first pass of tts and audio stuff

pi/bin/video-publisher.sh

@@ -17,6 +17,11 @@ VIDEO_WIDTH="${VIDEO_WIDTH:-1920}"
 VIDEO_HEIGHT="${VIDEO_HEIGHT:-1080}"
 VIDEO_FPS="${VIDEO_FPS:-30}"
 VIDEO_BITRATE="${VIDEO_BITRATE:-3000000}"
+AUDIO_ENABLE="${AUDIO_ENABLE:-0}"
+AUDIO_DEVICE="${AUDIO_DEVICE:-default}"
+AUDIO_RATE="${AUDIO_RATE:-16000}"
+AUDIO_CHANNELS="${AUDIO_CHANNELS:-1}"
+AUDIO_BITRATE="${AUDIO_BITRATE:-64000}"
 # Flip the camera 180deg (supported by rpicam-vid/libcamera-vid)
 FLIP_ARGS=(--rotation 180)
 
@@ -44,6 +49,43 @@ fi
 FLIP_ARGS=(--rotation 180)
 
 run_pipeline() {
+	if [[ "${AUDIO_ENABLE}" -eq 1 ]]; then
+		"${LIBCAMERA_BIN_PATH}" \
+			--inline \
+			--timeout 0 \
+			--width "${VIDEO_WIDTH}" \
+			--height "${VIDEO_HEIGHT}" \
+			"${FLIP_ARGS[@]}" \
+			--framerate "${VIDEO_FPS}" \
+			--bitrate "${VIDEO_BITRATE}" \
+			--codec h264 \
+			--profile baseline \
+			--denoise cdn_off \
+			--nopreview \
+			--awb custom \
+			--awbgains 1.5,1.8 \
+			--output - \
+			| "${FFMPEG_BIN_PATH}" \
+				-hide_banner \
+				-loglevel warning \
+				-fflags nobuffer \
+				-use_wallclock_as_timestamps 1 \
+				-f h264 \
+				-i pipe:0 \
+				-f alsa \
+				-thread_queue_size 512 \
+				-ac "${AUDIO_CHANNELS}" \
+				-ar "${AUDIO_RATE}" \
+				-i "${AUDIO_DEVICE}" \
+				-c:v copy \
+				-c:a aac \
+				-b:a "${AUDIO_BITRATE}" \
+				-ac "${AUDIO_CHANNELS}" \
+				-ar "${AUDIO_RATE}" \
+				-flush_packets 1 \
+				-f mpegts \
+				"${PUBLISH_URL}"
+	else
 		"${LIBCAMERA_BIN_PATH}" \
 			--inline \
 			--timeout 0 \
@@ -71,6 +113,7 @@ run_pipeline() {
 				-flush_packets 1 \
 				-f mpegts \
 				"${PUBLISH_URL}"
+	fi
 }
 
 while true; do

pi/install_roverd.sh

@@ -146,6 +146,41 @@ log "Installed roverd systemd unit"
 install_video_deps
 ensure_pwm_overlay
 
+# Enable Google AIY v1 sound card, ALSA defaults, and TTS engines
+install_audio_support() {
+	local boot_config
+	if ! boot_config="$(find_boot_config)"; then
+		log "WARNING: unable to locate /boot config.txt; please enable googlevoicehat-soundcard overlay manually"
+	else
+		if ! grep -Eq '^\s*dtoverlay=googlevoicehat-soundcard\b' "$boot_config"; then
+			local backup="${boot_config}.roverd.$(date +%Y%m%d%H%M%S).bak"
+			cp "$boot_config" "$backup"
+			{
+				echo ""
+				echo "# Added by roverd installer to enable Google AIY v1 sound card"
+				echo "dtoverlay=googlevoicehat-soundcard"
+				echo "dtparam=audio=off"
+			} >> "$boot_config"
+			log "Enabled googlevoicehat-soundcard overlay in $boot_config (backup at $backup). Reboot required."
+		else
+			log "googlevoicehat-soundcard overlay already present in $boot_config"
+		fi
+	fi
+	if [[ -f pi/asound.conf ]]; then
+		install -m 0644 pi/asound.conf /etc/asound.conf
+		log "Installed ALSA config to /etc/asound.conf"
+		alsa_reload_notice=1
+	else
+		log "WARNING: pi/asound.conf missing; skipping ALSA config install"
+	fi
+	log "Installing TTS/audio packages (flite, espeak)..."
+	apt-get update
+	apt-get install -y --no-install-recommends flite espeak
+	if [[ "${alsa_reload_notice:-0}" -eq 1 ]]; then
+		log "ALSA config updated; reboot recommended for overlay + audio changes"
+	fi
+}
+
 # Install video publisher assets
 install -D -o root -g root -m 0755 pi/bin/video-publisher.sh /usr/local/bin/video-publisher
 install -m 0644 pi/systemd/video-publisher.service /etc/systemd/system/video-publisher.service
@@ -158,11 +193,18 @@ VIDEO_WIDTH=1280
 VIDEO_HEIGHT=720
 VIDEO_FPS=30
 VIDEO_BITRATE=3000000
+AUDIO_ENABLE=1
+AUDIO_DEVICE=default
+AUDIO_RATE=16000
+AUDIO_CHANNELS=1
+AUDIO_BITRATE=64000
 ENV
 	chown roverd:roverd /var/lib/roverd/video.env
 	chmod 0640 /var/lib/roverd/video.env
 fi
 
