Live Interpretation

Hold a shortcut and speak into your microphone. EasyChat recognizes and translates your speech, then outputs it through a virtual audio device. Ten recognition languages are supported.

Requirements

Live Interpretation needs an ASR model and a virtual audio driver.

ASR model: Recognizes speech. Ten recognition languages are supported.
Virtual audio driver: Sends EasyChat's translated speech to a virtual microphone that other applications can use.

Install an ASR model

Before using Live Interpretation, install an ASR model.

Install a virtual audio driver

EasyChat uses VB-CABLE as its virtual audio driver.

Windows

Windows driver

Download the Windows virtual audio driver: VB-CABLE

macOS

macOS driver

Download the macOS virtual audio driver: VB-CABLE

Extract the driver archive, open VBCABLE_Setup_x64.exe, then select Install Driver.

How to use it

Add a Live Interpretation shortcut on the Shortcuts page.
Open Live Translation -> Live Interpretation, choose your settings, select the physical microphone you want to use, choose the language you will speak, then select Start.
In the application where you need Live Interpretation, set the microphone input device to CABLE Output.

Alternatively, set CABLE Output as the input device in your system's Settings -> Sound -> Input to use it in every application.

Hold the shortcut while speaking. EasyChat recognizes your speech, translates it, and outputs the result through the virtual audio device.

Note

When you hold down the hotkey, the software will emit a Beep sound, indicating that voice recognition is in progress;
When you release the hotkey, the software will emit a Beep sound, indicating that voice recognition has ended.
Once the translation is finished, the software will emit a Beep sound, indicating that the translation is complete.
There are three distinct Beep sounds in total, representing the start of recognition, the end of recognition, and the completion of translation.

After the translation is complete, TTS (Text-to-Speech) synthesis will occur (taking about 1~3 seconds), and the translated result will be output through the virtual audio device.