Audio Creator
Text-to-speech, voice cloning, music generation, and sound effects — all AI-powered.
Overview
Audio Creator bundles four separate audio generation tools into one screen. Each tool occupies its own tab and handles a distinct type of audio output. All generated audio is downloadable as MP3 or WAV and can be integrated with Video Creator for full audiovisual productions.
Text to Speech
Convert written text into natural-sounding speech. 30+ voices, adjustable speed and pitch. Preview before downloading.
Voice Cloning
Upload a 30-second audio sample and create a custom voice that can then be used for TTS generation.
Music Generation
Describe a mood, genre, tempo, and duration. AI generates original background music — royalty-free and unique each time.
Sound Effects
Describe a sound and AI generates it. From thunder cracks to office ambience to sci-fi soundscapes.
Screen previews
A quick look at the screens. Use the toggle in the corner of each preview to switch between the dark and light themes.
Text-to-speech workspace
Voice cloning from a sample
Music generation from a mood
Voice Clone
Cloning a voice
Open the Voice Clone tool in Audio Creator. In step 1, click Record and speak clearly for 5–30 seconds (a live waveform shows the level), or upload a .wav / .mp3 reference. Name the voice, confirm that you own it or have the speaker's permission, and click Clone voice. In step 2, choose the voice, type up to 4,000 characters and click Generate with cloned voice; you can play the result and download it as .mp3. Cloning runs on Fish Audio through the server, so it needs a FISH_API_KEY configured on the backend. Cloned voices are private to the Fish Audio account; this browser only remembers which ones you created.
Speech to Text
Transcribing audio
Open the Speech to Text tool in Audio Creator. Click Record to capture audio from your microphone (click Stop when finished), or click Upload audio file to pick an .mp3, .wav, .m4a or .webm file of up to 25 MB. Play the clip back, then click Transcribe. The transcript appears in a box you can copy. The audio is sent through the server to the Whisper model on the Modelion gateway; the API key never reaches your browser.
Text to Speech
Choosing the speech model
Text to Speech runs on the model selected in Settings → AI Models → Speech (TTS / STT) (Modelion or OpenRouter gateway models; free ones are listed first). Open the Text to Speech tool in Audio Creator, type up to 4,000 characters, enter the voice id supported by the model (for example alloy or adam), pick MP3, WAV, Opus, AAC or FLAC, and click Generate speech. You can play the result in place and download it. The text is sent through the server to the selected gateway provider.
The TTS tab lets you convert any written text into audio using one of 30+ AI voices. This is the fastest path from script to audio file — no recording equipment required.
Voice Selection
Browse voices by:
- Language — English, Spanish, French, German, Turkish, Japanese, and more.
- Gender — Male, Female, Neutral.
- Style — Conversational, Formal, News Anchor, Narration, Energetic, Calm.
Click a voice name to hear a short preview before selecting it for your conversion.
Adjustments
- Speed — 0.5× (very slow) to 2.0× (fast). 1.0× is the natural pace for most voices.
- Pitch — shift pitch up or down by up to 20 semitones. Use sparingly — extreme pitch shifts can degrade naturalness.
- Pauses — add markup tags like
<break time="1s"/>in your text to insert deliberate pauses for dramatic effect or clarity.
Long Text
For narrations longer than 500 words, split the text into paragraphs and generate each paragraph separately. This gives you finer control over pacing and allows you to regenerate individual sections without redoing the whole script.
Voice Cloning
Voice Cloning creates a custom AI voice from an audio sample of a real person speaking. The cloned voice can then be used as a TTS voice for any text.
Requirements for a Good Clone
- Minimum 30 seconds of clear speech. More sample = higher fidelity clone. 2–3 minutes is ideal.
- Clean audio — minimal background noise, no music, no other speakers. A quiet room recording is best.
- Natural speech — the sample should include varied sentence structures and natural pacing, not a monotone reading.
- Format — MP3, WAV, or M4A. Sample rate 16kHz minimum; 44.1kHz preferred.
Consent Required
Only clone voices for which you have explicit consent from the speaker. Cloning a voice without consent may violate privacy laws in your jurisdiction. TotalApp logs all voice clone creation events for audit purposes.
Once cloned, the voice appears in your Personal Voices list in the TTS tab and can be used exactly like any built-in voice.
Music Generation
Describe the music you need and the AI composes an original track. Generated music is royalty-free and unique — no two generations of the same prompt are identical.
Useful descriptors to include in your music prompt:
- Genre — jazz, classical, lo-fi hip hop, cinematic orchestral, electronic ambient, acoustic folk.
- Mood — uplifting, tense, nostalgic, mysterious, playful, melancholic, energetic.
- Tempo — "slow and relaxed", "medium tempo", "fast-paced and driving", or a BPM number like "120bpm".
- Instrumentation — "piano and strings", "solo acoustic guitar", "full orchestra", "synthesiser and drums".
- Duration — 15, 30, 60, or 120 seconds.
Example prompt: "Upbeat jazz trio, piano with double bass and brushed drums, 90bpm, warm and playful, 30 seconds, suitable for a product demo video background".
Sound Effects
Generate one-off sound effects by describing what you want to hear. The AI is good at both recognisable real-world sounds and abstract or fantastical effects.
Example prompts:
- "Thunder crack followed by rumbling that fades over 4 seconds"
- "Busy coffee shop ambience, espresso machine, background chatter, soft jazz in the distance"
- "Sci-fi door hiss opening, low mechanical whir"
- "Notification ding, short and pleasant, mobile app sound design style"
- "Footsteps on gravel, medium pace, 10 seconds"
Export and Integration
All audio from Audio Creator is exported as:
- MP3 — compressed, smaller file size. Good for web use and social media.
- WAV — lossless, larger file size. Use for professional production where quality is critical.
To pair audio with a video from Video Creator, download the audio from Audio Creator, then import it into the Video Creator timeline editor in the Audio Track slot. The Audio Track supports one background music track and one voiceover track simultaneously.