What a room needs for AI interpretation on attendees’ phones

No booth and no rented receivers. Each room needs a laptop, a way for the sound to get in and an on-site helper, plus a network, the speaker list and a notice for the room.

What you don’t need

No interpretation booth, no rented receivers, and no app for attendees: they scan a QR code with their own phones, pick a language in the browser, and bring their own earphones if they want the voice.

A laptop

One laptop with Chrome per room, Mac or Windows. A personal computer is easiest; avoid a work laptop locked down by company IT. Keep it plugged in with the lid open all day.

A way for the sound to get in

A room with a mixer: the venue’s AV team gives one line-level mono output from the mixer carrying every live microphone, the audience-question mic included, into the laptop through a USB audio interface (we use a MOTU M2). Keep walk-in music and speechless video out of that feed, and never route the interpretation back to the speakers.

A quiet room with no mixer: speakers wear wireless lavalier mics, or a USB mic sits on the table, plugged straight into the laptop (a wireless lavalier receiver can also plug into an iPhone or iPad running our free app 365UI Interpretation Audio); or a phone listens with its own microphone.

Right after ordering, decide how each room takes in sound, then buy what that needs, ideally a week ahead.

An on-site helper

One helper per room running at the same time, with no technical background needed: plug in, open the audio web page, and turn one knob on the interface as the page prompts (in a room on mics, move a mic closer or further). The page walks them through a soundcheck.

A network

The audio laptop needs a steady connection. If the venue Wi-Fi has a sign-in page, keep a phone hotspot as a backup.

Speakers and names

2 days before, send us the speakers’ names, titles and organizations in Chinese and English, and the event’s own organizations, sponsors and products; pasting from Excel works. We build a glossary from them, and the engine follows it for names.

A notice for the room

Before the start, tell speakers and attendees that the interpretation is done by AI and that microphone audio is processed in the United States and Singapore. A line on the opening slide, a sign by the audience-question mic and one sentence from the host are enough; the text is ready on your event page after you order.

Timeline

Read next

Try it free for 5 minutes

Markdown