Skip to content
Guideline
4 Jun 2021

Transcription Guidelines

Amberscript Academy Transcription Guidelines
Topics
Automatic transcription
Manual transcription

Do you want to have an audio file manually transcribed? You will be offered a choice between full verbatim or clean read transcription. What results can you expect per transcription type? And what guidelines do Amberscript’s transcribers follow?

This guide will also assist you in deciding which form of transcription best suits your needs and will shed some light on the transcription guidelines followed by the transcribers.

Clean read and full verbatim: Distinctions

Clean read (default): Goal – capture what has been said. End result: Grammatically correct transcript, with the good reading flow.

Examples: Organization/company meetings, municipality meetings, presentations, etc.

Full verbatim: Goal – capture how something has been said. End result: Precise transcript, as close to audio as possible.

Examples: Research interviews like patient-doctor conversations, speech analysis, legal documents like court hearings, focus group interviews where the emotions of the interviewee play an important role, etc.

English transcription guidelines in a nutshell:

  • Clean read transcriptions do not include: Speech errors, false starts, repetitions, slang words, stutters, filler words, etc.
  • In full verbatim transcriptions, the text is transcribed exactly as it sounds including speech errors, false starts, filler words, slang words, repetitions, and stutters.
  • The numbers zero up to and including nine are written out (i.e. one, three, nine)
  • The number 10 and above are written in digits (10, 11, etc.)
  • If a number is at the start of a new sentence, it is written out fully.
  • Times and dates are written as digits.

Punctuation guidelines to be followed:

.Full stop/periodEvery sentence ends with a full stop.
?Question markDo questions end with a question mark?
!Exclamation markUse an exclamation mark for emphasis!
,CommaCommas are used for short pauses, lists, and to improve reading flow.
EllipsisIf there is a pause … use ellipsis to indicate it.
Single dashU-use a single dash to indicate word interruptions and for hyphenations such as color-coded.
Double dashUse a double dash to mark incomplete sentences and–
“”Quotation marks“I’m a punctuation enthusiast!” she exclaimed.

Transcription tags used:

The following tags are included in a transcription when they are relevant for the transcription or context.

Full Verbatim Clean Read
Non-speech sounds like Mm-hmm or Uh-huh (affirmative)Mm-hmmYesYes
Audio file is impossible to understand after listening three times, at high volume and lower speed (0.7)[inaudible 00:00:00- 00:00:00]YesYes
Filler words like um, uh, er and ah[uh]Yes No
Music is played loudly and clearly – if relevant to context[music]Yes No
Video is played|video|Yes No
Continuous laughter for several seconds[laughter]Yes No
Continuous coughing for several seconds[coughing]Yes No
Long period of silence (3 seconds or longer)[silence 00:00:00- 00:00:00] Yes No
If relevant to transcription/context[crying]Yes No
If relevant to transcription/context[yelling]Yes No
Audience applauds speaker[applause]Yes No
When multiple people are speaking over each other causes any one of the speakers to become inaudible[crosstalk 00:00:00- 00:00:00] YesYes
When the audio includes non-English parts[foreign language 00:00:00- 00:00:00]YesYes
Our services allow you to create text and subtitles from audio or video.
  • Upload audio or video file
  • Automatic or manual speech to text
  • Edit and export text or subtitles