Video Translation Services
Espresso Translations builds those files to the conventions British viewers read without noticing — two lines at a time, a reading speed the audience can keep up with, line breaks made where the sentence breaks, and captions moved when they would sit over a face or over text already on screen. Subtitling is quoted per minute of running time, in writing, before anything starts.
ISO Certified 17100
Native Translators
Transparent Pricing
Twice Proofread
150+ Languages
Quote in 1 Hour
Request a Translation Quote Now
We’ll get back to you within 1 hour
Do you need translated subtitles or SDH? Translated subtitles carry the dialogue in another language; SDH, delivered as closed captions, stays in English and adds speaker identification and the sounds a deaf or hard-of-hearing viewer cannot hear. Most UK jobs need both. Espresso builds either to the conventions British viewers read without noticing, in 150+ languages under ISO 17100:2015 with a second linguist on every file, from £0.10 per word for scripts and transcripts, quoted free within 1 hour.
What should a UK video translation supplier get right?
Eight things decide whether professional video translation produces a file that is usable here rather than merely accurate. They are listed below.
2,000+
Linguists
ISO 17100
Certified quality standard
1 hour
Free quote
150+
Languages Supported
How does a video translation job actually run?
Work starts from a locked master and whatever text already exists. Where there is no transcript, the audio is transcribed and timecoded first; where a caption file already exists, it is checked against the picture before a word is translated. The translation process then works to the timings rather than to the page, because a subtitle that reads correctly and cannot be read in the time available has failed.
Condensing is part of the work, not a shortcut. English into German or French runs longer on the page, so the linguist writes to the reading speed instead of rendering word for word. A second linguist then revises the file against the source and against the picture, so linguistic quality is judged separately from the timing; in and out points are checked at the shot boundaries, and the deliverables are exported in whichever formats the destination requires.
Why does captioning matter so much in Britain?
Three separate pressures drive it. The Equality Act 2010 requires employers, service providers and educational bodies to make reasonable adjustments for disabled people, and an uncaptioned training or information video is a plain example of failing to. Public sector websites and apps have to meet WCAG 2.1 AA under the 2018 accessibility regulations, and that standard requires captions on pre-recorded video. Ofcom’s Code on Television Access Services sets subtitling, signing and audio description requirements for licensed broadcasters.
None of those rules binds a company’s marketing film, but together they set the standard British viewers are used to, and procurement teams increasingly write accessibility clauses into tenders. Separately, most social video is watched with the sound off, so a burned-in caption track is often the only version of a film the audience receives at all.
Who asks for video translation here?
Broadcasters, video production companies and independent producers delivering to a UK channel or a streaming platform. Universities, colleges and training providers captioning lecture capture, e-learning modules and the video inside a multilingual training course, where the terminology has to stay identical from the first module to the last. Employers putting induction, safety and compliance video in front of a workforce that does not all read English at the same speed. Public bodies working to the accessibility regulations. That variety of video is why the brief differs in each case, and the deliverable list differs with it.
The other half of the work runs outwards, towards international audiences. A British company with one film and six markets does not want six films. It wants one master, a set of translated subtitle tracks, and voiceovers wherever the script carries more detail than a viewer can read while watching.
How long does a video translation take?
How long it will take is settled in the quote, against the running time of the file and the deliverables you have asked for, rather than published as a window that would not survive contact with a real job. Translation of a transcript or an existing subtitle file follows our standard delivery of 2 to 3 working days, with 24-hour express where a transmission date will not move. Transcription, spotting and timing are scheduled on top of that, and a voiceover or dubbed version adds a recording session and a mix. The date is confirmed in writing before work begins.
What moves it is usually the material rather than the language. A file with no transcript has to be transcribed first. Poor location sound slows transcription down. Several target languages at once means several revision passes. On-screen text that has to be rebuilt, and burned-in versions that cannot be corrected after delivery, each add a production stage. So does every round of client sign-off, which is the one part of the schedule we do not control. An early warning and a clean transcript go a long way towards protecting the date.
What actually goes wrong with video translation?
