Hello everyone,
This is Portefaix. This is my very first devblog, so I'll keep it simple: today I want to talk about an aspect we often overlook… even though it plays a huge part in immersion: sound design.
There would be enough material for a novel (or at least three devblogs), but for this first entry I'd rather focus on the creative side, with a short detour into optimization — because a good sound isn't just a “cool sound”; it also needs to work well in the game.
What is sound design?
Sound Design means creating every aspect of a product's audio. In a video game, that includes:
- sound effects (SFX),
- ambience,
- interface (UI) sounds,
- voices (voice acting),
- and sometimes jingles / cues (short musical sequences that signal something).
The main categories of sound effects
In practice, we often group sounds by function (what they contribute to the game) and by nature (how they are used technically).
SFX
The big catch-all category: everything that doesn't fit neatly into the other categories.Ambience
Sounds that make the world believable: rain, wind, distant insects… Often less important for gameplay, but essential for immersion. When an ambient sound is spatialized (a waterfall, a mill, a specific source), we often refer to an “environment” subcategory.UI sounds
Non-diegetic sounds associated with the interface: menus, confirmations, notifications, and so on.Foley
Everything resulting from a character's actions: footsteps, clothing rustles, handling weapons, small impacts… In video games, foley is often grouped with SFX, but the important distinction is that foley is processed very little and often recorded in a studio.Voice acting
Spoken, understandable voices.Jingles / Sound Cues
Closely related to music: short, effective sequences often used to announce an event (or a reward).
Designing a vocalization (a practical example: monsters)
Now that we've mapped out the broad picture, let's get practical. In Hana, one of the major parts of sound design is creatures.
The first step is analysis: what feeling am I trying to convey?
A massive creature doesn't sound like a small, nervous animal. An aquatic creature doesn't have the same sonic texture as a dry, land-dwelling monster. Even at the same size, the intention can change everything.
In Hana, my goal is simple: believability. I want your brain to think “okay, that could exist,” even if the creature is fantastical.
To achieve that, we often rely on real-world references: our ears like to connect what they hear to something they already know.
From microphone to monster
Once the intention is clear, we need to “build” the vocalization. The most reliable way to keep it organic is to start with real sounds: animals, human voices, raw textures…
Here, we'll keep it simple: we record ourselves. Growls, rasps, whistles… everything goes.
Don't strain your voice: it's easy to injure yourself (and losing your voice for a monster would be a shame).
Next, we clean up the recording if needed and begin transforming it. A common approach is to use a sampler: a tool that lets us play a sound back at different pitches / speeds and apply effects.
The goal is a paradox: heavily distort the original sound until it becomes unrecognizable… while keeping it believable and clear in the game.
Above all, we need variations. A monster repeating exactly the same cry becomes a “sound robot” within 30 seconds. So we prepare several versions, and in the game we often add slight randomization (volume, pitch) to break up repetition.
A matter of optimization
Once we have a satisfying set of sounds, we export it according to specific rules. This serves both optimization and easier integration (whether I do it or someone else does).
In Hana, audio assets are imported uncompressed, and Unity handles compression. It's a standard trade-off: distributing everything uncompressed to players would quickly inflate the game's size and memory usage, often for only a minimal audible improvement.
Why aren't all sounds “3D” (spatialized)?
In 3D, the engine has to calculate the relative positions of source and listener (panning, attenuation, and so on). That uses CPU time. So we reserve 3D for sounds with a real spatial purpose.
In Hana, creature sounds play in 2D: the position of the monster or the player doesn't change how they sound. What matters is when the sound triggers (attack, hit, death, and so on).
Many ambient and UI sounds also work very well in 2D (wind, menus, notifications). This avoids creating countless unnecessary sound sources and leaves room for everything else.
Audio samples
Tip: listen with headphones if possible — that's where the details stand out best.
An “organic” take before processing, to preserve the source material.
The same source, shaped with a sampler + effects (pitch, timbre, texture).
A set of variations designed to avoid repetition.
A simple ambience to illustrate how an area “breathes.”
Conclusion
Hana's audio is still being developed: the universe is still finding its way in some respects, but the nature of the game is already clearly defined.
I have plenty of ideas for giving the world of Hana a truly distinctive sound. And I can't wait to see what audio possibilities the 3D version opens up (not just for monsters).
My eventual goal is to offer you an audio universe that can be a source of surprise… and immersion.
See you soon,
Portefaix