The Choicer Voicer

The judges in The Choicer Voicer are not people at all — they are computer-controlled panelists who sit and vote on your vocal impressions round by round, and the whole idea traces back to a streamer’s throwaway suggestion in December 2023. That single comment from Vinny Vinesauce, wondering aloud if someone should remake the old “Choicest Voice” minigame from Mario Party: Island Tour, is the reason this game exists at all.

From a Vinesauce Joke to The Choicer Voicer

Developer YeahMaybe was messing around with the Godot engine at the time and decided to take the suggestion literally. The first version was barely more than a meme: a handful of hard-coded audio clips and a basic scoring loop built to imitate the old Mario Party bit where players do their best impression of a character’s voice and a panel decides who nailed it.

What changed things was a decision to let the game load files at runtime instead of keeping everything hard-coded. Once voice packs could be swapped in from a folder instead of baked into the game, the scope kept widening. The developer has put it plainly: “Why customize just the voice packs?” That question ended up reshaping menus, the judge panel, the studio itself, and eventually the host character too.

A small team filled in around that expanding idea. AzureOtsu composes the music, leaning into a distinctly 2000s sound. Kiophen handled the artwork for a character named Shae. Studio Jimbly built out the GLTF import pipeline that lets players bring in their own studio models. None of this was planned from day one — it grew because the core loop of doing a voice and getting judged for it turned out to be worth building around.

Judges, Studios, and How Rounds Play Out

The main mode is a game show studio built for one to four players. Someone picks a prompt, players record their impression through a microphone, and the panel of computer judges votes on each performance. It plays out in short bursts — a full round takes a few minutes, which keeps it usable as a party format rather than something that demands a long sitting.

Recording works through live microphone input with waveform playback, so you can hear your own take before the judges weigh in. It is a simple loop by design: record, get judged, pass the controller. The comedy comes from the gap between what you meant to sound like and what actually came out of the microphone.

There is also a Twitch-facing option where a streamer’s chat votes on performances instead of the built-in judge panel, and a separate Dub Mode where players record a voiceover over a scene rather than performing a character impression cold. Both modes reuse the same recording core, just pointed at a different kind of judging or a different kind of prompt.

What You Can Actually Customize

• Voice packs — adding a new one is just dropping audio files into a folder
• The judge panel and host character
• The game show studio’s look, including imported GLTF models
• Menu aesthetics
• Content packs that bundle several of the above for others to download

Building Content Packs for The Choicer Voicer

Because the loading system pulls from folders rather than fixed files, making something new does not require touching code. A voice pack is just audio clips arranged the way the game expects them. That low barrier is presumably why content packs exist as a shareable unit in the first place — it lets a group of friends build a pack full of in-jokes without anyone needing to understand Godot.

This same openness is also where the game’s rough edges show. Microphone recording has known problems, particularly on surround-sound audio setups, and the developer has been upfront that this stems from a Godot engine limitation rather than something easily patched from the game’s side. It is the kind of issue that comes with building on an engine still maturing in this area, and it currently requires some trial and error with audio settings to work around.

The game currently sits at an early access alpha stage, with foundational code that the developer describes as a couple of years old at this point — a sign of how much has been layered on top since that first meme build. A “No Gameplay Demo” exists purely for testing whether your microphone is picked up correctly before you commit to a full session.

Pricing and Where It Stands Now

Access uses a pay-what-you-want model with a five dollar minimum, and it runs on Windows and Linux. Reception on itch.io currently sits at 4.8 out of 5 across 57 ratings, which for a game still in early alpha with acknowledged microphone bugs is a fairly strong signal that the core bit — impressions judged round by round — lands for the people trying it.

On the technical side, the toolchain behind it mixes Godot for the engine with GIMP for art, Audacity for audio cleanup, FL Studio for the score, and Blender for anything that needs modeling. The listing is upfront that no generative AI was used anywhere in its content, and right now everything is presented in English only, with mouse-driven menus throughout.

The Choicer Voicer is still a small, actively changing project, but its origin as a one-off tribute to a Mario Party minigame is easy to see in every round: pick a prompt, do the voice, let the panel decide, and pass it to the next person.