Open Text-to-Speech and Choose Your Voice
Open the Text-to-Speech editor
From the Noiz sidebar, click Text-to-Speech under the Product section. The workspace splits into two areas:
Left panel — enter your script and manage speakers
Right panel (Settings) — choose voices, models, and generation options
Choose a voice
In the right panel, you'll see your currently selected voice, recently used voices, and shortcuts to Clone Your Voice or Design New Voice.
Click any voice to select it. Each voice card shows supported languages and tags for age group, gender, tone, or use case. To browse the full catalog, open Voice Library from the sidebar under Playground and filter by language, gender, use case, and more.
Select an AI model
Below the voice settings, choose your AI model. Noiz currently offers Emotion Engine-v2, optimized for natural expressiveness and emotional nuance. The selected model applies to the entire session.
Write Your Script in Text-to-Speech Editor
Single speaker
Click inside the main text area and type or paste your script. Each segment supports up to 1,000 characters. A character counter appears at the bottom of the panel.
Multiple speakers
To simulate a conversation or multi-voice narration:
Click + Add Speaker below your current segment.
A new segment block appears with its own speaker label and text area.
Select a different voice for each segment using the speaker dropdown.
Repeat for each new speaker or scene break.
Using distinct voices for different speakers makes dialogue-style content more engaging and easier to follow.
Emotion tags (supported voices only)
Some voices support emotion tags that guide the tone of individual sentences. Place a tag at the very start of a sentence:
[happy]: You know what? For the first time, everything feels right.
Tags placed mid-sentence or at the end will not take effect. You can also use preset prompt chips at the bottom of the panel — Emotion Boost, Sad Monologue, E-commerce Promo, Excited Life Blogger — for ready-made sample scripts.
Pause markers and voice presets
Use pause markers in your script to control breath points and pacing between sentences. Once you've found a voice, speed, and delivery setting that works for a project, save it as a named voice preset so future sessions start from the same configuration.
Generate Audio With Text-to-Speech Speak It
Once your script and settings are ready, click Speak It. Noiz automatically generates two versions of your audio.
Play both and choose the one that best fits your content — they may differ in pacing, emphasis, or emotional delivery. Each version has its own playback and download button.
If neither version feels right, click Speak It again. Slight variation between generations is intentional — the model introduces natural randomness for more human-sounding output.
Preview and Download Text-to-Speech Output
A persistent playback bar at the bottom of the workspace shows your current position, a scrubber, a spectrogram, and thumbs-up/down rating buttons.
When you're satisfied, click the Download button to save the audio file. Files export in high-quality audio format compatible with major editing platforms including Premiere, DaVinci Resolve, and GarageBand.
Get Better Results From Text-to-Speech
Clarity. Keep sentences concise. Shorter, well-punctuated sentences produce cleaner, more natural output.
Pacing. Use punctuation deliberately. A period or comma creates a natural pause in the generated audio. Combine punctuation with pause markers when you need a specific beat.
Preview first. Always listen before downloading to confirm the tone and pacing match your intent.
Language. For non-English scripts, select a voice explicitly trained for that language rather than relying on multilingual tags alone.
Save your work. The editor does not auto-save between sessions. Keep your scripts in an external text file or document.
Paste your opening paragraph, hit Speak It, and listen to both versions before generating the full script.