Video Transcription Services

Video transcription services convert videos to text, speaker by speaker, timecoded so that any line can be traced back to the picture. Espresso Translations transcribes video and audio files from £5 a minute of runtime and quotes free of charge within 1 hour. Translating the finished text is a separate line on the quote: £0.10 a word over 2 to 3 working days, with a 24-hour express tier for urgent work, or £30 a certified page of up to 250 words turned round in 24 to 48 hours. A second linguist revises every translation against its source under ISO 17100:2015 — a standard that governs translation services and says nothing at all about transcripts.

From £5 a Minute of Video

SRT, VTT or an Editable Document

Speaker Labels and Timecodes

Translated Into 150+ Languages

150+ Languages

Quote Within 1 Hour

Espresso Translations on Google
Espresso Translations on Trustpilot
Espresso Translations ISO 17100:2015 certification for translation services

Ask Us for a Video Transcription Quote

We’ll get back to you within 1 hour

Forbes logo, an Espresso Translations client
IMDB logo, an Espresso Translations client
Jaguar logo, an Espresso Translations client
Universal logo, an Espresso Translations client
National Geographic logo, an Espresso Translations client
EY logo, an Espresso Translations client
Sixt logo, an Espresso Translations client
Amazon logo, an Espresso Translations client
EDF logo, an Espresso Translations client
FREENOW logo, an Espresso Translations client

What does a video transcription service do? It turns a recorded video into text — what was said, attributed to the person who said it, with timecodes tying every line back to the picture. Espresso Translations transcribes video and audio files from £5 a minute of runtime, charged on how long the file runs rather than by the word, and returns either an editable transcript in DOCX, PDF or TXT, or an SRT or VTT caption file, or both from a single pass. The free quote arrives within 1 hour and names the turnaround that file allows. Translating the text is quoted on its own, at £0.10 a word or £30 a certified page of up to 250 words. The point UK buyers most often have back to front is that a caption file is not a transcript: captions are cut into timed cues to be read off the screen, and leave out what the reader of a record needs — which is why an accessibility requirement usually asks for both.

What Do You Get From a Video Transcription Service?


A document holding everything that was said in the video, attributed to whoever said it and timecoded throughout. It is not a summary and not a set of notes: the transcript is the record, and anything circulated afterwards is drafted from it. Video transcripts carry one thing an audio transcript cannot, which is the picture as corroboration — a name caption, a slide, a document held up to the lens — and a video transcriber uses all of it to settle what the sound alone leaves open. Where two voices genuinely cannot be separated, or a word is lost under a passing lorry, the transcript says so at that timestamp rather than inventing a plausible line.

Three decisions shape the finished document, and settling them before the work starts is cheaper than revisiting them afterwards. The first is verbatim or clean read. The second is the labelling convention — real names where you can send a running order or a participant list, otherwise consistent roles such as Presenter, Speaker 2 or Interviewer. The third is where the timecodes fall: at every change of speaker, which suits a record somebody will have to navigate, or at fixed intervals, which suits continuous narration. None of it asks you to change the workflow you already have. The file goes out, the document comes back in the format you asked for.

A verbatim transcript keeps the hesitations, the repetitions and the false starts, because what is being preserved is what was said rather than what the speaker meant. A clean read takes the filler out and reads as prose; parts of the market sell the same thing as intelligent verbatim. Captions are a third document again — the same words, but cut into timed cues, split across lines for reading on a screen, and shortened wherever the speech outruns the eye. Ask for a transcript when somebody is going to read or quote it, ask for a caption file when the words go on the picture, and ask for both when an accessibility requirement is driving the job, because neither file does the other one’s work.

Three neighbouring services get confused with this one often enough to be worth stating plainly. Espresso Translations does not offer interpreting: an interpreter speaks, live, while an event is happening. Translation renders a written text into another language. Transcription writes down a recording after the event, which is also what makes a second pass over the words possible at all. Live captioning — text appearing while people are still talking — is a fourth thing and not this one. Everything on this page begins once a video or audio file exists.

Which File Do You Need: a Transcript, Captions or Subtitles?


It depends entirely on what the video has to do next, and the five common answers want five different files. People who set out to convert video to text usually have one of them in mind and ask for another, which is the commonest avoidable cost in this work: a caption file cannot be filed as a record, and a reading transcript cannot be loaded into a player. Tell us the destination and the format follows from it.

