FaceCue: Give Unity Characters a Line and They Give You a Face

FaceCue turns dialogue into Unity facial animation with phoneme-based lip sync, co-articulation, emotional and head motion, multilingual support and a rig mapping wizard.
FaceCue facial animation product artwork
Image: Firecrest Roost. Source: https://facecue.net/assets/facecue-share.403bdff2.jpg

For those who don’t know the tool: FaceCue is a Unity facial-animation system that uses dialogue audio and optional transcript information to coordinate lip, face, eye and head motion. The promise is less time keyframing a greeting and more time deciding whether the character actually means it.

Making a digital character say one sentence is easy. Making the face convincingly say it, blink, breathe and express an emotion without becoming an accidental horror short is rather less easy. FaceCue is Christos Nasioutzikis’s Unity facial animation system, designed to turn dialogue audio and optional transcript information into coordinated lips, eyes, expression, head movement and breathing. There is a video and browser demo for the curious and the appropriately skeptical. The face may finally catch up with the script. And a warning: Playing with the Web-Version stole a solid hour, but we have the character reciting something unprintable.

Face rig designer interface showing phoneme setup for "sh" sound with detailed mouth shape and configuration options.

Phonemes rather than a jaw-flapping metronome

FaceCue bases lip synchronization on phonemes, with co-articulation and linguistic rules that account for neighboring sounds. Ten languages, including German, are individually tuned, with more than 50 additional languages supported. A single dialogue line can change language mid-sentence by marking ranges in its transcript. That is useful for localization and multilingual characters, colourful swearing (Which always should be multilingual, because otherwise what’s the point in learning another language), although performance and naturalness across every language need hands-on testing. Not all vowels are created equal, whatever the character’s jaw might suggest.

Lip sync animation software interface showing audio waveform, phoneme breakdown, emotion cues, and viseme output graph.

Rig mapping without re-inventing every eyebrow

A setup wizard recognizes common rig families, including MetaHuman, Character Creator, Daz and VRoid, and builds a face map from hand-tuned recipes. Visemes are assembled from FACS shapes rather than blindly using preconfigured mouth shapes. This matters because rigs vary greatly in facial controls, naming conventions and useful expressions. The Face Rig Designer lets artists refine any part of the mapping, while the Clip Composer previews a baked line without entering Unity Play Mode. These are developer-described workflows, not benchmarked results.

Animation as a component of the pipeline

By coordinating mouth, eye, head and breathing motion from the same dialogue input, the system aims to reduce the mechanical cleanup between voice takes and believable characters. The baked preview also suggests a practical review step before integrating the result into a game scene or interactive performance. Complex acting remains an artistic task: phoneme alignment is not a substitute for intent, facial asymmetry or a performance director. A character can pronounce the sentence perfectly and still look like it is reading its tax return.

FaceCue OneClick Wizard detect rig family interface showing progress bars for speech, pseudo-speech, and eye look detection.

Availability, testing and the fine print

Current regional prices, taxes and licence conditions are listed in the Unity Asset Store. Full prize is 129€, currently discounted to 65€. A browser demo, Windows build, Android demo and documentation are linked from FaceCue Media and FaceCue Docs. The provided feature account is from Firecrest Roost; DP has not independently measured rig compatibility, language quality or runtime costs – a convincing line delivery still requires a little more than moving the jaw.

https://facecue.net/

The compatibility guide specifies Unity 2021.3 through the tested Unity 6.6 release. Baking runs in the Windows editor; baked cues drive the rig at runtime rather than storing animation keyframes. Live-viseme support is unavailable on Linux and WebGL, and Apple-silicon Macs use a fallback. The language guide distinguishes ten tuned languages, over fifty additional decoder languages and a less precise universal audio route. A transcript improves alignment; text alone is not the complete lip-sync input.

ProductFaceCue
DeveloperChristos Nasioutzikis / Firecrest Roost
HostUnity
InputDialogue audio; optional transcript information for alignment, according to the developer
OutputFacial and head animation for mapped character rigs
FeaturesPhoneme lip sync, expression, eye movement and co-articulation
DocumentationOfficial documentation