If I have to define Gris sound in a word I would call it dynamic. Of course minimalism has been always the reference, the softness of the sound along the game. But the main property of its sound is the way it changes with soft transitions inside the game. The dynamism when it has to adapt to the gameplay, to the music and giving you the space to enjoy the two main corner stones of the game: The art and the music.
Sound in Gris has two main tasks: Situate and narrate. We achieve the situation by using a rich range of ambiences and sound effects, with the iteration of the player and the technology we had. The narrative is supported in the proper audio direction previously defined, and in this case we have taken music as one of the main support vehicles to tell the story through sound. Therefore, the sound direction is composed by three parts, each one of them with its own references:
- Technical part: As main references Inside, Battlefield 1, The last of Us
- Minimalistic part: Abzû, Monument Valley, Hidden Folks, Journey, Alto’s Odyssey, Arrival (film)
- Cinematographic sound design: God of War, The Last of Us, Ghost in the Shell (film)
As Gris is an independent development, the resources have been limited. Fortunately, Nomada Studio has been always trying to take care of the sound with technology or giving resources for it, but of course everything is way far from, for instance, the resources we have at Ubisoft to make AAA games. This is the software we have used for Gris:
- Logic Pro X
- Izotope RX 6
Regarding the gear:
We just used the material seen above for Gris, working in my room. A Macbook Pro, Beyerdynamic DT770 headphones, a Rode NTG2, a contact microphone, an Audient iD14 audio interface and a Zoom H6 recorder. Unfortunately for us, we didn’t have speakers or a good room (in acoustic terms) or a good bunch of microphones, but nevertheless they have worked pretty well.
Creative Part: References, creation and recordings
From the very beginning we knew that music and art were the corner stones of the project. Sound design should give the player the experience of situation, realism and join the player in the story, but always giving the main role to them. Consequently we took “strange” decisions. For instance when the player achieves a new color there is just art and music. Sounds and ambiences are muted, or their volume reduced a lot, to let art and music tell the story
On the other hand, we have tried to put something personal in the project. For example, I made some ambiences recordings in a forest in El Espinar (Segovia, Spain), my father’s village where he was born and he grew up. Other example, foley related with sand was made with sand from Castellón’s Beach (España). My brother studied there and I asked him to bring me some sand. Branch sounds were made with branches from my parent’s garden. Some of the Bosses screams have screams from Gemma Gammara, Berlinist singer. The crows that you can listen to along the game were recorded in Japan, in Ueno Park in Tokyo and in the Osaka Castle Park during a trip I did with my girlfriend, just to name some of them. Obviously not all the sounds have this “Personal” thing, but our intention was always to bring this love to the project as much as possible.
In other matters there is the minimalistic style, which is so important in Gris. For sound we tried to spotlight it through three different ways:
- Limiting the number of voices that are being played at the same time: By limiting I don’t mean muting, what I mean is trying to keep just the number of voices needed in each moment avoiding saturation or information overload.
- Taking care of the layers: If there is an ambience we want it soft but giving the right information. We didn’t want sounds to be saturated with layers, we searched for the right layers and details so the sound could have the information needed without being an “attention seeker”. For instance, sand footsteps are not footsteps, they were made using a slide for guitars inside of a bowl to make them soft and light.
- Controlling each object volume and being “generous” with the dynamic range and the loudness levels (I’ll talk about that later)
Last but not least, the “cinematographic” sound design: Low frequencies, whooshes, hits, stingers.. I think these kind of details in some certain parts helped to tell what we wanted.
Technical Part: Wwise
Fortunately we have been able to work with Wwise. In my humble opinion is the best solution for audio in games nowadays: It’s reliable, the code is stable and works pretty well with different engines, it’s very versatile and it gives the sound designer a lot of technical and creative possibilities. Implementation in Unity was quite easy and it helped us achieve the soundscape we wanted. In fact, for the first demo we did three years ago we used Fmod. It worked well, but in the end we decided to use Wwise. I’m not going to talk deeply about the tech part because I don’t want to make this post endless, but I want to show you some stuff from our Wwise project.
We always wanted to keep the project in order, to make it easy to understand. We followed a specific nomenclature for every kind of sound and we divided the project in five main Work Units:
Ambiences, cinematics, foley, HUD and sound effects. The main goal was to organise the project based on levels, or based on general elements of the game like stars, destructible objects or abilities.
For ambiences we wanted to make it different from interior and exterior, when you are underwater, etc. We have used a lot states and RTPCs in order to control parameters like distance to camera or altitude:
Foley in Gris is 2D. This means foley is going to be always there, regardless the distance from Gris to the listener (camera). At the beginning we were not convinced at all about this because with the different cameras foley always had the same volumen, which was quite ugly. So we decided to put them in 3D with a big roll-off. But that didn’t convince us at all, it made weird situations with the camera. So we decided to put them back in 2D but with two RTPCs modifying the volume and applying a low pass filter based on the distance from Gris to the camera. And it worked!
However, sfx related with swimming are modified by their own RTPCs and not by States:
Regarding the mixing, the dynamic in the game is “generous” (to put it in some way). We wanted a dynamic game, so when there’s a quite moment player could enjoy it with calm sounds, but when the player is in a tense situation (for instance a boss fight) the dynamic is different and the loudness levels are higher. Sound in Gris has an integrated media around -26/-25 LUFS, and the short-term is between -39 LUFS and -18 LUFS (approximately) based on what’s happening in the game.
Gris is a game made with few resources. This means there are things that don’t reach the expectations (for example the dynamic of the sound in certain parts, or maybe some sound effects don’t have the expected quality). However we are so proud of the game we made, about everything not just about the sound. On my side and related with my area (sound) I want to highlight a few things. The music composed by Berlinist is the most beautiful music I have listened to in years (Marco albano, Gemma Gamarra and Luigi Gervasi are genius). Roger Mendoza has been the key to make the sound works so well in the game, he has been a corner stone for the audio side. And the help from Damián Sánchez (www.sonotrigger.com) was fundamental at the end of the project. We wouldn’t be able to reach the quality we were looking for without him, and he made a great work. All in all, I have been so lucky to work with a such talented, professional and amazing team. And in spite of the limitations we had, we are really happy with the result.
Thanks for playing Gris!