Guide
How to record a sermon for an accurate transcript
A sanctuary is built for people to hear, not for microphones. The same echo that carries a preacher's voice to the back row blurs it on a recording. A little setup decides whether the transcript is quick and clean, and with graded pricing it also decides what it costs.
Take the feed from the sound desk
If your church has a mixing desk, record from it. A soundboard feed carries the pulpit microphone directly, before the room adds echo and the congregation adds noise. Most desks have a recording or USB output; a cable into a laptop or a small recorder is enough.
If there is no desk, a lapel microphone on the preacher is the next best thing. Clip it about a hand's width below the chin, away from a tie or a Bible that might rub against it, and check the level during a loud passage so it doesn't distort.
Avoid the phone in the pew
A phone in the congregation hears everything the preacher's microphone doesn't need: the room's echo, children, fans, the PA hum and people shuffling. Recordings like that often grade C or D, and some words are lost however carefully they are transcribed.
If a phone is all you have, put it close to the pulpit or in front of a loudspeaker, not halfway back. Use flight mode, check storage and battery for the whole service, and record a short test first.
Readings and congregational responses
Scripture readings are often given from a lectern or the floor with no microphone. Ask readers to use one, or to stand close to the pulpit microphone. Prayers and announcements from the floor have the same problem.
If the congregation's responses matter to you, mic the congregation lightly: one microphone facing the room, mixed in low under the pulpit. Too much and every cough, child and conversation joins the recording.
Scripture references are transcribed as they were spoken. We don't look up a verse and paste in the text, and we don't correct a paraphrase to match a Bible translation, so the clearer the reference sounds, the better it reads on the page.
Split the service into separate files
Send the sermon, the Bible study and the announcements as separate files rather than one long recording of the whole morning. Each file is graded on its own, so a noisy section doesn't affect the grade of a clean one, and you only pay for the parts you need.
Worship sets are long, loud and rarely what people want in a transcript. Trim them out before you upload, or tell us where the sermon starts. You pay by the audio minute.
A YouTube or livestream link is fine too: paste it in the upload form and the video is imported for you.
What makes a recording grade A, B, C or D
Every upload is graded from A to D on how clearly the words can be recognised, background noise, recording level, distortion and overlapping voices. A soundboard feed or a well-placed lapel microphone usually grades A or B and costs the standard rate.
A phone in the pew, a very echoey hall or a busy question-and-answer session can grade C or D. Then only that file costs 30% or 70% more, and you see each file's grade and the reason before you pay.
Language, consent and what happens to the recording
Church recordings are transcribed in English first. If a service moves between English and Kiswahili, or is in another language, it is checked by a person before we quote. We transcribe what was said in the language it was said in; we do not translate.
Consent is the church's to obtain, including from parents or guardians when children are recorded in a Sunday school class, a choir or a dedication. We do not publish or reuse recordings, and we never use them to train AI models.
Audio and video are deleted 90 days after upload, or sooner if you ask. Send a list of names and places with the recording; supporting documents are free and help spellings come out right first time.