Frequently asked questions
The questions we hear most before and after events, roughly in order of how often they come up: capturing the audio, the venue network, how listeners are counted and how they listen, then delay, the venue screen, online meetings, transcripts and refunds.
What equipment captures the sound best?
The ways in, best first:
- The venue mixer: a MOTU M2 audio interface into a laptop (Mac or Windows). The steadiest, clearest sound; stage and audience-question mics all come in, and speakers wear nothing extra
- No mixer: speakers wear wireless lavalier mics such as a DJI Mic 3, or a USB mic such as a Shure MV88+ sits on the table, plugged into a laptop
- No laptop: a phone within half a metre of the speaker, listening with its own microphone. An iPhone in Safari has tested steady, and keeps listening with the screen locked or another app in front; Android phones differ a lot from model to model, so precheck with that very phone before the event, and don’t lock it while it sends
- No laptop and the speakers move about: an iPhone or iPad with a wireless lavalier, in our app 365UI Interpretation Audio. The receiver plugs into the phone; if it comes out, the app sends silence and says so rather than using the phone’s own microphone
- Not recommended: a mic plugged into a phone and used through a browser. On an iPhone, browsers (including Safari) record the phone’s own microphone instead, and the audio page refuses it; Android takes a while to find the mic, and the page cannot tell if it is unplugged
- Only a Mac laptop at hand: a MacBook’s own microphone works, placed like a phone within half a metre of the speaker. The page does not take a Windows laptop’s own microphone or a headset’s: they mostly hear the room around them.
- The whole event is on Zoom, Google Meet or Teams: no equipment at all; paste the meeting link on your event page and our bot joins to hear it.
Not sure? Ask the venue two things: do this room’s microphones go into a mixer, and can the mixer give us one line-level output? If both are yes, use the mixer.
The sound is poor or the translation is off: what usually causes it?
- The device is too far from the speaker: a phone’s or laptop’s own microphone hears everything around it. Put it within half a metre of the speaker, the closer the better; further away the translated voice slows down and nearby chatter gets translated too.
- Audience questions miss the mic: what no microphone hears is not translated. Pass the mic to whoever asks, or have the host repeat the question.
- The mixer’s feed is too quiet or muted: ask the venue AV for a feed with every speaker and question mic in it, ideally a post-fader aux or matrix send that does not follow the master; at the sound check, say a line into each mic. When the sound is low, the ingest page asks you to turn the GAIN knob on the sound card.
- The wrong sound source: pick the room’s mics and sound card from the device list on the ingest page; sharing a tab’s sound is only for meetings or streams running in a web page. Picked by mistake, listeners hear silence.
- No pause for the break: on a break, press “Start a break (pause interpretation)” on the ingest page, and in a room with worn mics switch them off, or private chat gets translated too.
- Speakers never pause: a short pause between sentences keeps the translated voice closer behind.
At the pre-event check, the helper follows the ingest page’s prompts one small step at a time until it says “Remote hears you clearly”.
What if the venue network is poor?
- The computer or phone sending the sound: if the venue Wi-Fi uses a sign-in page, keep a phone hotspot as backup. When the network drops, the ingest page says to sign in to the Wi-Fi again or join the hotspot, and it reconnects itself.
- Bandwidth is rarely the problem: each listener’s phone uses a few dozen kbps. With a full hall, the usual limit is how many devices the Wi-Fi takes at once. Ask the venue three things ahead: how many access points the room has, how many phones it has carried at once before, and whether the Wi-Fi needs a sign-in page.
- Listeners cannot connect: the page reconnects by itself; if it keeps failing, switch Wi-Fi or use mobile data. Local attendees on mobile data and visitors from abroad on the venue Wi-Fi (roaming data costs money) also takes load off the Wi-Fi.
- Visitors from mainland China: on international roaming from China Mobile, China Unicom or China Telecom, their data usually goes back through mainland China first, so the listening page may not open or may keep dropping. Ask them to use the venue Wi-Fi, or a local SIM or eSIM.
- If it breaks off: when an outage on our side passes 5% of a slot, the system refunds in proportion, no claim needed; problems caused by a venue network, equipment or venue audio not prepared as asked are not refunded.
