Real-time subtitles in multiple languages means one speaker, one microphone, and an audience that reads the talk in several languages at the same moment. Each attendee picks their own language and follows it on their own screen, usually a phone. The speaker changes nothing. The system listens once and writes many times.
One talk, many languages at the same time
Classic live captioning works in one language. Someone, human or machine, writes down what the speaker says, and the text appears on a screen. That already helps people who are deaf or hard of hearing, people in a noisy room, and anyone who reads a second language better than they hear it.
Multilingual subtitles add a step. The words spoken on stage are transcribed, then translated, then sent to each person in the language they chose. A Japanese engineer, a Brazilian buyer and a German journalist sitting in the same row read the same sentence, each in their own language, a couple of seconds after it was said.
The key words are "each in their own". A shared screen at the front of the room can carry one language, two at most before it becomes hard to read. Personal subtitles remove that limit. The number of languages no longer depends on the size of the screen.
What happens between the microphone and the phone
The audio is captured once. The sound usually comes from the room's mixing desk, the same feed that goes to the loudspeakers. A wireless lavalier or a USB table microphone works too when there is no desk. A clean signal matters more than anything else in the chain.
Speech recognition turns sound into text. The system writes each sentence in the language being spoken. Good systems also detect when the speaker changes, so a panel does not read like one long monologue.
Translation runs in parallel. Each finished sentence is translated into every language that someone in the room has selected. Adding a language does not slow the others down, because the work happens side by side.
Delivery goes to each device. The attendee opens a page in their browser, chooses a language, and the subtitles scroll as the talk goes on. In the simplest tools, this starts with a QR code, with no app to install and no account to create.
The whole trip takes a little time. With Glot, the translation delay is around two seconds, short enough to follow the argument, laugh at the joke and look up at the slide it refers to.
Why personal subtitles beat a single translated screen
Nobody waits for their language. On a shared screen, the organiser has to decide which language is shown. Everyone else is left out. On personal devices, the choice belongs to each attendee.
Reading comfort. A phone held at reading distance is easier to follow than a line of text at the far end of a large hall. People can scroll back if they missed a sentence.
Mixed audiences are the norm. International conferences rarely have two neat language groups. They have a long tail: a few attendees from many countries. Personal subtitles serve that long tail without extra planning.
The room screen still has its place. It works well for the main language of the event, for accessibility, and for people who left their phone in their bag. We cover that setup in our article on live captions on the big screen. The two approaches work together.
What to check before choosing a tool
The list of languages, and what it costs. Some tools price each language separately. Others include all of them. Glot covers more than 60 languages, all included in the price, so a late registration from a new country does not change the quote.
How attendees connect. Every extra step loses people. An app download in a conference hall with weak Wi-Fi is a real barrier. A QR code that opens a web page is the shortest path.
Speaker identification. On a panel, subtitles without names are hard to follow. Check that the tool shows who is speaking.
Specialist vocabulary. Product names, acronyms and people's names are where automatic translation stumbles first. A glossary prepared before the event helps a lot, as we explain in our piece on custom glossaries for live events.
Audio options. Some attendees prefer to listen rather than read. A translated voice in their earphones, with the option to switch back to the original sound, suits people who find reading tiring over a full day.
Data protection. Talks often contain unpublished information. Ask where the audio is processed and whether the provider complies with the GDPR.
Where multilingual subtitles reach their limits
Real-time machine translation is good, but it is not a human interpreter. Jokes built on wordplay, heavy irony or a speaker who mumbles away from the microphone will lose something on the way. Very sensitive settings, such as diplomatic negotiations or court hearings, still call for professional interpreters.
Audio quality is the other limit. If a speaker walks off with a handheld microphone pointed at the ceiling, no system will recover the words. A short sound check with the technician before the first session is the best investment you can make.
Finally, subtitles ask attendees to read. For a keynote that is fine. For a full day of workshops, offering a translated voice as well as text keeps people comfortable until the last session.
How to set it up for your event
The setup is lighter than most organisers expect. Connect the audio, create the session, then print the QR code at the entrance and put it on the opening slide. Attendees scan it as they sit down and choose their language. It takes a few minutes per room, and Glot closes each session with a written summary that attendees can keep.
The price is set per room, by the day or by the hour, and stays the same whatever the number of languages. You can compare the formulas on our pricing page.
One price per room, whatever the number of languages.
See Glot pricing