Character Components

Per-character conversation, voice input, and TTS playback.

---

UEstuaryCharacterComponent

UActorComponent — the per-character conversation interface. Created automatically on AEstuaryCharacter. This is the component you use most.

Methods

Function	Returns	Description
SendText(const FString& Text)	FString	Send a user message. Returns the generated message_id.
SayLine(const FString& Text, bool bTextOnly = false)	void	Speak a scripted line via TTS (skips the LLM, auto-interrupts).
Interrupt()	void	Abort the currently-streaming response. No-op if no active turn.
UpdatePreferences(const FEstuaryPreferences& Prefs)	void	Update session-level behavior flags.
SetEnableAnimation(bool bEnable)	void	Opt into ARKit-52 lipsync. Call before Connect.
SetEnableBodyAnimation(bool bEnable)	void	Opt into body animation (experimental). Call before Connect.
IsRegistered() const	bool	True when registered with the subsystem and routing is active.
GetCurrentMessageId() const	FString	message_id of the most recent turn.
GetCharacterId() const	FString	Character ID inherited from the owning actor.
GetLiveLinkSubjectName() const	FName	Resolved Live Link subject for lipsync.

Properties

Property	Type	Description
LiveLinkSubjectOverride	FName	Override the Live Link subject name. None = use Character ID.

Events (BlueprintAssignable)

Delegate	Payload	Fires when
OnBotResponse	FEstuaryBotResponse	Each streaming text chunk (bIsFinal on the last).
OnBotVoice	FEstuaryBotVoice	Each streaming TTS audio chunk.
OnSttResponse	FEstuarySttResponse	Each partial/final transcription during voice input.
OnInterruptAcked	FEstuaryInterruptPayload	The server confirms an interrupt.
OnBotAnimation	FEstuaryBotAnimation	A renderable ARKit-52 lipsync frame.
OnBotPose	FEstuaryBotPose	A renderable body-pose frame (experimental).
OnVisionResponse	FEstuaryVisionResponse	A vision (VLM) result for a sent camera frame.
OnCameraCaptureRequested	FEstuaryCameraCapture	The server requests an image capture.

CharacterComponent->OnBotResponse.AddDynamic(this, &AMyActor::HandleResponse);
FString Id = CharacterComponent->SendText(TEXT("Hello!"));

---

UEstuaryAudioPlayerComponent

UAudioComponent — decodes and plays streaming TTS audio (bot_voice) with a short pre-buffer for pop-free playback. Add it to the character actor; it auto-wires to the conversation stream on BeginPlay.

Methods

Function	Returns	Description
EnqueueBotVoice(const FEstuaryBotVoice& Voice)	void	Queue a TTS chunk (called automatically).
EnqueueRawPcm16(const TArray<uint8>& PcmBytes, int32 RemoteSampleRate)	void	Queue raw PCM16 audio you produced.
FlushAllForMessage(const FString& MessageId)	void	Drop queued audio for a message (used on interrupt).
SetEnableAnimation(bool bEnable)	void	Enable the amplitude envelope tap used by lipsync.
GetCurrentAudioTime() const	float	Seconds elapsed in the current utterance.
GetPlaybackTimeSeconds() const	float	Total playback time.
IsBotAudioActive() const	bool	True while TTS is playing.
GetCurrentJawWeight() const	float	Current amplitude-driven jaw weight (lipsync fallback).

---

UEstuaryMicrophoneComponent

UActorComponent — captures the player's microphone and streams it for transcription. Cross-platform (Win64 / Mac / iOS / Android). Created automatically on AEstuaryCharacter.

Microphone capture is fixed at 16 kHz mono for transcription.

Methods

Function	Description
StartListening()	Begin the voice session and stream microphone audio.
StopListening()	End the session and trigger final transcription.
SetMuted(bool bMuted)	User mute. Never toggled automatically.
SetSuppressed(bool bSuppressed)	Half-duplex guard — stop sending while the bot speaks (WebSocket voice).
SetAutoBargeIn(bool bEnable)	Enable/disable auto-interrupt when the user speaks. Default on.

Properties (read-only)

Property	Type	Description
bIsCapturing	bool	True between StartListening and StopListening.
bIsMuted	bool	Current user-mute state.
bIsSuppressed	bool	Current half-duplex suppression state.
bAutoBargeIn	bool	Whether auto barge-in is enabled.

// Push-to-talk
void AMyPawn::OnTalkPressed()  { MicrophoneComponent->StartListening(); }
void AMyPawn::OnTalkReleased() { MicrophoneComponent->StopListening(); }

See Voice Connection for the full voice workflow.