The failures below are the ones that come back on a QC report or from a viewer.
- Reading speed: The subtitle is accurate and nobody can read it. A line has to stay on screen long enough for the audience to take it in, which is why a faithful rendering of fast dialogue is often the wrong answer and a condensed one is the right one.
- Text expansion: English into German, French or Spanish runs longer on the page. Dropped into the same timings without condensing, the subtitle either spills to a third line or clears the screen before it has been read.
- Line breaks in the wrong place: Breaking between an article and its noun, or between a verb and its object, makes a subtitle harder to read even when every word in it is correct. The BBC guidelines break at the grammatical joint instead.
- Speaker identification: In an SDH file the viewer cannot hear who is talking. UK broadcast practice solves that with colour and with name labels, and a file that drops them is unusable for the audience it was made for.
- Sound that carries meaning: A knock at the door, a phone, a shift in the music. Where the story turns on a sound, an SDH file carrying dialogue alone has lost the plot point rather than merely trimmed it.
- Position clashes: Subtitles default to the bottom of the frame, which is where lower thirds, straps and on-screen captions already live. Moving the subtitle off them is part of the job, not an afterthought.
- Frame rate and timecode: A file timed against a 25 frames-per-second master drifts once it is laid onto a master conformed at a different rate. The drift is invisible at the head of the programme and glaring by the end of it.
- Burned-in captions: Once a caption is rendered into the picture it cannot be corrected, swapped for another language or switched off. Anything burned in has to be signed off before the render rather than after it.
- Americanised output: Machine subtitling defaults to US spelling, US date order and US measurements. On a file made for a British audience that reads as an error in the first line and undermines everything after it.
- Specialist terminology: Clinical, legal and engineering material has fixed wording, and a subtitle that paraphrases it to hold the reading speed changes what the film says. Expert video work of that kind goes to a linguist who already works in the field.
What do captions and translated subtitles get you?
The gains below are the ones British clients ask us to evidence.
- Reasonable adjustments evidenced: Captioned induction, training and information video is one of the clearest ways to show that a workplace or a service has been made usable by deaf and hard-of-hearing people.
- Public sector conformance: Pre-recorded video on a public sector site needs captions to reach WCAG 2.1 AA, and an audio-described version to go further than that.
- Sound-off viewing: Most social video is watched muted. A burned-in caption track is frequently the only version of a film the audience actually receives.
- Text you can search: A timed transcript gives a search engine, an internal knowledge base and an auditor something to read. A video file on its own gives all three of them nothing.
- Audiences already at home: Britain is multilingual before anybody crosses a border. Subtitling the same film into the languages spoken in your own catchment reaches people the English version does not.
- One master, several markets: Subtitle tracks and voiceover versions hang off a single edit, so an export version costs a deliverable rather than a second production.
- Tender and framework requirements: Accessibility clauses turn up routinely in public and corporate procurement, and a supplier who can deliver SDH and translated tracks to spec is answering the question the buyer asked.
- Training people finish: Captions help a viewer reading in a second language as much as they help a viewer who cannot hear the audio, and both groups complete a course they can follow.
- Fewer repeat questions: A product or process video that can be read as well as heard cuts the follow-up that would otherwise land on a support desk.
- An archive you can reuse: Caption files and timed transcripts are assets. Re-cut the film next year and the text already exists to conform the new version against.
What does video translation cost in the UK?
Two figures cover most of it: subtitling from £5 per video minute, and transcription from £5 per audio minute, both measured against the running time of the finished file. Voiceover and dubbing carry no published rate, because a recorded voice brings a voice artist, a studio and a mix that a caption file does not; what they cost follows the running time, the number of voices the script needs and whether the original voices are being replaced. Where a script or transcript is translated as text rather than timed to picture, it is quoted at £0.10 per word. Where a transcript has to be produced as a signed document for a tribunal bundle or a regulatory file, it is priced by the page and the figure is in the quote before work starts.