+install_audio_support
+
 systemctl daemon-reload
 systemctl enable roverd.service
 systemctl enable video-publisher.service

pi/roverd/cmd/roverd/main.go

@@ -21,7 +21,7 @@ func main() {
 	if err != nil {
 		log.Fatalf("load config: %v", err)
 	}
-	if err := roverd.UpdatePublisherEnv(cfg.Media); err != nil {
+	if err := roverd.UpdatePublisherEnv(cfg.Media, cfg.Audio); err != nil {
 		log.Fatalf("prepare media env: %v", err)
 	}
 
@@ -55,7 +55,7 @@ func main() {
 
 	adapter := roverd.NewSerialAdapter(serialPort, logger)
 
-	mediaSupervisor := roverd.NewMediaSupervisor(cfg.Media, logger)
+	mediaSupervisor := roverd.NewMediaSupervisor(cfg.Media, cfg.Audio, logger)
 	if mediaSupervisor != nil {
 		mediaSupervisor.Start(ctx)
 	}

pi/roverd/command.go

@@ -7,6 +7,7 @@ type helloMessage struct {
 	MaxWheelSpeed int               `json:"maxWheelSpeed"`
 	Media         MediaConfig       `json:"media"`
 	CameraServo   CameraServoConfig `json:"cameraServo"`
+	Audio         AudioConfig       `json:"audio"`
 }
 
 type sensorMessage struct {
@@ -24,6 +25,7 @@ type inboundMessage struct {
 	SensorStream *sensorStreamPayload `json:"sensorStream,omitempty"`
 	Media        *mediaCommand        `json:"media,omitempty"`
 	Servo        *servoPayload        `json:"servo,omitempty"`
+	TTS          *ttsPayload          `json:"tts,omitempty"`
 }
 
 type driveDirectPayload struct {
@@ -51,6 +53,14 @@ type servoPayload struct {
 	PulseUs *int     `json:"pulseUs,omitempty"`
 }
 
+type ttsPayload struct {
+	Text   string `json:"text"`
+	Engine string `json:"engine,omitempty"`
+	Voice  string `json:"voice,omitempty"`
+	Pitch  int    `json:"pitch,omitempty"`
+	Speak  bool   `json:"speak,omitempty"`
+}
+
 type ackMessage struct {
 	Type   string `json:"type"`
 	ID     string `json:"id"`