What the video is for, and which file that means
What you want to do with itThe file we deliverWhat that file has to carry
Read, quote or file the recordAn editable DOCX, PDF or TXT transcriptSpeaker labels throughout, timestamps at each change of speaker, and a marked gap at anything unresolved
Put captions on the screen in the original languageAn SRT or VTT sidecar fileCue in and cue out times, lines split where the sense breaks, and a reading speed an audience can actually keep up with
Publish the video to an audience in another languageA translated subtitle file, one per languageThe source file’s cue timings, with every line re-cut to fit a language that may run a fifth longer than English
Record a voiceover, or brief a dubbing studioA timecoded script, speaker by speakerTimecodes a voice artist can work to, and a pronunciation note against every name, place and technical term
Meet an accessibility requirementA caption file and a full text transcript togetherNon-speech sound described, every speaker identified, and the whole audio content represented rather than summarised
Uploading a video file for a free transcription quote
Second pass over a video transcript against the picture
Project manager pricing a video transcription on the runtime of the file
Video transcript delivered with speaker labels, timecodes and an SRT caption file

SRT and VTT: the Two Caption Files Almost Everything Accepts

SubRip (.srt) and WebVTT (.vtt) are the two sidecar formats worth asking for, and between them they cover nearly every player, platform and editing suite in use. An SRT file is about as simple as a format gets: a cue number, a start and end time, the text, a blank line, repeat. WebVTT is the one the web platform itself understands, loaded through the HTML video element’s track element, and it carries styling and positioning that SRT has no way to express — useful when burnt-in titles already occupy the bottom of frame. Both are plain text, so both stay editable after delivery: you can correct a name in a cue without coming back to us. We deliver either, or both, and we will match the cue conventions of a file you are already using if you send us an example.

Burnt-In Subtitles or a Sidecar File?

A sidecar file sits alongside the video and the viewer switches it on; burnt-in text is part of the picture and cannot be switched off, translated or corrected without a new export. Sidecar wins almost every time — it is accessible, it is searchable by the platform, it can be swapped for another language, and it can be fixed. Burning in earns its place in exactly two situations: a social feed that plays silently and strips caption tracks, and a delivery specification that demands it. We supply the caption file either way, because burning in is a video editing step performed on your master, and a master is the one thing we would rather not touch.

What a Transcript Does That a Caption File Cannot

A caption file is written to be read in two-second glances, so it is cut short, split across cues and stripped of anything that would slow the eye down. That is the right decision for the screen and the wrong one for a record. An editable video transcript keeps the full sentence, names the speaker every time the voice changes, and runs as continuous readable text, which is what makes it quotable in a report, searchable and usable as evidence of what was actually said. It is also the file a translator works from, because a translator needs the whole sentence before deciding how it ends. When a video matters in both ways — on screen and on file — the honest answer is both documents, produced in one pass over the same audio so that they cannot drift apart.

Who Sets a Captioning Standard in the UK, and What Nobody Sets

Two regimes matter here and they apply to different people. Broadcast television is governed by Ofcom’s Code on Television Access Services, drawn up under section 303 of the Communications Act 2003, which sets how much of a channel’s output has to be subtitled, signed and audio described; the figures differ by channel and audience reach, and they are obligations on the broadcaster rather than on a supplier. Everything published on a website instead runs through the Public Sector Bodies (Websites and Mobile Applications) (No. 2) Accessibility Regulations 2018, SI 2018/952, which oblige public sector bodies to meet an accessibility requirement that the Government’s own guidance expresses as WCAG level AA. For video that means WCAG 2.1 success criteria 1.2.2, captions for prerecorded media, at Level A, with 1.2.1 and 1.2.3 covering the text alternative — which is the transcript. The deadlines have long passed: 23 September 2019 for sites published from 23 September 2018, 23 September 2020 for everything older, and 23 June 2021 for mobile applications. There is one exemption people miss, and it is narrow: prerecorded time-based media published before 23 September 2020 is out of scope, so the archive is exempt and the next thing you publish is not. What nobody publishes is an accuracy figure. There is no British standard for transcription or captioning accuracy, no ISO service standard for either, and no percentage a supplier can hold itself to and be compared on — so this page quotes none, and a supplier that quotes one without naming the test it came from has told you nothing you can check.

Which UK Organisations Order Video Transcription?