Translation for video content is priced against the material as much as against the minutes. Four things move a video quote more than anything else: whether a timed transcript already exists, how many target languages come off the same master, whether a burned-in render is needed as well as a caption file, and how specialist the content is. Express delivery is priced against the schedule it actually requires and stated in the quote, never added afterwards as a percentage. Every line is itemised in writing before work begins.
What changes the price of a video job?
The cost drivers are listed below.
- Running time: Subtitling and voiceover are quoted by the minute of finished programme, so the duration of the master is the first number in any quote.
- Whether a transcript exists: A timed, corrected transcript removes a whole production stage. Material that arrives as picture and sound alone has to be transcribed and spotted before a word is translated.
- Which deliverable you need: A translated caption file, an SDH track, a voiceover read over the original audio and a dub matched to the picture are four different jobs at four different costs, in that order.
- Language pair: Common pairs cost less because more qualified linguists work in them, and a rare pair costs more for the same reason. Both are priced before work starts rather than discovered afterwards.
- Text expansion: A language that runs longer than English needs more condensing to hold the reading speed, and that shows up in the subtitling rate rather than in a word count.
- Audio quality: Overlapping speech, heavy background noise and remote-recorded contributors all slow transcription down. Clean, high-quality location sound is the cheapest thing a client can supply.
- Number of languages: Each additional language is a separate translation and a separate revision pass. Transcription and spotting are done once at the start, which is what stops the cost multiplying straight through by the number of versions.
- On-screen text and graphics: Titles, straps and captions baked into the edit have to be rebuilt rather than subtitled, which brings video editing into a translation job.
- Burned-in versions: Rendering captions into the picture produces a separate master for every language, and each one has to be viewed and signed off before it goes out.
- Revision rounds: Changes after a file has been timed cost more than changes to the script before it is timed. Agreeing terminology up front is the cheapest round of the lot.
Can you get video subtitles translated for free?
Yes, up to a point. YouTube will auto-caption a track and machine-translate it, and several browser tools will do the same, at no cost and in minutes. For a rough internal cut, or to find out roughly what a file in a foreign language says, that is a reasonable use of them and it will save time and money.
What free tools do not do is hold a reading speed, break a line at the grammatical joint, identify a speaker, carry the sounds that matter, move a subtitle off a lower third or keep British spelling. Transcription is the stage you can automate. The others are the parts a UK viewer notices and a platform QC rejects, and none of them is a translation problem — they are the craft a caption file is judged on.
What does audio and video translation cost per minute?
Subtitling is quoted from £5 per video minute and transcription from £5 per audio minute, both measured against finished running time rather than word count. Recorded voice carries no published rate: a voiceover read over the original audio, or a dub that replaces the voices and matches the delivery to the picture, is costed from the running time, from how many voices the script needs and from whether a studio session and a mix are part of it. Where the same content is handled as text — a script, a transcript or an existing subtitle file — it is £0.10 per word.
A quote fixes which of those you are buying before work starts, so a caption job does not quietly become a voiceover one. Transcription and translation are itemised as separate lines, alongside spotting, revision by a second linguist and the export formats. Call us with the running time and the target languages, or send the file itself, and the figure comes back within 1 hour.
Which video translation services can you order?
The subtitling services, voiceover options and transcripts below can be ordered on their own or as a set off one master.
- Translated subtitle tracks: The dialogue in another language, timed to the picture and delivered as a caption file the destination will accept. The commonest order, and what most export work needs.
- SDH and closed captions: Same-language captions carrying speaker identification and the sounds that matter to the story, written for deaf and hard-of-hearing viewers rather than for language access.
- Burned-in captions: Captions rendered into the picture for platforms that strip caption tracks or autoplay muted, delivered as a finished master for each language.
- Voiceover: A translated script read by voice talent over the original audio, which stays audible underneath. Used for factual, corporate and training material where the script carries detail a subtitle cannot hold. Casting is a short list you listen to, not a hunt to find the perfect voice.
- Dubbing: The original voices replaced and the delivery matched to the picture, for material where the audience should be watching rather than reading. The most production-heavy option on this list.
- Timecoded transcription: The spoken content as text, timed to the frame. It is the substrate every other deliverable is built on and a useful asset on its own.