pi/roverd/config.go

@@ -53,6 +53,18 @@ type BatteryConfig struct {
 	Urgent int `yaml:"urgent"`
 }
 
+type AudioConfig struct {
+	CaptureEnabled bool   `yaml:"captureEnabled" json:"captureEnabled"`
+	CaptureDevice  string `yaml:"captureDevice" json:"captureDevice,omitempty"`
+	SampleRate     int    `yaml:"sampleRate" json:"sampleRate,omitempty"`
+	Channels       int    `yaml:"channels" json:"channels,omitempty"`
+	Bitrate        int    `yaml:"bitrate" json:"bitrate,omitempty"`
+	TTSEnabled     bool   `yaml:"ttsEnabled" json:"ttsEnabled"`
+	DefaultEngine  string `yaml:"defaultEngine" json:"defaultEngine,omitempty"`
+	DefaultVoice   string `yaml:"defaultVoice" json:"defaultVoice,omitempty"`
+	DefaultPitch   int    `yaml:"defaultPitch" json:"defaultPitch,omitempty"`
+}
+
 type MediaConfig struct {
 	PublishURL     string   `yaml:"publishUrl" json:"publishUrl,omitempty"`
 	PublishPort    int      `yaml:"publishPort" json:"-"`
@@ -89,6 +101,7 @@ type Config struct {
 	MaxWheelMMs int               `yaml:"maxWheelSpeed"`
 	Media       MediaConfig       `yaml:"media"`
 	CameraServo CameraServoConfig `yaml:"cameraServo"`
+	Audio       AudioConfig       `yaml:"audio"`
 }
 
 func LoadConfig(path string) (*Config, error) {
@@ -127,6 +140,17 @@ func LoadConfig(path string) (*Config, error) {
 			HomeAngle:    0,
 			NudgeDegrees: 2,
 		},
+		Audio: AudioConfig{
+			CaptureEnabled: true,
+			CaptureDevice:  "default",
+			SampleRate:     16000,
+			Channels:       1,
+			Bitrate:        64000,
+			TTSEnabled:     false,
+			DefaultEngine:  "flite",
+			DefaultVoice:   "rms",
+			DefaultPitch:   50,
+		},
 	}
 	if err := yaml.Unmarshal(data, &cfg); err != nil {
 		return nil, err
@@ -180,6 +204,7 @@ func LoadConfig(path string) (*Config, error) {
 	if err := validateServoConfig(&cfg.CameraServo); err != nil {
 		return nil, fmt.Errorf("cameraServo: %w", err)
 	}
+	validateAudioConfig(&cfg.Audio)
 	return &cfg, nil
 }
 
@@ -219,6 +244,30 @@ func clampFloat(value, min, max float64) float64 {
 	return value
 }
 
+func validateAudioConfig(cfg *AudioConfig) {
+	if cfg.CaptureEnabled && cfg.CaptureDevice == "" {
+		cfg.CaptureDevice = "default"
+	}
+	if cfg.SampleRate <= 0 {
+		cfg.SampleRate = 16000
+	}
+	if cfg.Channels <= 0 {
+		cfg.Channels = 1
+	}
+	if cfg.Bitrate <= 0 {
+		cfg.Bitrate = 64000
+	}
+	if cfg.DefaultEngine == "" {
+		cfg.DefaultEngine = "flite"
+	}
+	if cfg.DefaultVoice == "" {
+		cfg.DefaultVoice = "rms"
+	}
+	if cfg.DefaultPitch <= 0 {
+		cfg.DefaultPitch = 50
+	}
+}
+
 func derivePublishURL(serverURL, roverName string, port int) (string, error) {
 	if roverName == "" {
 		return "", errors.New("missing rover name for publishUrl")

pi/roverd/media_env.go

@@ -9,24 +9,36 @@ import (
 
 const publisherEnvPath = "/var/lib/roverd/video.env"
 