How are listeners counted? With 1,000 at the event, do we need the 1,000 tier?
The listener count is the number of phones listening or reading captions at the same time, not everyone at the event; people reading the venue screen do not count. Often only part of the audience needs the translated voice while the rest follow the speaker and the screen, so choose the tier for that part.
- Say 1,000 people attend and two or three hundred of them need the translated voice: the 500 tier is enough, and everyone else reads the bilingual captions on the venue screen.
- The venue screen costs nothing extra and does not count as a listener.
- Once a room reaches its limit, extra phones cannot join, and there is no bill afterwards. If more people arrive than expected, raise the room’s tier on your event page at any time; it takes effect once the difference is paid.
- Rooms with voice for 900 or more listeners at once must be booked at least 24 hours ahead.
Do listeners need headphones? What if they hear nothing?
- Please listen with headphones, not the phone’s speaker, so as not to disturb people nearby or the speaker on stage. Without headphones, read the captions on the phone or the venue screen.
- The page says it is playing but there is no sound: check the headphones are connected and turn up the phone’s media volume.
| The page says | What to do |
|---|---|
| “Tap to enable audio” | Tap “Enable audio”. |
| “Waiting for the floor” | The session has not started yet or is on a break; audio will start automatically once a speaker begins. |
| “Reconnecting” | The network is unsteady and the page reconnects automatically; if it keeps failing, switch Wi-Fi or use mobile data. |
| “Interpretation in this room has ended for today” | This room is done for today; the page lists any other rooms still running. |
How far behind the speaker is the translated voice?
- In the samples on our home page the translated voice starts about two seconds after the speaker; those are clear recordings of one person speaking at an even pace.
- A real venue is a little slower: in the main hall of a forum in Toronto in October 2026, the median was 2.9 seconds and 95% were within 4.4 seconds. Long stretches without a pause are slower.
- Captions arrive about the same time as the voice, never after it.
How do the venue-screen captions work? What if there is only one screen?
- Open the screen link on the computer connected to the projector: it shows captions in two languages full screen, with no sound.
- With only one screen and the slides played from one computer: install the 365UI Captions app on that computer (Mac and Windows, free) and the captions float along the bottom of the slides. The venue helper guide shows how.
Does it work when the whole event is on Zoom, Teams or Google Meet?
- Yes, on Zoom, Teams and Google Meet alike: paste the full meeting link on your event page; five minutes before each slot our bot joins to hear the meeting and posts the listening link in the meeting chat. No computer, helper or equipment needed.
- Where a company’s Teams or Google Workspace keeps out people from other organizations and anonymous guests, the bot cannot join; check with the meeting organizer first.
- Zoom and Teams meetings can show captions in one language; in Google Meet and other meeting apps, share the venue-screen caption page. Attendees who want other languages use the listening page.
What do I hear when the speaker uses the language I picked?
The speaker’s own voice, not a second reading; when the speaker switches to another language, the translated voice comes back by itself. The captions then show the speaker’s own words.
How long are transcripts kept after the event?
- Choose one on your event page: kept 30 days (download transcripts and room audio, then deleted); keep the transcripts (free; the room audio is still deleted after 30 days); or not kept. Without a choice they are kept 30 days.
- When kept for 30 days, we email you a week before they are deleted (except monthly captions sessions). The After the event part of your event page downloads every transcript in one archive.
- After the event you can still switch to keeping the transcripts: tick the confirmation in After the event and turn it on.
How do cancellations and refunds work?
- Cancel at least 14 days before the event starts for a full refund; at least 7 but fewer than 14 days, half; fewer than 7 days, no refund.
- When an outage on our side passes 5% of a slot during the event, we refund in proportion; past 50%, the whole slot fee. The system works it out, no claim needed.
Can we try it first?
Yes. Press the free trial on the home page and, once signed in, open a 5-minute trial room with no card: speak at your computer, listen on your phone.
See also
AI assistant
Hi, I’m the 365UI AI assistant. I answer from our interpretation site, its guides and our company site, 365ui.com, and you can ask follow-ups. Try:
I do not answer questions about anything else, and I cannot look up an order or an event; for those, write to support@365ui.com.
We keep conversations for 90 days to improve the answers, without your IP address; see the privacy notice.