Skip to content

fix: honor explicit TTS voice selections #445

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
rogerchappel merged 1 commit into from
May 13, 2026
Merged

fix: honor explicit TTS voice selections #445

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 8 additions & 1 deletion src/app/api/tts/route.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@ import { readFile, unlink } from "node:fs/promises";
import { tmpdir } from "node:os";
import { join } from "node:path";
import { randomUUID } from "node:crypto";
import { normalizeAgentVoiceSettings, type AgentVoiceSettings } from "@/lib/tts-voices";
import { isExplicitServerVoice, normalizeAgentVoiceSettings, type AgentVoiceSettings } from "@/lib/tts-voices";

export const dynamic = "force-dynamic";

Expand Down Expand Up @@ -64,6 +64,13 @@ export async function POST(request: NextRequest) {
if (result) return result;
}

if (isExplicitServerVoice(voice)) {
return Response.json(
{ error: `Selected ${voice.provider} TTS voice is unavailable`, fallback: "none" },
{ status: 503 }
);
}

// 2. Try local TTS CLI
const localResult = await tryLocalTTS(text, voice);
if (localResult) return localResult;
Expand Down
69 changes: 43 additions & 26 deletions src/app/chat/page.tsx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,7 +37,13 @@ import {
publishAgentModeDiagnostic,
recordVoiceCrashBreadcrumb,
} from "@/lib/agent-mode-diagnostics";
import { DEFAULT_AGENT_VOICE_SETTINGS, normalizeAgentVoiceSettings, type AgentVoiceSettings } from "@/lib/tts-voices";
import {
DEFAULT_AGENT_VOICE_SETTINGS,
isExplicitServerVoice,
normalizeAgentVoiceSettings,
shouldUseDeviceTts,
type AgentVoiceSettings,
} from "@/lib/tts-voices";
import { isOpenClawHeartbeatAck, isOpenClawHeartbeatArtifact } from "@/lib/openclaw-heartbeat-artifacts";

/** Append <!--task_card --> markers for parsed task references not already embedded. */
Expand Down Expand Up @@ -628,10 +634,6 @@ function isNativeCapacitorApp() {
return platform === "ios" || platform === "android";
}

function shouldUseNativeSpeech(voice: AgentVoiceSettings) {
return voice.provider === "auto" || voice.provider === "browser" || voice.provider === "say" || voice.preferNative === true;
}

function blobToBase64(blob: Blob) {
return new Promise<string>((resolve, reject) => {
const reader = new FileReader();
Expand Down Expand Up @@ -2247,7 +2249,7 @@ export default function ChatPage() {

const voices = window.speechSynthesis.getVoices();
const selectedVoice = (resolvedVoiceSettings.voiceId || resolvedVoiceSettings.voiceName) &&
(resolvedVoiceSettings.provider === "browser" || resolvedVoiceSettings.provider === "say" || resolvedVoiceSettings.preferNative)
shouldUseDeviceTts(resolvedVoiceSettings)
? voices.find((v) =>
v.voiceURI === resolvedVoiceSettings.voiceId ||
v.name === resolvedVoiceSettings.voiceId ||
Expand Down Expand Up @@ -2282,6 +2284,8 @@ export default function ChatPage() {
) => {
const kind = options.kind ?? "filler";
const token = kind === "filler" ? fillerAudioTokenRef.current : undefined;
const usesExplicitServerVoice = isExplicitServerVoice(resolvedVoiceSettings);
const useDeviceSpeech = shouldUseDeviceTts(resolvedVoiceSettings);
try {
if (resolvedVoiceSettings.enabled === false) {
recordTtsBreadcrumb("play.voice-disabled", { kind });
Expand All @@ -2294,14 +2298,14 @@ export default function ChatPage() {
setIsPlayingAudio(true);
activeAudioKindRef.current = kind;

if (ttsModRef.current === "disabled") {
if (ttsModRef.current === "disabled" && !usesExplicitServerVoice) {
recordTtsBreadcrumb("play.disabled", { kind });
setIsPlayingAudio(false);
activeAudioKindRef.current = null;
return;
}