-func UpdatePublisherEnv(cfg MediaConfig) error {
+func UpdatePublisherEnv(media MediaConfig, audio AudioConfig) error {
-	if cfg.PublishURL == "" {
+	if media.PublishURL == "" {
 		return fmt.Errorf("media publishUrl missing")
 	}
-	if cfg.VideoWidth <= 0 || cfg.VideoHeight <= 0 || cfg.VideoFPS <= 0 || cfg.VideoBitrate <= 0 {
+	if media.VideoWidth <= 0 || media.VideoHeight <= 0 || media.VideoFPS <= 0 || media.VideoBitrate <= 0 {
 		return fmt.Errorf("invalid media dimensions/bitrate")
 	}
 	if err := os.MkdirAll(filepath.Dir(publisherEnvPath), 0o755); err != nil {
 		return err
 	}
 	var buf bytes.Buffer
-	fmt.Fprintf(&buf, "PUBLISH_URL=%s\n", cfg.PublishURL)
+	fmt.Fprintf(&buf, "PUBLISH_URL=%s\n", media.PublishURL)
-	fmt.Fprintf(&buf, "VIDEO_WIDTH=%d\n", cfg.VideoWidth)
+	fmt.Fprintf(&buf, "VIDEO_WIDTH=%d\n", media.VideoWidth)
-	fmt.Fprintf(&buf, "VIDEO_HEIGHT=%d\n", cfg.VideoHeight)
+	fmt.Fprintf(&buf, "VIDEO_HEIGHT=%d\n", media.VideoHeight)
-	fmt.Fprintf(&buf, "VIDEO_FPS=%d\n", cfg.VideoFPS)
+	fmt.Fprintf(&buf, "VIDEO_FPS=%d\n", media.VideoFPS)
-	fmt.Fprintf(&buf, "VIDEO_BITRATE=%d\n", cfg.VideoBitrate)
+	fmt.Fprintf(&buf, "VIDEO_BITRATE=%d\n", media.VideoBitrate)
+	fmt.Fprintf(&buf, "AUDIO_ENABLE=%d\n", boolToInt(audio.CaptureEnabled))
+	fmt.Fprintf(&buf, "AUDIO_DEVICE=%s\n", audio.CaptureDevice)
+	fmt.Fprintf(&buf, "AUDIO_RATE=%d\n", audio.SampleRate)
+	fmt.Fprintf(&buf, "AUDIO_CHANNELS=%d\n", audio.Channels)
+	fmt.Fprintf(&buf, "AUDIO_BITRATE=%d\n", audio.Bitrate)
 	if err := os.WriteFile(publisherEnvPath, buf.Bytes(), 0o640); err != nil {
 		return err
 	}
 	return nil
 }
+
+func boolToInt(v bool) int {
+	if v {
+		return 1
+	}
+	return 0
+}

pi/roverd/media_supervisor.go

@@ -13,13 +13,14 @@ import (
 
 type MediaSupervisor struct {
 	cfg           MediaConfig
+	audio         AudioConfig
 	logger        *log.Logger
 	client        *http.Client
 	checkInterval time.Duration
 }
 
-func NewMediaSupervisor(cfg MediaConfig, logger *log.Logger) *MediaSupervisor {
+func NewMediaSupervisor(cfg MediaConfig, audio AudioConfig, logger *log.Logger) *MediaSupervisor {
-	if err := UpdatePublisherEnv(cfg); err != nil {
+	if err := UpdatePublisherEnv(cfg, audio); err != nil {
 		logger.Printf("media supervisor: update env failed: %v", err)
 	}
 	if !cfg.Manage || cfg.Service == "" {
@@ -35,6 +36,7 @@ func NewMediaSupervisor(cfg MediaConfig, logger *log.Logger) *MediaSupervisor {
 	}
 	return &MediaSupervisor{
 		cfg:           cfg,
+		audio:         audio,
 		logger:        logger,
 		client:        client,
 		checkInterval: interval,
@@ -45,7 +47,7 @@ func (m *MediaSupervisor) Start(ctx context.Context) {
 	if m == nil {
 		return
 	}
-	if err := UpdatePublisherEnv(m.cfg); err != nil {
+	if err := UpdatePublisherEnv(m.cfg, m.audio); err != nil {
 		m.logger.Printf("media supervisor: update env failed: %v", err)
 	}
 	if m.cfg.HealthURL == "" || m.client == nil {
@@ -76,7 +78,7 @@ func (m *MediaSupervisor) HandleAction(ctx context.Context, action string) error
 	if m == nil {
 		return errors.New("media supervisor disabled")
 	}
-	if err := UpdatePublisherEnv(m.cfg); err != nil {
+	if err := UpdatePublisherEnv(m.cfg, m.audio); err != nil {
 		return err
 	}
 	switch action {

pi/roverd/roverd.sample.yaml

@@ -37,3 +37,13 @@ cameraServo:
   homeAngle: 0
   nudgeDegrees: 2
   allowRawPulse: false
+audio:
+  captureEnabled: true
+  captureDevice: default
+  sampleRate: 16000
+  channels: 1
+  bitrate: 64000
+  ttsEnabled: false
+  defaultEngine: flite
+  defaultVoice: rms
+  defaultPitch: 50