- On-screen text localisation: Titles, straps, slides and graphics rebuilt in the target language so the picture says what the audio says.
- Audio description scripts: A written account of what is happening on screen, for blind and partially sighted viewers, timed to fit the gaps in the dialogue.
- E-learning caption sets: Course video captioned and translated module by module, with terminology held consistent across the whole programme rather than file by file.
- Certified transcript translation: A translated transcript with a signed statement of accuracy naming the linguist who produced it, for use where a document rather than a video is what gets filed. £30 per page of up to 250 words, back in 24 to 48 hours.
What is subtitling, exactly?
Subtitling is the business of working out what can be read in the time available and then writing it. The transcript is timed against the original video first, with in and out points set at the shot boundaries wherever the edit allows, and the translation is written into those timings. That constraint is the whole craft: the text has to say what was said, fit two lines, and clear the screen before the next line is due.
An automatic pass is quick and costs almost nothing, and on clean single-speaker audio it produces a usable first draft. It does not hold a reading speed, does not know that a line has broken in the wrong place, and cannot tell a sound that matters from background noise. Our Subtitle Translation service puts human translators on that draft, or on the raw audio, and produces a file that reads the way a British viewer expects it to.
How do you get English subtitles onto a video?
The eight steps below are the workflow a video translation project follows through the studio.
- Send the master and anything written. Provide us with a locked cut, plus any script, transcript or existing caption file. Sending the picture before it is locked means the timings have to be done twice.
- Transcribe and timecode. The audio becomes text with in and out points, including who is speaking. This is the stage an automatic tool gets most nearly right, and it still has to be checked against the audio.
- Spot the file. Decide where each subtitle appears and disappears, against the shot boundaries and the reading speed rather than against the waveform.
- Translate into the timings. The linguist writes to the space available, condensing where the target language runs longer, and holds terminology consistent across the whole programme.
- Revise. A second qualified linguist checks the file against the source and against the picture — both are professional linguists working into their first language. ISO 17100:2015 requires that step, and it is where most errors are caught.
- Check on the picture. Play the file back in full. This quality assurance pass moves positions off graphics, fixes clashes with on-screen text, and cuts back anything that cannot be read at speed. It is the difference between a file checked professionally and one that has merely been exported.
- Proof the language itself. Spelling stays British, dates read DD/MM/YYYY, measurements stay metric and money stays in pounds.
- Export and deliver. SRT and WebVTT for web and social, EBU-STL or EBU-TT for broadcast delivery, a burned-in master where the platform needs one, and the timed transcript alongside.
Human or machine translation: what does a video actually need?
Machine translation is good enough now that the question deserves a straight answer. Anyone who has used a phone translator abroad knows what it does well: a short, plain sentence in a common language pair, rendered instantly and for nothing. The caption tracks that YouTube, Vimeo and most conferencing platforms generate are the same technology pointed at speech, and on clean audio, one speaker and unremarkable vocabulary they get surprisingly close. The output is usually fluent, which is exactly what makes the lines it gets wrong hard to spot. They fail predictably on names, on specialist terms, on overlapping speech, on accents the model has seen little of, on idiom, and on anything where a sound rather than a word carries the meaning. They also hold no reading speed and break a line wherever it happens to fall, because neither of those is a translation problem and neither is something the model was asked to do.
What a human translator adds is judgement about what the line is for. A qualified linguist working into their own first language condenses fast dialogue so it can be read at all, holds a term steady across a whole programme, decides which sound matters to the story, and sees when a rendering that is literally correct would land wrongly on a British audience. Under ISO 17100:2015 a second linguist then revises that work against the source, which is where most of the remaining errors are caught. The practical rule is what the file is for. An internal review copy or a rough search index is a reasonable use of an automatic track, and using one there saves real money. A published video, a course a student is assessed on, a safety briefing or anything a regulator or a tribunal might read back is not: those need professional subtitlers on the file, because nobody can tell you afterwards which lines the machine got wrong.
Do you transcribe video as well as translate it?
