Voice Detection

VoiceReactController provides the core voice detection system with configurable sensitivity presets and gameplay-friendly features.

Detection Presets

VoiceReact includes three built-in sensitivity presets that cover most use cases.

Very Sensitive

Best for horror games where enemies should react to quiet sounds.


voiceController.SetDetectionPreset(DetectionPreset.VerySensitive);

Characteristics:

Detects whispers and quiet speech easily
Start threshold: -25 dB
Lower noise margin (2 dB)
Shorter silence duration (0.6s)

Use Cases:

Horror games (monsters hear everything)
Tense stealth scenarios
Forced silence mechanics

Balanced (Default)

Good default for most gameplay scenarios.


voiceController.SetDetectionPreset(DetectionPreset.Balanced);

Characteristics:

Reliable detection without false positives
Start threshold: -20 dB
Moderate noise margin (3 dB)
Standard silence duration (0.8s)

Use Cases:

General gameplay
Stealth games with moderate difficulty
Voice-controlled interactions

Less Sensitive

Best for noisy environments or when you want players to speak clearly.


voiceController.SetDetectionPreset(DetectionPreset.LessSensitive);

Characteristics:

Requires louder, clearer speech
Start threshold: -15 dB
Higher noise margin (5 dB)
Longer silence duration (1.0s)

Use Cases:

Noisy environments (trade shows, arcades)
Competitive multiplayer (reduce false positives)
Voice commands that need clarity

Detection Properties

Check if Player is Speaking


if (voiceController.IsPlayerSpeaking)
{
    Debug.Log("Player is currently speaking");
}

Get Current Volume


// Raw decibel level (-60 to 0 dB)
float db = voiceController.CurrentDB;
 
// Normalized volume (0-1 for UI meters)
float volume = voiceController.NormalizedVolume;

Speech Duration


// How long has the current speech lasted?
float duration = voiceController.SpeechDuration;
 
// How long since they last spoke?
float timeSinceSpeech = voiceController.TimeSinceLastSpeech;

Noise Floor

Background noise level automatically tracked:


float noiseFloor = voiceController.NoiseFloorDB;
Debug.Log($"Background noise: {noiseFloor} dB");

Detection Cooldown

Prevent spam by requiring silence between detections:


[SerializeField] private VoiceReactController voiceController;
 
void Start()
{
    // Require 2 seconds of silence between detections
    voiceController.detectionCooldown = 2f;
}

Or set cooldown dynamically:


void OnSpeechEnd(float duration)
{
    // Force cooldown after speech
    voiceController.SetDetectionCooldown(3f);
}

Check cooldown status:


if (voiceController.IsInCooldown)
{
    Debug.Log("Detection is in cooldown");
}

Silence Detection

Check if player has been silent for a specific duration:


void Update()
{
    // Reward for staying quiet
    if (voiceController.HasBeenSilentFor(10f))
    {
        Debug.Log("Player has been silent for 10 seconds!");
        AwardStealthBonus();
    }
}

Volume Normalization

Enable normalized volume for UI displays:


voiceController.enableVolumeNormalization = true;
 
void Update()
{
    // Get 0-1 value perfect for UI sliders/meters
    float normalizedVolume = voiceController.GetNormalizedVolume();
    volumeMeter.fillAmount = normalizedVolume;
}

Normalization automatically tracks min/max volume during the session and scales current volume to 0-1 range.

Custom Configuration

For full control, use Custom preset and configure manually:


voiceController.SetDetectionPreset(DetectionPreset.Custom);
 
// Access low-level MicrophoneInputHandler for custom thresholds
var handler = voiceController.GetComponent<MicrophoneInputHandler>();
 
// Create custom settings
var customSettings = new VoiceDetectionSettings
{
    startThresholdDB = -22f,
    endThresholdDB = -32f,
    minAmplitudeThreshold = 0.003f,
    significantAmplitudeThreshold = 0.06f,
    noiseMarginDB = 3.5f,
    minPercentageNonZero = 0.05f,
    isConfigured = true
};
 
handler.ApplyDetectionSettings(customSettings);

Understanding Detection Parameters

Start Threshold (dB)

Volume level required to start detecting speech.

Lower = more sensitive (detects quieter sounds)
Higher = less sensitive (requires louder speech)
Range: -60 to 0 dB

End Threshold (dB)

Volume level below which detection ends.

Should always be lower than start threshold
Creates hysteresis to prevent flickering on/off
Range: -60 to 0 dB

Noise Margin (dB)

Additional buffer above background noise floor.

Higher = more immunity to noise
Lower = more sensitive but risk false positives
Range: 1-10 dB

Amplitude Thresholds

Min Amplitude Threshold: Minimum audio amplitude to be considered above noise.

Range: 0.001 to 0.01
Lower = more sensitive

Significant Amplitude Threshold: Amplitude required for valid speech detection.

Range: 0.01 to 0.5
Higher = requires clearer speech

Common Patterns

Dynamic Difficulty

Adjust sensitivity based on game difficulty:


public void SetDifficulty(int level)
{
    switch (level)
    {
        case 1: // Easy
            voiceController.SetDetectionPreset(DetectionPreset.LessSensitive);
            break;
        case 2: // Normal
            voiceController.SetDetectionPreset(DetectionPreset.Balanced);
            break;
        case 3: // Hard
            voiceController.SetDetectionPreset(DetectionPreset.VerySensitive);
            break;
    }
}

Contextual Sensitivity

Change detection based on game state:


void OnEnterStealthZone()
{
    // In stealth areas, be more sensitive
    voiceController.SetDetectionPreset(DetectionPreset.VerySensitive);
}
 
void OnEnterSafeZone()
{
    // In safe areas, less sensitive
    voiceController.SetDetectionPreset(DetectionPreset.LessSensitive);
}

Volume-Based Mechanics

React to different volume levels:


void Update()
{
    if (voiceController.IsPlayerSpeaking)
    {
        float db = voiceController.CurrentDB;
 
        if (db > -15f)
        {
            // Shout (very loud)
            TriggerShoutMechanic();
        }
        else if (db > -25f)
        {
            // Normal speech
            TriggerNormalMechanic();
        }
        else
        {
            // Whisper (quiet)
            TriggerWhisperMechanic();
        }
    }
}

UI Volume Meter

Display real-time volume in UI:


using UnityEngine.UI;
 
public class VolumeMeterUI : MonoBehaviour
{
    [SerializeField] private VoiceReactController voiceController;
    [SerializeField] private Image fillImage;
    [SerializeField] private Color quietColor = Color.green;
    [SerializeField] private Color loudColor = Color.red;
 
    void Update()
    {
        // Get normalized volume (0-1)
        float volume = voiceController.GetNormalizedVolume();
 
        // Update fill amount
        fillImage.fillAmount = volume;
 
        // Color based on volume
        fillImage.color = Color.Lerp(quietColor, loudColor, volume);
    }
}

Debugging

Enable debug mode to see real-time detection data:


var handler = voiceController.GetComponent<MicrophoneInputHandler>();
handler.EnableDebugMode(true);

This shows:

Current dB level
Noise floor
Detection thresholds
Audio waveform visualization
Detection state

Next Steps

Voice Listeners Events API Reference