Anyone who has to make a video readable, searchable, accessible or usable in another language. In practice the files arrive from universities with lecture capture, public bodies with webcasts and published video, broadcasters and production companies working through post-production deliverables, marketing and internal communications teams, and solicitors and employers holding footage that may end up in a file. The subject decides the vocabulary and the vocabulary decides the pace, so each job goes to a transcriber who already knows the terms.

Universities, Lecture Capture and the Recorded Teaching Archive

Recorded teaching arrives in volume rather than one file at a time: a module’s worth of lecture capture, a conference’s keynotes, a series of recorded seminars. Most universities are public sector bodies for the purposes of the 2018 Accessibility Regulations, so published recordings carry the same captioning and text-alternative obligations as any other public sector video, and the Equality Act 2010 duty to make reasonable adjustments sits underneath that for any student who needs one, whether or not the recording is published.

What we deliver for teaching material is usually a caption file for the player and an editable transcript for the module handbook, produced in one pass. Terminology is the real work: a lecture on pharmacology, tort or structural engineering is dense with words that a general transcriber will approximate and a subject-matter one will spell correctly, and a drug name or a case citation transcribed by ear is worse than a marked gap. Send us the reading list, the slide deck or a glossary and the names go in right the first time.

Public Bodies, Webcasts and Published Video

Councils, NHS trusts, regulators, agencies and arm’s-length bodies all publish video, and all of them sit inside the 2018 Accessibility Regulations. Local government has a second reason to hold a text record: the Openness of Local Government Bodies Regulations 2014 made it lawful for anyone attending a meeting of a local government body to film, record and report it, and the webcast most councils now run is the authority’s own version of the same record. A searchable transcript of a planning committee or a full council meeting is what turns a year of recordings into something a clerk, a journalist or a resident can actually use, where a folder of video files can only be re-watched.

We produce the caption file for the player, the full text transcript for the minute book and website, and a translation where the authority serves a community in another language. Two practical notes. Published video needs the speakers identified by name and role, because a reader six months later will not recognise a voice; and a transcript can be redacted line by line, which a recording cannot, so it is the file to work from when something has to be withheld before publication.

Broadcasters, Producers and Post-Production Deliverables

Production companies and broadcasters order transcripts at two points: during the edit, when a searchable text of every rush is what makes a 60-hour shoot navigable, and at delivery, when the distribution specification asks for a script and a subtitle file. Ofcom’s access services code governs how much of a channel’s output must carry subtitling, signing and audio description, so for a broadcaster the caption file is a compliance deliverable rather than a nice-to-have, and it arrives on a schedule.

A timecoded, speaker-attributed script is also the document a localisation chain is built on: subtitles in each territory’s language, a dubbing script if the programme is being revoiced, and a voiceover script where a narrator is being replaced. Each of those is a different cut of the same transcript, which is the argument for having the transcript made properly once rather than three times. We work from the delivered master or from a viewing copy with burnt-in timecode, whichever you can send.

Marketing, Training and Internal Communications Video

A marketing or training video is usually transcribed for three reasons at once: captions, because most social video plays with the sound off; searchable text, because a page carrying the words of the video can be found by somebody who would never find the video; and reuse, because the transcript is the raw material of the blog post, the knowledge-base article and the e-learning module that follow. Training video adds a fourth: a text version of a mandatory course is what lets a learner with a hearing impairment, or one working in a noisy depot, complete it at all.

For internal communications the recording is often an all-hands briefing or a recorded town hall, and the transcript is what reaches the people who could not attend. Where a workforce reads in several languages, the transcript is translated and the subtitle file is re-cut for each one, which is a cheaper and far more reliable route than asking people to follow a second language on the fly.

Solicitors, Employers and Footage That May End Up in a File

Recorded investigation meetings, video-recorded witness statements, site and incident footage, and video calls held during a grievance or disciplinary process all produce files that end up in a bundle or an employment tribunal file. The requirement there is a full verbatim transcript with speakers attributed and a timecode on every exchange, because the dispute is usually about one sentence and a reader has to be able to go back to it. One hard boundary: do not send us a recording you made of a court hearing, including a remote one you attended — section 85B of the Courts Act 2003 has made that an offence since 28 April 2022, and a transcript of proceedings runs through the official route instead. Where a recording of a hearing is legitimately in your hands, court transcription services is the page that covers it.

