode a lot to check and fix graphics quicker. Post launch, I have more time to enjoy gameplay and am playing our patch build checks relying solely on my improving ability to feel the beat.
What lasting impact do you think Hi-Fi RUSH will have on Tango and the sorts of games it might develop in the future?
Johanas: Even before the game was released, internally, everyone was so proud of the work we had done. So, when we approached the announcement, I told the team that if we can go from horror games to this, and pull it off, we can pretty much do anything if we keep a strong vision and are committed to executing with high quality.
In that sense, I would never take anything off the table, in terms of future projects, if inspiration strikes.
Thanks for taking the time to chat with us. Where can people find out more about Hi-Fi RUSH and Tango Gameworks?
Johanas: We have websites, Twitter, Instagram, our studio page, but more importantly, if you run out on a busy street and just yell to the sky “Hi-Fi RUSH!!!!” one of our PR representatives will appear and answer any questions about the game to you.
Get Unreal Engine today!
Get the world’s most open and advanced creation tool.
With every feature and full source code access included, Unreal Engine comes fully loaded out of the box.