Yes, and the transcript is a deliverable in its own right rather than a stage you pay for twice. A timecoded transcript is what every other deliverable on this page is built from, and it is what a client needs when the video is evidence, a record or a source document: research interviews, board and committee recordings, training sessions, podcasts and witness footage. Because a transcript is text, it is searchable, quotable and auditable in a way video files on their own never are — it gives a search engine something to read, an internal system something to find and a compliance record something to point at, which is the same reason a caption track earns its cost twice over.
There are two ways to have it done and the difference matters. Strict verbatim transcription keeps every false start, repetition and filled pause, and is asked for wherever the manner of speech is itself part of what is being examined — research interviews, disciplinary hearings, anything that may be quoted back word for word. An intelligent, clean-read transcript takes those out and is what most training, marketing and editorial work wants, because it is readable. Say which one you need when you send the file: converting one into the other afterwards means listening to the recording again. Our transcription service is quoted per minute of running time, and the transcript can be translated at the same time or later.
Which subtitle formats should you ask for?
A UK job is exported in a wide range of caption formats. The usual deliverables are listed below.
- SubRip (.srt): The default caption file for web, social and most players. Plain text with timings, no styling, and accepted almost everywhere.
- WebVTT (.vtt): The HTML5 caption format. It carries positioning and basic styling, which matters when subtitles have to be kept clear of on-screen graphics.
- EBU-STL (.stl): The long-standing European broadcast subtitle format, still asked for by UK facilities and archives.
- EBU-TT and IMSC: The XML-based subtitle formats used for current broadcast and streaming delivery, where positioning, colour and timing all have to travel with the file.
- Burned-in master: The captions rendered into the picture. One master per language and no way to switch them off, so they are signed off before the render.
- Timed transcript: The spoken content as a document with timecodes, for archive, search, compliance records and as the basis of any later re-version.
How should you brief a video translation supplier?
Seven things settle a video brief. Translation companies scope and price video work differently from one another, so answering these before you ask for a price is what makes the quotes you get back comparable.
- Name the deliverable. Translated subtitles, SDH, voiceover, dubbing or a burned-in master. Those are different jobs and there is no single translation solution that covers all five, so asking for video translation on its own leaves the supplier guessing which one you meant.
- Say where it will play. Broadcast, a streaming platform, a learning system, a website or social. The destination decides the file format and often the caption rules as well.
- Send a locked cut. Timings are set against the picture, so a re-edit after spotting means redoing the work rather than amending it. A locked cut also means the date you were quoted is the date you get.
- Say what text already exists. A script, a transcript or an existing caption file removes a production stage and makes the quote simpler. Saying one exists when it does not is the commonest cause of a slipped date.
- Ask what rules the file will be built to. Reading speed, line count, positioning and speaker identification. A supplier who cannot answer that has not thought about the viewer.
- Ask who revises it. ISO 17100:2015 requires a second qualified linguist to revise the first one’s work against the source. On a marketplace you are buying one freelancer’s time; under ISO 17100 you are buying two linguists on the same file. “We proofread everything” is not the same answer.
- Get the quote in writing first. Running time, languages, deliverables and date, itemised before anything starts. Please click the quote button to upload the master, or contact us with the running time and the target languages if the file is too large to send. Ours comes back within 1 hour.
What is included in every video translation?
Whatever the deliverable, three things come with it: the files your destination will actually take, two linguists on every one of them, and a figure agreed before anything starts.
Files your platform takes
Subtitles arrive as SubRip or WebVTT for web and social, as EBU-STL or EBU-TT where a UK broadcaster or streaming service is the destination, and burned into the picture where captions would otherwise be stripped. A timecoded transcript comes with every job, for archive, search and compliance records. Spelling, dates and measurements stay British, which is the first thing an automatic track gets wrong.
Two linguists, 150+ languages
Work goes to a linguist writing into their first language, from a network of more than 2,000, each with at least five years’ experience. A second linguist then revises the file against the source and against the picture, which is what ISO 17100:2015 requires and where most caption errors are caught. Espresso works in over 150 languages, and one master can carry as many of them as the release needs.