On a video where people switch languages, multilingual transcription writes each contribution down in the language it was spoken in before any language translation begins, so the record carries the language as well as the content. Espresso Translations works in 150+ languages through a network of 2,000+ linguists, each with at least five years of professional experience, and is ATC certified and an ATC member.

How Is Video Transcription Priced in the UK?


On runtime, from £5 a minute: a video is charged on how long it runs, not on how many words are inside it. That matters more on video than on anything else, because density varies wildly — a two-hander interview and a ten-person panel of the same length contain very different amounts of speech, and runtime pricing puts that risk on us rather than on you. A 48-minute lecture is quoted on 48 minutes whether the lecturer talked throughout or spent twelve of them on a video clip.

What moves a quote off the floor is the recording and the document you want from it, rather than a list of supplements: how many people speak and how much they overlap, how the audio was captured, verbatim against clean read, where the timecodes fall, whether you want a transcript, a caption file or both, and whether a translation follows. Send the file and an itemised quote comes back within 1 hour, naming the turnaround that recording allows.

Delivery formats are yours to choose. The transcript comes as DOCX, PDF or TXT — editable in every case — and captions as SRT or VTT, cued to your file. We accept the audio and video files you already have: MP4, MOV, WebM, AVI and MKV on the video side, MP3, WAV and M4A on the audio side, a video link where the file is too large to email, and physical media or an older tape format where the recording predates all of that. An accurate video transcript does not depend on the container the footage arrives in, only on what was recorded inside it.

Translation sits on its own line, because not every transcript needs one. Standard work runs at £0.10 a word over 2 to 3 working days, with a 24-hour express tier for urgent jobs, and same-day delivery is available on short items that can be translated and revised inside a working day. A certified translation is £30 a page of up to 250 words, turned round in 24 to 48 hours, with the signed certificate of accuracy included in that page price rather than charged on top. Where a destination outside the United Kingdom insists on a sworn translation, that is £55 a page over 7 to 14 working days. Subtitling has its own floor of £5 a video minute, and voiceover is quoted per project because the cost depends on the artist, the studio and the length of the script rather than on runtime alone.

The figures below hold whatever the recording turns out to be. For a set of files — a term of lectures, a season of episodes, a quarter of webinars — tell us how many there are and how long each one runs, and the quote covers the whole set.

Video transcription and transcript translation: UK prices and delivery
ServicePriceDelivery
Transcribing the video, by the minute of runtimeFrom £5 a minuteStated in the quote for your file
Translating the transcript or the subtitle fileFrom £0.10 a word2–3 working days
Certified translation, per pageFrom £30 a page (up to 250 words)24 to 48 hours
QuoteFreeWithin 1 hour

Other Recordings We Transcribe

Video is one subject inside our transcription services, which also cover interview transcription of research, press and panel recordings, legal transcription, conference call transcription of recorded multi-party calls, academic transcription of lectures and research recordings, medical transcription, document transcription of handwritten and scanned material, and dictation typed up from a single voice. Turning video and audio to text is one team’s work here: audio transcription of a sound-only recording is the same job without a picture, an audio recording from a dictaphone or a phone reaches the same team, and video to text transcription differs only in having the screen as corroboration. It is human transcription throughout — a person produces every line and a second person checks it against the recording — and all of it is quoted on how long the audio runs. Where a file is urgent rather than large, quick transcription is the tier to ask for.

How Does a Video Become a Finished Transcript?


In four stages, and the first one decides the other three. We transcribe video and audio by the same method, so what the recording is — one camera or several, a lapel microphone or a room mic, two speakers or a panel of nine — is what the quote is built on, so it belongs at the start rather than being discovered halfway through.

78% PLACE ORDER

Step 1: Send the Video, and Tell Us What It Is For

Upload your file through the secure form, or send us a link to it where the file is large: MP4, MOV, WebM, AVI and MKV are all fine, as are MP3, WAV and M4A if you have already pulled the audio off, and a private online video link works as well as an attachment. Three things sent with it save real money. A running order or cast list, so speakers go in under their own names. A glossary of product names, drug names, place names and acronyms, spelled as you spell them. And the destination — reading transcript, caption file, subtitle translation, dubbing script — because that is what decides the format and the conventions.