pi/roverd/tts.go

@@ -0,0 +1,75 @@
+package roverd
+
+import (
+	"context"
+	"fmt"
+	"os/exec"
+	"strings"
+	"time"
+)
+
+func (c *WSClient) handleTTSPayload(payload *ttsPayload) error {
+	if payload == nil {
+		return fmt.Errorf("tts payload required")
+	}
+	if !c.cfg.Audio.TTSEnabled {
+		return fmt.Errorf("tts disabled on rover")
+	}
+	if payload.Speak == false {
+		return nil
+	}
+	text := strings.TrimSpace(payload.Text)
+	if text == "" {
+		return fmt.Errorf("tts text required")
+	}
+	if len([]rune(text)) > 512 {
+		text = string([]rune(text)[:512])
+	}
+
+	engine := strings.ToLower(strings.TrimSpace(payload.Engine))
+	if engine == "" {
+		engine = strings.ToLower(strings.TrimSpace(c.cfg.Audio.DefaultEngine))
+	}
+	if engine == "" {
+		engine = "flite"
+	}
+
+	voice := strings.TrimSpace(payload.Voice)
+	if voice == "" {
+		voice = strings.TrimSpace(c.cfg.Audio.DefaultVoice)
+	}
+	pitch := payload.Pitch
+	if pitch <= 0 {
+		pitch = c.cfg.Audio.DefaultPitch
+	}
+	pitch = clampInt(pitch, 0, 99)
+
+	ctx, cancel := context.WithTimeout(context.Background(), 12*time.Second)
+	defer cancel()
+
+	var cmd *exec.Cmd
+	switch engine {
+	case "espeak", "e":
+		args := []string{}
+		if pitch > 0 {
+			args = append(args, "-p", fmt.Sprintf("%d", pitch))
+		}
+		args = append(args, text)
+		cmd = exec.CommandContext(ctx, "espeak", args...)
+	case "flite", "f":
+		args := []string{}
+		if voice != "" {
+			args = append(args, "-voice", voice)
+		}
+		args = append(args, "-t", text)
+		cmd = exec.CommandContext(ctx, "flite", args...)
+	default:
+		return fmt.Errorf("unsupported tts engine: %s", engine)
+	}
+
+	out, err := cmd.CombinedOutput()
+	if err != nil {
+		return fmt.Errorf("tts exec failed: %w (%s)", err, string(out))
+	}
+	return nil
+}

pi/roverd/wsclient.go

@@ -71,6 +71,7 @@ func (c *WSClient) sendHello(ctx context.Context, conn *websocket.Conn) error {
 		MaxWheelSpeed: c.cfg.MaxWheelMMs,
 		Media:         c.cfg.Media,
 		CameraServo:   c.cfg.CameraServo,
+		Audio:         c.cfg.Audio,
 	}
 	c.log.Printf("sending hello (camera servo enabled=%v pin=%d)", msg.CameraServo.Enabled, msg.CameraServo.Pin)
 	return writeJSON(ctx, conn, msg)
@@ -150,6 +151,8 @@ func (c *WSClient) dispatch(ctx context.Context, msg *inboundMessage) error {
 			return fmt.Errorf("camera servo disabled")
 		}
 		return c.handleServoCommand(msg.Servo)
+	case msg.TTS != nil:
+		return c.handleTTSPayload(msg.TTS)
 	default:
 		return fmt.Errorf("unsupported command type: %s", msg.Type)
 	}

server/public/index.html

@@ -11,8 +11,8 @@
     <meta name="apple-mobile-web-app-status-bar-style" content="black-translucent" />
     <meta name="apple-mobile-web-app-title" content="Multi Roomba Rover" />
     <title>Multi Roomba Rover</title>
-    <script type="module" crossorigin src="/assets/index-A0W3Zfns.js"></script>
+    <script type="module" crossorigin src="/assets/index-Ch0OrDUK.js"></script>
-    <link rel="stylesheet" crossorigin href="/assets/index-DOoEkJaz.css">
+    <link rel="stylesheet" crossorigin href="/assets/index-BxUpjkzh.css">
   </head>
   <body>
     <div id="root"></div>

server/src/services/chatService.js

@@ -6,6 +6,7 @@ const { getRole } = require('./roleService');
 const { describeAssignment } = require('./assignmentService');
 const roverManager = require('./roverManager');
 const { getNickname } = require('./nicknameService');
+const { issueCommand } = require('./commandService');
 
 const RATE_LIMIT_WINDOW_MS = 8000;
 const RATE_LIMIT_MAX = 5;
@@ -66,6 +67,7 @@ function buildMessage(socket, text, meta = {}) {
     roverId,
     fromDiscord: Boolean(meta.fromDiscord),
     text,
+    tts: meta.tts || null,
   };
 }
 
