For decades a multilingual conference meant the same shopping list: soundproof booths at the back of the hall, a team of professional interpreters, and a bin of receiver headsets to hand out and collect. It worked, and it still works. It was also expensive, heavy to transport, and impossible to scale up at short notice.
There is now a lighter way, and the sensible position is not that one replaces the other but that most events were using the heavy option because it was the only one that existed.
The problem with the traditional setup
Booth interpretation is superb, and the cost structure is unforgiving. You pay for the booths, the receivers, the technicians who run them, and interpreters for every language — usually in pairs, because a simultaneous interpreter cannot work alone for more than about half an hour without quality falling away.
Three consequences follow, and every organiser will recognise them.
- Adding a language late is expensive. A request a fortnight before the event means finding two more interpreters and another channel, and the budget moves in a step rather than a slope.
- Breakout sessions go uncovered. The rig lives in the main hall. Six parallel workshops would need six rigs, so in practice the workshops get nothing — and workshops are where people actually talk.
- Small sessions cost nearly as much as large ones. The setup is the same whether forty people or four hundred are in the room.
The result is that organisers offer the two or three languages they can justify, and quietly accept that everyone else will manage in a second language.
The browser-based alternative
Instead of receivers, attendees use their own phones. Instead of booths, the speaker's microphone feeds a system that transcribes and translates as they talk, sending subtitles — and optional spoken audio — to each person in the language they choose.
Setup takes minutes from a laptop. Adding a language is a setting rather than a logistics project. And because there is no rig, there is nothing to move between rooms.
What you gain
- Every session covered. A breakout room of twenty gets the same treatment as the main stage. For most conferences this is the single largest change, because it doubles or triples the amount of the programme that is accessible at all.
- Languages on demand. Offer twenty and find out which five people actually use, rather than guessing in advance and being wrong.
- Nothing to hand out or collect. No deposit-taking, no queue at the desk, no counting receivers back in at the end of a long day, no replacing the ones that went home.
- A searchable record. The transcript of every session can exist in every language afterwards — which for many events turns out to be worth more than the live translation, because it is what lets someone who never attended find the talk months later.
What it asks of you instead
Being honest about the trade: the effort does not disappear, it moves.
- Audio becomes your responsibility. An interpreter in a booth can cope with a mediocre feed because a human fills gaps from context. A speech engine cannot. Every room needs a proper microphone on the speaker, and a room microphone will not do.
- Roving microphones for questions. Audience questions are the part every organiser forgets. Without a passed microphone, the answer appears in the transcript and the question does not.
- Network coverage in every room, including the basement one nobody visited during the site inspection.
- Panels need discipline. Four people talking over each other is the hardest material there is. Ask the chair to name who speaks next.
Practical notes for organisers
- Put the code on the badge. Lanyards are the one thing every attendee carries all day and looks at when bored. Better than the programme, and far better than a sign at the back.
- Brief the speakers in writing, a week ahead. Wear the headset, speak to the microphone, send us your proper nouns.
- Collect names and terms in advance. Organisation names, place names, technical vocabulary, the speakers' own surnames. On difficult audio a supplied word list makes a substantial difference; on clean audio it costs nothing.
- Test in each room during setup, not on the morning. Read a transcript in each one.
- Decide about the archive before you begin and tell the speakers. Some will want their session public; some will not.
When booths still make sense
To be fair to the older model, and because overselling this helps nobody: for diplomatic, legal or regulated settings where a certified human interpreter is required, booths remain the standard and should. The same applies where a single mistranslated sentence carries contractual or legal weight, and at events where a substantial part of the audience will not use a phone at all.
There is also a hybrid worth considering for larger conferences: human interpreters for the two main languages on the main stage, and browser-based translation for everything else — every breakout, every workshop, and the fifteen other languages in the room. That combination covers far more of the programme than either approach alone, usually for less than adding two more booth languages.
Try it at your next event
InMyTongue streams live subtitles and spoken translation to every attendee's phone, in the language they choose, straight from a browser with nothing to install — in as many rooms as you need. New accounts include free credit and no card is needed to try it. See how it works.