Step 2: The Quote, and the Six Things It Settles

A project manager watches part of the recording before pricing it, and the quote settles six things: runtime, how many voices there are, how the audio was captured, which document or documents you want, where the timecodes fall, and whether anything is being translated afterwards. It reaches you within 1 hour, itemised, with a turnaround stated for that file rather than promised in advance of seeing it.

ORIGINAL EN A 文 TRANSLATED AR FR DE ES JA ZH AR CERTIFIED TRANSLATION

Step 3: Transcription, Then a Second Pass Against the Picture

A transcriber works through the recording, going back over the overlapping stretches until they are resolved or recorded as unresolved, and using what is on screen — a name caption, a slide, a document held up to camera — to settle spellings that audio alone cannot. A second person then checks the text against the video rather than against the first draft, which is the only way a mis-heard figure or a mis-attributed line is reliably caught. Nothing is filled in to look complete. Where a recording carries sensitive information, the method does not change but the access list does.

Step 4: Delivery, Captions, and What Happens to Your File

The transcript arrives in the format you asked for, with the labelling and timecode conventions agreed in the quote, and the caption file arrives cued to your video so it loads without adjustment. A translation follows where you need one. Confidentiality is handled by NDA and encrypted transfer, access is limited to the transcriber, the reviewer and the project manager, and we delete our copies on request once the job closes. If you have a retention period of your own to keep, the master stays with you rather than living on a supplier’s server.

Can You Transcribe a Video to Text Yourself?


Yes, and for some files you should. There are three routes that cost nothing but your time, and knowing where each one stops is the whole point of asking. The first is the automatic captions a platform generates: upload your video, wait a few minutes, and it will generate a video transcript as a caption track, which you can open and copy the text out of — a YouTube video is the usual case, and a YouTube link alone is enough to pull the track where one exists. The second is dictation: playing the file while a document’s speech-to-text feature listens, which works on a quiet single voice and on nothing more demanding than that. The third is manual transcription — typing it out yourself with a foot pedal or keyboard shortcuts, which is slow, as accurate as your attention, and the honest baseline the other two are measured against. Anyone searching for how to transcribe video to text is weighing exactly these three against paying somebody, and the sensible answer depends on the recording rather than on principle. The quickest way to transcribe video is rarely the cheapest once the repairs are counted.

What they have in common is where they stop. All three convert audio to text in a single video-to-text pass with no second listen; none labels a speaker reliably once a third voice joins; none knows a technical term it has never met; and none can tell you which lines it was unsure about. The platform route has one more problem that catches people out, because what you copy out is a stream of short cues with no punctuation, no paragraphing and no speaker attribution, so turning it into readable prose is itself an afternoon’s work. Count the real cost in minutes per minute of runtime: a transcriber working properly spends several minutes on every minute of audio, and so will you, plus the manual typing. So if you need to transcribe a webinar you only want a rough searchable text of, do it yourself. For a lecture series, a tribunal file, a published video or anything a reader will quote, the hours spent repairing free output are usually longer than the job would have taken to do properly, and the accurate text is what you were buying in the first place.

Where an AI Transcript Breaks, and Why It Is Always the Same Minutes

Automatic transcription handles the easy minutes well and fails on exactly the ones that matter, which is the single most useful thing to know about it. Speech recognition degrades when two people talk across each other, when the microphone is across the room, when a speaker has a strong regional accent or is working in a second language, when the terminology is specialist, and when background noise — traffic, air conditioning, a room full of people — sits on top of the voice. It produces a confident, fluent, plausible line in all five cases, because a transcription tool built on speech recognition has no way of knowing that it is guessing, and a fluent wrong word is far more dangerous in a record than a marked gap. That is also why comparing transcription software on how it handles a clean studio interview tells you nothing: every option handles that file, and none of them handles the other one. A person goes back, listens again, uses the argument and the turn-taking to work out who spoke, checks a name against what is on screen, and leaves a marked gap at the timecode where the audio genuinely does not resolve — which is what separates an accurate transcript from a fluent one. That is the difference you are buying, and it is concentrated in a small number of minutes in every file.

Which Other Espresso Services Pair With a Video Transcript?


A video transcript is usually a step rather than a destination — towards subtitles, a translated record, a revoiced version or a certified document. Whichever of these matches the job in front of you is the better place to carry on.

Certified Translation Services

Is the transcript, or its translation, going to a university, a tribunal or the Home Office with a certificate attached? Certified Translation Services covers that route, at £30 a page of up to 250 words.