if (isNativeCapacitorApp() && (shouldUseNativeSpeech(resolvedVoiceSettings) || (ttsModRef.current === "native" && resolvedVoiceSettings.provider === "auto"))) {
if (isNativeCapacitorApp() && (useDeviceSpeech || (ttsModRef.current === "native" && resolvedVoiceSettings.provider === "auto"))) {
if (kind === "filler" && token !== fillerAudioTokenRef.current) {
recordTtsBreadcrumb("native-speech.stale-filler", { characters: text.length });
return;
Expand Down Expand Up @@ -2329,7 +2333,7 @@ export default function ChatPage() {
}

// If server TTS is unavailable, or this session prefers a device/browser voice, go straight to browser speech.
if (ttsModRef.current === "browser" || resolvedVoiceSettings.provider === "browser" || resolvedVoiceSettings.preferNative) {
if ((ttsModRef.current === "browser" && !usesExplicitServerVoice) || useDeviceSpeech) {
recordTtsBreadcrumb("play.browser-fallback", { kind });
playBrowserTTS(text, kind, token);
return;
Expand All @@ -2345,7 +2349,9 @@ export default function ChatPage() {
if (response.status === 503) {
recordTtsBreadcrumb("server.fetch.unavailable", { kind, status: response.status });
if (isNativeCapacitorApp()) {
ttsModRef.current = resolvedVoiceSettings.provider === "auto" ? "native" : "disabled";
if (!usesExplicitServerVoice) {
ttsModRef.current = resolvedVoiceSettings.provider === "auto" ? "native" : "disabled";
}
if (resolvedVoiceSettings.provider === "auto") {
await speakNativeVoiceText({
text,
Expand All @@ -2361,10 +2367,16 @@ export default function ChatPage() {
activeAudioKindRef.current = null;
}
} else {
// Server has no TTS backend, switch to browser mode
console.log("[TTS] Server unavailable, using browser speechSynthesis");
ttsModRef.current = "browser";
playBrowserTTS(text, kind, token);
if (usesExplicitServerVoice) {
console.log("[TTS] Server unavailable; no browser fallback for explicit server voice");
setIsPlayingAudio(false);
activeAudioKindRef.current = null;
} else {
// Server has no TTS backend, switch to browser mode
console.log("[TTS] Server unavailable, using browser speechSynthesis");
ttsModRef.current = "browser";
playBrowserTTS(text, kind, token);
}
}
return;
}
Expand Down Expand Up @@ -2437,15 +2449,20 @@ export default function ChatPage() {
activeAudioKindRef.current = null;
return;
}
// Network error — try browser fallback on regular web only.
playBrowserTTS(text, kind, token);
// Network error — try browser fallback on regular web only when it can honor the selected voice class.
if (usesExplicitServerVoice) {
setIsPlayingAudio(false);
activeAudioKindRef.current = null;
} else {
playBrowserTTS(text, kind, token);
}
}
}, [assignAudioObjectUrl, markFirstAudioStarted, playBrowserTTS, recordTtsBreadcrumb, resolvedVoiceSettings, revokeAudioObjectUrl]);

const prefetchTTS = useCallback(async (text: string) => {
if (isNativeCapacitorApp()) return;
try {
if (resolvedVoiceSettings.enabled === false || resolvedVoiceSettings.provider === "browser" || resolvedVoiceSettings.preferNative) return;
if (resolvedVoiceSettings.enabled === false || shouldUseDeviceTts(resolvedVoiceSettings)) return;
const response = await fetch("/api/tts", {
method: "POST",
headers: { "Content-Type": "application/json" },
Expand Down Expand Up @@ -2483,7 +2500,10 @@ export default function ChatPage() {
isSpeakingQueueRef.current = false;
return;
}
if (ttsModRef.current === "disabled") {
const usesExplicitServerVoice = isExplicitServerVoice(resolvedVoiceSettings);
const useDeviceSpeech = shouldUseDeviceTts(resolvedVoiceSettings);

if (ttsModRef.current === "disabled" && !usesExplicitServerVoice) {
recordTtsBreadcrumb("queue.disabled");
ttsQueueRef.current = [];
isSpeakingQueueRef.current = false;
Expand All @@ -2499,25 +2519,22 @@ export default function ChatPage() {

