The questions we keep getting.

About 20 minutes the first time. You create an account, name your service, paste a list of languages your congregation speaks, and connect an audio cable from your sound board to the booth computer. We provide a step-by-step on the first call.

No. They scan a QR code or visit a short URL. The listener experience is a regular web page — no app, no login, no permissions beyond opening the browser tab.

No new microphone. The supported input is a direct audio feed from your sound board into the booth computer, so whatever mic your pastor already preaches with carries straight through. A phone or laptop microphone, or a speaker placed in front of the computer, is not supported — translation and voice quality depend on a clean direct feed.

Yes. Book a free live test. We’ll set you up with a temporary URL, run a real Sunday with you, and review the transcripts together afterward. Zero commitment.

We quote the measured number, not a ceiling. Every service is graded afterwards against the church’s own recording: on recent services, 89–93% of content words were heard right in English, and 92–94% of caption meaning carried into Spanish. We’re honest about the failure modes: jokes, idioms, and unfamiliar proper nouns are where humans still beat us.

Live captions in 190+ languages, with natural spoken voice in 60+ — the most coverage on Spanish, Portuguese, Mandarin, Korean, Vietnamese, Tagalog, Arabic, French, and Russian. New languages added on request, most within 48 hours.

Yes. Every account has a glossary editor. Add proper nouns, theological terms, and language-specific overrides. Glossaries apply at both transcription and translation, so "Yahweh" stays "Yahweh" across every language.

Speech-to-text handles most major accents well — Southern US, Indian English, African English, Spanglish, etc. For unusual dialects we recommend running a 10-minute test recording first; we can tune from there.

No. Your audio and transcripts are never used to train any model — ours or our vendors’.

You do. Recordings live in object storage tied to your account. You can export, download, or delete them at any time. We retain only what you ask us to retain.

Compliant. We’re happy to sign DPAs. Listener identity is not required and not stored by default.

No. MyOpenDoor Translate runs as a hosted service — your booth computer sends the audio feed and every guest joins from a web page. If on-premises is a hard requirement for your organization, email us and we’ll talk it through.

The booth reconnects by itself — no click from your volunteer — and if the hearing, translation, or voice service behind it has a hiccup, it swaps to a backup automatically. Listeners see a "reconnecting" badge and their page picks back up. Words spoken while the connection is down are not replayed, so a wired connection for the booth computer is the best insurance.

A moderator can pause, edit, or take the stream offline at any moment from the broadcaster console. You can also flag a transcript line after the fact and we’ll review it together to improve your glossary.

Faithfulness and control. Generic tools translate word-by-word instead of as whole sentences, quietly paraphrase or “smooth” what was said, and have no concept of your congregation’s vocabulary. MyOpenDoor Translate is purpose-built for live, in-the-room translation — every word delivered faithfully, with a glossary you control.

Fair concern. We’re a small founder-run company, not chasing venture growth at the cost of our customers. Your recordings and transcripts are yours — download them from the console any time, and they stay yours whatever happens to us.