@@ -73,7 +75,20 @@ function broadcastMessage(message) {
   publishEvent({ source: 'chat', type: 'chat:message', payload: message });
 }
 
-function handleIncoming({ text } = {}, socket, cb = () => {}) {
+function normalizeTtsOptions(raw = {}) {
+  if (!raw || typeof raw !== 'object') return null;
+  const speak = raw.speak !== false;
+  if (!speak) return null;
+  const engine = typeof raw.engine === 'string' && raw.engine.toLowerCase() === 'espeak' ? 'espeak' : 'flite';
+  const voice = typeof raw.voice === 'string' ? raw.voice.trim() : undefined;
+  let pitch = Number.isFinite(raw.pitch) ? Math.round(raw.pitch) : undefined;
+  if (typeof pitch === 'number') {
+    pitch = Math.max(0, Math.min(99, pitch));
+  }
+  return { speak, engine, voice, pitch };
+}
+
+function handleIncoming({ text, tts } = {}, socket, cb = () => {}) {
   const role = getRole(socket);
   // if (role === 'spectator') {
   //   cb({ error: 'Spectators cannot chat' });
@@ -104,12 +119,39 @@ function handleIncoming({ text } = {}, socket, cb = () => {}) {
     cb({ error: 'Message looks like spam' });
     return;
   }
-  const message = buildMessage(socket, clean, { fromDiscord: false });
+  const roverId = resolveRoverId(socket?.id);
+  const ttsOptions = normalizeTtsOptions(tts);
+  const message = buildMessage(socket, clean, { fromDiscord: false, roverId, tts: ttsOptions });
   logger.info('Chat message', { socket: socket.id, roverId: message.roverId });
   broadcastMessage(message);
+  maybeSpeak(socket, message, ttsOptions);
   cb({ success: true });
 }
 
+function maybeSpeak(socket, message, ttsOptions) {
+  if (!ttsOptions || !message?.roverId) return;
+  const record = roverManager.rovers.get(message.roverId);
+  const audio = record?.meta?.audio || {};
+  const ttsEnabled = Boolean(audio.ttsEnabled);
+  if (!ttsEnabled) return;
+  if (!roverManager.canDrive(message.roverId, socket)) return;
+  try {
+    issueCommand(message.roverId, {
+      type: 'tts',
+      tts: {
+        text: message.text,
+        engine: ttsOptions.engine,
+        voice: ttsOptions.voice,
+        pitch: ttsOptions.pitch,
+        speak: true,
+      },
+    });
+    logger.info('TTS sent', { roverId: message.roverId, engine: ttsOptions.engine, socket: socket.id });
+  } catch (err) {
+    logger.warn('TTS send failed', { roverId: message.roverId, error: err.message });
+  }
+}
+
 function sendExternalMessage({ text, nickname = 'Discord', role = 'admin', roverId = null }) {
   const clean = typeof text === 'string' ? text.trim() : '';
   if (!clean || clean.length > 256) {

server/src/services/roverManager.js

@@ -118,6 +118,7 @@ function getRoster() {
     maxWheelSpeed: record.meta?.maxWheelSpeed,
     media: record.meta?.media,
     cameraServo: record.meta?.cameraServo,
+    audio: record.meta?.audio,
     locked: record.locked,
     lockReason: record.lockReason,
     lastSeen: record.lastSeen,

webui/src/components/ChatPanel.jsx

@@ -25,13 +25,27 @@ function displayName(message) {
   return message.nickname || message.socketId?.slice(0, 6) || 'unknown';
 }
 