Price and date up front
Subtitling is quoted from £5 per video minute and transcription from £5 per audio minute, while a script or transcript handled as text is £0.10 per word. A certified page of up to 250 words is £30, back in 24 to 48 hours. Translation of an existing transcript or caption file runs to the standard 2 to 3 working days, with 24-hour express where a transmission date will not move, and the quote itself is free within 1 hour.
What does each sector need from a video translation?
Six kinds of buyer account for most British video work, and each arrives with a different constraint rather than a different language. The deliverable list follows the constraint.
E-learning and training
Course video is translated as a programme rather than as a folder of separate files: one glossary and one set of subtitle conventions from the first module to the last. Captions are delivered as files a learning platform can load, or burned in where it cannot. Where the narration carries more detail than a viewer can take in while watching, a translated voiceover does the work the subtitle track cannot.
Broadcast and streaming
Delivery to a UK channel or a streaming service means EBU-STL or EBU-TT rather than a web caption file, positions kept clear of straps and lower thirds, and reading speeds built to the BBC Subtitle Guidelines. Ofcom’s Code on Television Access Services sets subtitling, signing and audio description requirements for licensed broadcasters. Masters here are commonly 25fps, so every frame rate the film will exist at has to be named before spotting starts.
Marketing and advertising
Most social video is watched with the sound off, so a burned-in caption track is frequently the only version of a film the audience receives. Straps, titles and end cards are rebuilt in the target language rather than subtitled, which brings a production stage into a translation job. One master then carries a subtitle track per market instead of a reshoot.
Public sector and access
Public sector websites and apps work to WCAG 2.1 AA under the 2018 accessibility regulations, which asks for captions on pre-recorded video at Level A and audio description at Level AA. The Equality Act 2010 duty to make reasonable adjustments reaches induction, safety and information video that no procurement clause mentions. SDH tracks and audio description scripts are ordered alongside the translated ones, not instead of them.
Medical and pharmaceutical
Clinical and device material has fixed wording, and a subtitle that paraphrases it to hold the reading speed changes what the film says. Those files go to a linguist who already works in the field, with terminology agreed before spotting rather than argued after it. Patient-facing video usually needs a same-language SDH track as well as a translated one.
Legal and regulatory
Where the recording is the evidence, the timecoded transcript is normally what gets filed and the video is exhibited alongside it. Strict verbatim transcription keeps every false start and filled pause, which is what a hearing or an investigation asks for, and a clean read is a different document. A certified transcript translation carries a signed statement of accuracy naming the linguist, at £30 per page of up to 250 words.
What are the four steps from file to delivery?
A video job starts with the audio rather than the text, which is the one thing that makes it different from a document job. The four stages below are what a quote is built against.
Send the master
Provide a locked cut plus any script, transcript or existing caption file, over secure transfer rather than email. Say which languages you need, where the film will play and every frame rate it will exist at. A link is enough for us to quote from; the file itself is what the timing is done against.
Transcribe and quote
Where no transcript exists the audio is transcribed and timecoded first, including who is speaking, which is a stage a document job never has. Running time, languages and deliverables then set the figure, itemised line by line and back within 1 hour. Terminology is agreed here, before anything is timed, because a change after spotting costs more than the same change before it.
Spot, translate, revise
In and out points are set against the shot boundaries and the reading speed, and the translation is written into those timings rather than onto the page. A second qualified linguist then revises the file against the source and against the picture, as ISO 17100:2015 requires. The file is played back in full so positions move off graphics and anything unreadable at speed is cut back.
Deliver every version
Exports go out in whatever the destination takes — SubRip, WebVTT, EBU-STL, EBU-TT, a burned-in master per language — with the timed transcript alongside. Standard delivery is 2 to 3 working days for translation of an existing transcript or caption file, with 24-hour express where a date will not move. Anything burned into the picture is signed off before the render, because it cannot be changed afterwards.
Which related service might you need instead?
Video translation sits next to several jobs that are ordered on their own terms. If what you are handing over is described better below, start there.
