How it works

One event. Three ways to receive it.

Yazyk is an app one operator runs on a computer in the room. It takes a speech feed from the sound desk, translates it, and sends the result to the delivery paths you choose. Afterward its run log becomes an event receipt.

From spoken feed to a checkable record.

The input, language, and output stay explicit for the operator.

Longer guides cover setting up the computer, the feed from the sound desk, running a service, and what to do when the internet drops. Read the guides.

Sign in and the app is on your account's Downloads page, for Windows and for macOS. The macOS build is signed and notarized by Apple, so macOS asks you to confirm the first open instead of refusing it. Windows is the proven path; the macOS build has not run a live event. Ask us to set it up with you.

  1. Open the app at the desk

    Yazyk runs on one computer in the room, and every choice below is made on that screen by the person operating it.

  2. Choose the speech feed

    Use the sound desk feed that carries the speaker to the computer, then confirm the selected input responds to speech.

  3. Choose the language

    Select what is being spoken and the language to translate it into. Sixty languages are available, in any pairing.

  4. Choose delivery

    Send translated audio to an FM transmitter, make it available to phones, publish captions, or use the paths together.

  5. Read the receipt

    After the event, review coverage, listener delay, translated speech, and each recorded interruption.

Choose what fits your room.

The room output is useful when a venue already owns an FM system. Phones support listeners without a receiver. Captions serve people who need or prefer text.

FM

Computer to transmitter

Select the audio output, then run a male-to-male cable from the computer to the existing transmitter.

Browser

Phone audio

Listeners open your browser listener and hear the translated audio on their own device.

Text

Live captions

Captions can carry translated text. Same-language captions carry the source speech as an accessibility service, which is not translation.

The run log becomes an event receipt.

A readable after-event summary of what was delivered and what stopped.

Four things your team can inspect

  • How much of the event had a live translation pipeline
  • How much translated speech was produced
  • The delay measured at listener playback
  • Every interruption, with its time and duration