Read aloud
TransOne can read text out loud, whether it is the original you typed or the translation it produced, in a voice that fits the language. This is pure text to speech: it speaks the text as it is, it does not translate it.
Speak buttons
A Speak button sits almost everywhere text appears:
- in the input box footer of the translation window;
- on every result card;
- on the dictionary, encyclopedia, study and summary cards;
- in History;
- in the Vocabulary notebook, where each word can also be read aloud as its flashcard turns over.
Click it and the text is read aloud. While it speaks, the button turns into Stop, so click it again to stop.
Speak selection and Speak clipboard
Two actions read text aloud without opening the window at all.
- Speak Selection reads the text you have selected in another app.
- Speak Clipboard reads the text you last copied.
Neither has a factory keyboard shortcut. To use them, either pick them from the menu bar icon, or record a shortcut of your own in Settings ▸ Shortcuts. See Keyboard shortcuts. Press the same shortcut again to stop.
There is also a quicker way to read a selection aloud: select the text, press ⌥W for Selection Actions, and choose Speak (or press 4).
Speak Selection needs Accessibility, the macOS permission that lets TransOne read text you have selected in another app. Speak Clipboard needs no permission at all, because copying already hands the text over. See Permissions.
Speech engines
Choose the engine in Settings ▸ Speech Synthesis (open Preferences from the menu bar icon, then choose Speech Synthesis in the sidebar).
Only one speech engine is active at a time. Turning another one on switches to it. The engine that is on cannot be switched off by itself, since TransOne always needs one: hovering its switch says "At least one engine must stay on. Turn on another to switch." The last engine in the list cannot be deleted either.
| Engine | Needs a key | Notes |
|---|---|---|
| System (On-Device) | No | Default. Offline, uses your Mac's own voices. |
| OpenAI | Base URL, Model, Voice, API Key | Supports voice instructions. |
| Tencent Cloud | Secret ID, Secret Key, Region, optional voice type | |
| Volcano Engine | App ID, Access Token, Cluster, Voice type | Create a speech app and enable TTS. |
| Google Cloud | API Key, optional voice name | Enable the Text-to-Speech API first. |
| Microsoft Azure | Subscription key, Region, optional voice name | Create a Speech resource. |
Voices and speed
The default System (On-Device) engine uses the voices built into macOS and works offline. It gives you full control:
- a voice picker, set to Automatic or any voice installed on your Mac;
- Speed, Pitch and Volume sliders;
- a Preview Voice button to hear your settings, which turns into Stop while it is playing.
When you leave the voice on Automatic and do not pin a specific voice, TransOne detects the language of the text and picks a matching voice, so the original and the translation are each read in the right accent.
System voices pause naturally between sentences, and a little longer between paragraphs, so a long passage is easy to follow. Numbers like "3.5", web addresses and abbreviations like "e.g." do not cause a pause. The pauses get shorter as you raise Speed.
TIP
Want more or better voices? macOS lets you add more voices in its own settings, under spoken content. Any voice you install there shows up in the voice picker here.
Related
- Keyboard shortcuts to give Speak Selection a shortcut.
- Selection actions to read a selection with ⌥W.
- Permissions to understand what Speak Selection needs.