// Kick off prefetch of the NEXT sentence while this one plays
const upcoming = ttsQueueRef.current[0];
if (upcoming && ttsModRef.current !== "browser") {
if (upcoming && (ttsModRef.current !== "browser" || usesExplicitServerVoice)) {
prefetchTTS(upcoming);
}

try {
const browserSpeechAllowed =
!isNativeCapacitorApp() &&
"speechSynthesis" in window;
const usesExplicitServerVoice =
resolvedVoiceSettings.provider === "openai" ||
resolvedVoiceSettings.provider === "elevenlabs";
const useBrowserForFastStart =
!usesExplicitServerVoice &&
!hasStartedResponseAudioRef.current &&
browserSpeechAllowed &&
!resolvedVoiceSettings.preferNative &&
!useDeviceSpeech &&
resolvedVoiceSettings.provider !== "browser";

if (browserSpeechAllowed && (ttsModRef.current === "browser" || resolvedVoiceSettings.provider === "browser" || resolvedVoiceSettings.preferNative || useBrowserForFastStart)) {
if (browserSpeechAllowed && ((ttsModRef.current === "browser" && !usesExplicitServerVoice) || useDeviceSpeech || useBrowserForFastStart)) {
hasStartedResponseAudioRef.current = true;
// Browser TTS with queue continuation
if (browserSpeechAllowed) {
Expand Down Expand Up @@ -2545,7 +2562,7 @@ export default function ChatPage() {
utterance.rate = resolvedVoiceSettings.speed ?? 1.15;
const voices = window.speechSynthesis.getVoices();
const selectedVoice = (resolvedVoiceSettings.voiceId || resolvedVoiceSettings.voiceName) &&
(resolvedVoiceSettings.provider === "browser" || resolvedVoiceSettings.provider === "say" || resolvedVoiceSettings.preferNative)
useDeviceSpeech
? voices.find((v) =>
v.voiceURI === resolvedVoiceSettings.voiceId ||
v.name === resolvedVoiceSettings.voiceId ||
Expand Down Expand Up @@ -2586,7 +2603,7 @@ export default function ChatPage() {

// Check if we have a prefetched audio for this exact sentence
hasStartedResponseAudioRef.current = true;
if (isNativeCapacitorApp() && (shouldUseNativeSpeech(resolvedVoiceSettings) || (ttsModRef.current === "native" && resolvedVoiceSettings.provider === "auto"))) {
if (isNativeCapacitorApp() && (useDeviceSpeech || (ttsModRef.current === "native" && resolvedVoiceSettings.provider === "auto"))) {
revokePrefetchedAudio("native-queue-play");
recordTtsBreadcrumb("queue.native-speech.play.start", { characters: next.length });
const status = await speakNativeVoiceText({
Expand Down
13 changes: 13 additions & 0 deletions src/lib/tts-voices.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,3 +76,16 @@ export function normalizeAgentVoiceSettings(value: unknown): AgentVoiceSettings
preferNative: typeof record.preferNative === "boolean" ? record.preferNative : DEFAULT_AGENT_VOICE_SETTINGS.preferNative,
};
}

export function isExplicitServerVoice(voice: AgentVoiceSettings) {
return voice.provider === "openai" || voice.provider === "elevenlabs";
}

export function shouldUseDeviceTts(voice: AgentVoiceSettings) {
return (
voice.provider === "auto" ||
voice.provider === "browser" ||
voice.provider === "say" ||
(voice.preferNative === true && !isExplicitServerVoice(voice))
);
}
Loading