Legal Translation Services

Is the footage part of an investigation, a bundle or a contract dispute? Legal Translation Services takes the language side once the transcript exists, and keeps the terminology consistent across every file in the matter.

Subtitle Translation Services

Is the video going out to an audience that reads another language? Subtitle Translation Services takes the transcript and produces a cued subtitle file per language, at its own floor of £5 a video minute. Subtitling is not translation with timecodes bolted on: every line has to be re-cut to a reading speed in a language that may run a fifth longer than the English did.

Voiceover and Audiovisual Services

Is the video being revoiced rather than subtitled? Voiceover works from the timecoded script the transcript produces, and is quoted per project rather than per minute, because the artist, the studio time and the script length all move the figure. The wider audiovisual translation page sets out how transcript, subtitle file and voice script fit together on one production.

Localisation Services

Is the video one piece of a product, a campaign or a course that has to work in another market? Localisation services covers the rest of it — interface text, on-screen graphics, dates, units and the examples that only make sense in one country — so the subtitled video does not arrive beside an app that is still in English.

Transcription Services

Something sound-only, or in a single language that is not English? The Transcription Services hub sets out the rest of the range, and video translation is the page to read if the finished video, rather than the transcript, is what you need in another language.

2,000+

Linguists

ISO 17100

Certified quality standard

1 hour

Free quote

150+

Languages Supported

What Else Do UK Buyers Ask About Video Transcription?


What does it cost to transcribe a video?

From £5 a minute of runtime, metered on how long the file plays rather than on how much is said inside it, so a crowded panel and a single-presenter piece of the same length are charged alike. A caption file and a reading transcript produced from the same pass are quoted together rather than twice. If the text then has to be translated, that is £0.10 a word over 2 to 3 working days, with a 24-hour express tier, or £30 a certified page of up to 250 words turned round in 24 to 48 hours. Quotes cost nothing, arrive within 1 hour and set out the turnaround available for your file.

How long does it take to transcribe a video?

Several times the runtime, and the multiplier depends on the recording rather than on effort. A clean two-voice interview moves quickly; a nine-person panel with crosstalk, a dense technical lecture or audio captured on a camera’s built-in microphone all mean going back over passages repeatedly, and then there is the second pass against the picture before anything is delivered. That is why the turnaround is stated in the quote for your file, after a project manager has watched part of it, instead of being promised in advance. Urgent work goes on the express tier and the quote says what that allows for your recording.

Which video or audio file formats can I send?

MP4, MOV, WebM, AVI and MKV on the video side, and MP3, WAV and M4A if you have already extracted the audio. Where the file is too large to email, send a link to it from whatever storage you use. Older material is fine too: we have worked from tape formats, optical discs and digitised archive footage, and if you are unsure whether a file will open, send a short sample first and we will tell you before you commit to anything. Audio or video makes no difference to the price, which is set by runtime.

Will I get an SRT file or a document?

Whichever you ask for, and both is a common answer. An editable transcript comes as DOCX, PDF or TXT, with speaker labels and timecodes; captions come as SRT or VTT, cued to your file so they load without adjustment. Say what the video is for when you send it — reading, filing, captioning, subtitle translation, a dubbing script — and the format follows from that. If you already use a particular cue convention, send an example and we will match it.

Can you transcribe video to text in another language?

Yes, and the order of the two jobs matters. Whatever was spoken is written down in that language first, which keeps the record faithful and shows who switched language and when. Only then is the text translated into whichever of our 150+ languages you need to read. Transcription starts from £5 a minute of runtime; translation is quoted apart from it at £0.10 a word, or £30 a certified page of up to 250 words, and a second linguist revises the translation against its source under ISO 17100:2015, which is a translation standard rather than a transcription one.

How do you handle several speakers in one video?

Each change of voice is labelled and timecoded, and sending a running order or a cast list is what makes the labels real names rather than Speaker 1 and Speaker 2. Video helps here in a way audio alone does not: a name caption, a seating position or somebody turning to camera often settles an attribution that the sound leaves ambiguous. Where two people genuinely talk across each other and neither can be separated, the transcript records that at the timecode instead of attributing a line on a guess.

Is a transcript enough to meet the accessibility regulations?