+const FLITE_VOICES = ['kal', 'rms', 'slt', 'ksp', 'bdl'];
+const ESPEAK_PITCHES = Array.from({ length: 10 }, (_, idx) => idx * 10);
+
 export default function ChatPanel({ hideInput = false, hideSpectatorNotice = false, fillHeight = false }) {
   const { session } = useSession();
   const { messages, sendMessage, registerInputRef, onInputFocus, onInputBlur, blurChat } = useChat();
   const [draft, setDraft] = useState('');
   const [sending, setSending] = useState(false);
+  const [speak, setSpeak] = useState(false);
+  const [engine, setEngine] = useState('flite');
+  const [voice, setVoice] = useState('rms');
+  const [pitch, setPitch] = useState(50);
   const canChat = session?.role !== 'spectator';
   const listRef = useRef(null);
+  const currentRoverId = session?.assignment?.roverId || null;
+
+  const rover = useMemo(
+    () => session?.roster?.find((entry) => String(entry.id) === String(currentRoverId)) || null,
+    [currentRoverId, session?.roster],
+  );
+  const ttsSupported = Boolean(rover?.audio?.ttsEnabled);
 
   const sorted = useMemo(() => messages.slice(-200), [messages]);
 
@@ -40,6 +54,22 @@ export default function ChatPanel({ hideInput = false, hideSpectatorNotice = fal
     listRef.current.scrollTop = listRef.current.scrollHeight;
   }, [sorted]);
 
+  useEffect(() => {
+    if (ttsSupported) {
+      setSpeak(true);
+    } else {
+      setSpeak(false);
+    }
+  }, [ttsSupported]);
+
+  const ttsPayload = useMemo(() => {
+    if (!ttsSupported || !speak) return null;
+    if (engine === 'espeak') {
+      return { speak: true, engine, pitch };
+    }
+    return { speak: true, engine, voice };
+  }, [engine, pitch, speak, ttsSupported, voice]);
+
   async function handleSend(event) {
     event.preventDefault();
     if (!canChat || hideInput) return;
@@ -47,7 +77,7 @@ export default function ChatPanel({ hideInput = false, hideSpectatorNotice = fal
     if (!clean) return;
     setSending(true);
     try {
-      await sendMessage(clean);
+      await sendMessage(clean, ttsPayload);
       setDraft('');
       blurChat();
     } catch (err) {
@@ -109,6 +139,52 @@ export default function ChatPanel({ hideInput = false, hideSpectatorNotice = fal
             placeholder={canChat ? 'Type a message…' : hideSpectatorNotice ? '' : 'Spectators cannot chat'}
             disabled={!canChat}
           />
+          {ttsSupported && (
+            <div className="flex items-center gap-0.5">
+              <label className="flex items-center gap-0.25 text-xs text-slate-300">
+                <input
+                  type="checkbox"
+                  checked={speak}
+                  onChange={(e) => setSpeak(e.target.checked)}
+                  className="accent-cyan-500"
+                />
+                <span>Speak</span>
+              </label>
+              <select
+                value={engine}
+                onChange={(e) => setEngine(e.target.value)}
+                className="field-input text-xs"
+              >
+                <option value="flite">flite</option>
+                <option value="espeak">espeak</option>
+              </select>
+              {engine === 'flite' ? (
+                <select
+                  value={voice}
+                  onChange={(e) => setVoice(e.target.value)}
+                  className="field-input text-xs"
+                >
+                  {FLITE_VOICES.map((v) => (
+                    <option key={v} value={v}>
+                      {v}
+                    </option>
+                  ))}
+                </select>
+              ) : (
+                <select
+                  value={pitch}
+                  onChange={(e) => setPitch(Number(e.target.value))}
+                  className="field-input text-xs"
+                >
+                  {ESPEAK_PITCHES.map((p) => (
+                    <option key={p} value={p}>
+                      pitch {p}
+                    </option>
+                  ))}
+                </select>
+              )}
+            </div>
+          )}
           <button type="submit" disabled={!canChat || sending} className="button-dark disabled:opacity-50">
             {sending ? '...' : 'Send'}
           </button>

webui/src/context/ChatContext.jsx

@@ -51,9 +51,9 @@ export function ChatProvider({ children }) {
   }, [playSound, session?.socketId, socket]);
 
   const sendMessage = useCallback(
-    (text) =>
+    (text, tts = null) =>
       new Promise((resolve, reject) => {
-        socket.emit('chat:send', { text }, (resp = {}) => {
+        socket.emit('chat:send', { text, tts }, (resp = {}) => {
           if (resp.error) {
             reject(new Error(resp.error));
           } else {