What is transcription?
Transcription is the speech in an audio or video file written down as text. It can be verbatim, with every hesitation, repetition and sound recorded, or clean, without filler words and punctuated like written text. A complete transcript also shows who is speaking and the timestamp of each line.
Also called: audio transcription, transcript
The same line, verbatim and clean
Verbatim works for research and records. Clean works for reading, quoting and publishing. Both keep the meaning.
| Type | Text |
|---|---|
| Verbatim | Um... we, we just started like that, you know? [laughs] No plan at all. |
| Clean | We just started like that, with no plan at all. |
Why transcription is the foundation
Text can be read, searched and copied. Once the speech is transcribed, an hour-long interview becomes a document you skim in minutes, and the right quote shows up in a search.
Almost everything that comes next is built on it: captions, chapters, paper edits, clips and summaries. A transcript with the timing of every word is what keeps captions in sync and makes a clip start on the right word.
Transcription in FilmScribe
We transcribe audio and video in 90+ languages, with automatic language detection, the timing of every word, speakers separated and sounds marked, like [laughter] and [music]. You fix anything right in the text, and the fix carries over to captions and exports. Download TXT, PDF, SRT or FCPXML, on every plan.
Transcription and speakers A file or a YouTube link, with no length limit, every word timed and every line attributed. See transcription →Getting a transcript you can trust
- 01
Agree on the style up front
Before work starts, settle verbatim or clean, how to mark unclear words, whether to include timestamps and speaker names, and how to spell recurring names. Changing halfway means reworking the file.
- 02
Mark what you cannot hear
Never guess at a word. Mark it as [inaudible] or [unclear] with the timestamp, so whoever reviews it can listen to that spot instead of trusting a guess.
- 03
Record for the transcript
Clean audio improves accuracy more than anything you fix later. A mic close to each speaker, a quiet room and no music under the speech prevent most errors.
- 04
Spot-check against the audio
Pick a few random passages and listen while reading. If you find several errors per minute, plan a full review; if you find none, focus on names and numbers.
Related terms
See all terms →Frequently asked questions
What is the difference between verbatim and clean transcription?
Verbatim records everything that was said, including hesitations, repetitions and sounds. Clean drops the filler and tidies the punctuation without changing the meaning. Research and legal work call for verbatim; articles, captions and scripts call for clean.
Is a transcript the same as a translation?
No. A transcript keeps the original language of the speech. A translation turns it into another language. On Studio and Agency, we translate transcripts into 9 languages with the same timing.
How long does it take to transcribe an hour of audio?
By hand, four to six hours for an hour of speech. In FilmScribe, the transcript is ready in minutes, and you just review it.
Captions, clips and paper edits. All in one place.
Upload a video or paste a YouTube link. We transcribe every word with its timing, and you take the result to social media or to your editor.
See the result on your own recording before adding a card.