Subtitle translation
When the caption file is the whole deliverable: a translated subtitle track timed to a picture that is already finished. Quoted from £5 per video minute. Choose it where the audio stays exactly as it is and only the text changes. Subtitle Translation.
Audio translation
When the source is sound rather than picture — a recording, a podcast, an interview, a conference floor. Transcription is quoted from £5 per audio minute, and the transcript can be translated at the same time or later. Audio Translation.
Audiovisual translation
When one release needs several deliverables at once: subtitles, a voiceover version, on-screen text rebuilt and an audio description script, all off a single master. Choose it where the film is going to several markets and several formats together. Audiovisual Translation.
Video transcription
When you need the words as a document rather than as captions — a searchable, quotable record of what was said, timed to the frame. It is also what every other deliverable is built from, so ordering it first can remove a stage from a later job. Video Transcription.
E-learning translation
When the video sits inside a course rather than standing on its own, and the slides, quizzes and handbook have to say what the narration says. Terminology is held across the whole programme rather than file by file. eLearning Translation.
Marketing translation
When the script has to land rather than merely be accurate — a campaign film, a launch video, a social cut-down. The copy is written for a British audience and the on-screen text is rebuilt rather than subtitled. Marketing Translation.
What Else Do UK Buyers Ask About Video Translation?
Answers to the questions UK buyers ask once the brief is agreed — SDH and access obligations, file formats and frame rates, working from a video with no script, and what to send for a quote.
Do you translate e-learning and training videos?
Yes, and course video is the commonest kind of video work we are asked for after marketing film. A training programme is translated as a programme rather than as a set of separate files: one glossary, one set of subtitle conventions and the same terminology from the first module to the last, so a learner moving between modules is not reading two different vocabularies. Captions come as subtitle files your learning platform can load, or burned into the picture where it cannot, and a translated voiceover is available where the narration carries more detail than a viewer can take in while watching. Subtitling is quoted from £5 per video minute and the figure comes back within 1 hour.
Do captions on their own meet WCAG 2.1 AA, or is audio description needed too?
Captions are not the whole of it at AA. The W3C asks for captions on pre-recorded video at Level A and adds audio description for pre-recorded video at Level AA, so a public sector body working to the 2018 accessibility regulations is being asked for both. We write the caption or SDH track and the audio description script, timed to fit the gaps in the dialogue; recording that description is a separate production step and is quoted as one. Settle which of the two a film needs before the edit is locked, because a description script is written against finished picture and cannot be started twice.
Our master is 25fps — what happens if the platform wants another frame rate?
The subtitle file is timed against one frame rate and has to be conformed when the picture moves to another. UK masters are commonly 25fps and a good deal of web and streaming delivery is not, so a file that sat perfectly on the edit drifts progressively once it is laid onto the converted version — invisible in the first minute and glaring by the last. Tell us every frame rate the film will exist at before spotting starts, and each version is exported and played back on the picture it will actually run on. Retiming after delivery is a production stage rather than a correction, so it is scheduled and quoted as one.
Can you work from a link, or do you need the master file?
A link is enough to quote from and not enough to work from. Running time, audio quality and how much on-screen text there is can all be judged from a streamed copy, so a link and the target languages are usually all it takes to put a figure in front of you within 1 hour. Timing and export need the file itself, in the cut and at the frame rate that will be published, because a compressed web copy is a different length in frames from the master it came from. Files move over secure transfer rather than email, under signed non-disclosure agreements where the material is unreleased.
Can you certify a translated transcript for a UK court or tribunal?
What you receive is a translated transcript with a signed statement of accuracy naming the linguist who produced it, at £30 per certified page of up to 250 words, back in 24 to 48 hours. There is no sworn translator status in Britain: no state register of translators exists here, unlike France, Spain or Italy, so certification in the UK means a named linguist’s signed statement rather than a court-appointed seal. Whether that meets the requirement in front of you is the court’s or the tribunal’s decision, and the wording they want is worth asking them for before the file is ordered. Where the video itself is the evidence, the timecoded transcript is normally what gets filed and the recording is exhibited alongside it.