Usually not on its own. For prerecorded video with sound, WCAG’s captions criterion, 1.2.2, sits at Level A and asks for captions, while 1.2.1 and 1.2.3 cover the text alternative that a transcript provides — so the two files answer different criteria and most public sector publishing needs both. The Public Sector Bodies (Websites and Mobile Applications) (No. 2) Accessibility Regulations 2018 are what make this binding on public sector bodies, and the Government’s guidance expresses the standard as WCAG level AA. Whether a particular video is in scope is a question for your accessibility lead; we supply the caption file and the transcript, and we will not tell you that you are compliant.

Can you add the subtitles to the video for me?

We deliver the caption file rather than a re-exported video, and that is deliberate. A sidecar SRT or VTT can be switched on, swapped for another language and corrected; text burnt into the picture can be none of those things without a fresh export from your master, and a master is the one file we would rather leave in your hands. Burning in is a video editing step your editor or platform performs in minutes using the file we send. Where the destination is a social feed that strips caption tracks, tell us and we will cut the lines shorter so they sit well once they are on the picture.

Why not just use the automatic captions?

Use them where nothing turns on a sentence. They cost nothing, they arrive in minutes and they are a reasonable searchable index of a clean, single-voice recording. They degrade in five predictable places: overlapping speech, a microphone far from the speaker, a strong accent or second-language English, specialist terminology, and background noise. In all five an AI transcription engine produces a fluent, confident, wrong line rather than flagging a doubt, and a plausible wrong word in a record is worse than a marked gap. What you also get back is a caption track, not a document — short cues with no punctuation, no paragraphs and no speaker labels — so tidying it into something readable is itself a job.

Is there a published accuracy standard for video transcription?

No. There is no British standard for transcription or captioning accuracy, no ISO service standard that sets one, and no test a buyer can hold two suppliers to and compare. The percentages that circulate in this market come from individual practitioners’ examinations or from vendors’ own testing, not from any delivery tolerance anyone is bound by. So we quote no accuracy figure and this page states none. What we will tell you is the method: a person transcribes, a second person checks the text against the recording rather than against the first draft, and anything the audio does not resolve is left marked at its timecode.

Should I ask for verbatim or a clean read?

Ask for verbatim where the words may be quoted, challenged or read in a tribunal, because verbatim keeps the hesitations, repetitions and interruptions that a dispute often turns on. Ask for a clean read where the transcript is for reading, publishing or reuse, and the filler is noise — some of the market sells that as intelligent verbatim, and it is the same thing. If the transcript is going to become captions, a clean read is almost always the right call, because a caption line has no room for a false start.

Can you transcribe a recording of a court hearing?

Not one you recorded yourself. Section 85B of the Courts Act 2003 has made it an offence since 28 April 2022 to make an unauthorised recording of court proceedings transmitted to you, or of a person remotely attending them, and the same conduct is a contempt of court; transcripts of proceedings run through the official route instead, and Espresso Translations claims no official or authorised court transcription status. Where a recording of a hearing is legitimately in your hands, or where the file is your own investigation meeting rather than a hearing, our court transcription and legal transcription pages cover that work.

Do you provide interpreting for a live event?

No. Espresso Translations does not offer interpreting. People search for the two services together, so the distinction is worth having plainly: an interpreter speaks, live, while an event is happening, and live captioning puts text on screen while people are still talking. Everything on this page starts after a recording exists, which is also what makes a second pass over the words possible. A video you have already filmed is squarely the work we do, whatever language it was filmed in.

How do I get the quality of the recording right before filming?

Get the microphone close to the speaker and everything else follows. A lapel or boom microphone beats a camera’s built-in one by a wide margin, and a quiet room beats a good microphone in a noisy one — traffic, air conditioning and a hard-floored space with echo are the three that cost the most time. Beyond that: ask people not to talk over one another and to leave a beat before replying, keep speakers facing the microphone, record a separate track per person if your kit allows it, and say the names of participants at the top so there is an audio reference for the labels. None of this is a condition of the work; it is simply what makes a recording cheaper and faster to transcribe, and we will tell you honestly after listening if a file is going to be difficult.

Contact Us


Get in touch with our team for professional translation services. We're here to help you with all your translation needs.

Email
info@espressotranslations.com

Phone
+44 203 488 1841

Espresso Translations Google Reviews rating
Espresso Translations Trustpilot rating
ISO 17100 certified translation services

Request a Quote Now

We’ll get back to you with a competitive quote.