Captions / subtitles

Captions for video is the creation of accurate, time-synchronized text of what sounds in the video: lines, speaker names if needed, and important non-speech sounds (e.g. '[laughter]', '[phone ringing]') for the full picture. We do it for deaf and hard-of-hearing users (a WCAG accessibility criterion), and also for those watching without sound (metro, office) — there are many such now. Honestly about quality, this is key: good captions are HUMAN work: transcription, correct punctuation, accurate synchronization with speech, a readable pace. Automatic captions (YouTube auto-recognition, etc.) are only a DRAFT: they err on terms, names, accents, with noise and multiple speakers, and publishing them as 'accessibility' without checking is dishonest. We can take auto as a base, but finish it manually. Honestly about volume: the price depends on the DURATION of the video (minutes) and complexity (number of speakers, terminology). Honestly about languages: captions in each language are separate work (translation+synchronization); one set of captions ≠ all languages. Honestly about the process: new videos require new captions — this is maintenance. Honestly about the boundary: captions convey speech and sounds as text, but that is not the same as audio description of the visuals for blind users (814) — different tasks for different needs. Honestly about the effect: accessibility + retention of those without sound; we do not guarantee direct sales growth. Honestly about access: source videos are needed. An important boundary: this is captions for video; transcripts of audio are 812, audio description is 814. Picture this: instead of 'without sound the video is useless, and deaf users are excluded' — text synchronously accompanies the video. The base price starts from 15,000 ₽ (depends on duration; each language separate).

• Deaf and hard-of-hearing users do not understand the video — no captions.
• Without sound (metro, office) the video is useless — audience is lost.
• Auto-captions with errors on terms, names and with noise.
• Captions are out of sync and unreadable by pace.

• Transcription of speech with correct punctuation
• Accurate synchronization of captions with the video
• Marking important non-speech sounds ([laughter], [phone ringing])
• Speaker identification if needed
• Finishing auto-captions manually (if there is a base)
• Indicating boundaries (each language separate; auto — only a draft)
• Checking readability and pace
• Handover in the needed format and review with you

• The video is clear to deaf and hard-of-hearing users
• Content is accessible to those watching without sound
• Captions are accurate and in sync (without auto-recognition errors)
• The captions criterion is closed (each language/new video — separate)

• We get the video, clarify languages and complexity; estimate by duration
• We transcribe, synchronize, mark sounds, check the pace
• We hand over in the needed format, review boundaries (languages/new videos) with you

• Auto-captions can just be turned on — why pay?

  Auto-captions are a draft: they err on terms, names, accents, with noise and multiple speakers. Publishing them as 'accessibility' is dishonest — a deaf user gets distorted meaning. We can take auto as a base, but finish it manually: transcription, punctuation, accurate synchronization. That is real accessibility.

• Will you make captions in all languages at once?

  Each language is separate work: it is translation plus its own synchronization. One set of captions does not cover all languages. Tell us the needed languages — we will calculate honestly for each; it is usually more reasonable to start with the main one and the language of the key audience.

• Is this the same as audio description?

  No, different tasks. Captions convey speech and important sounds as TEXT — for those who cannot hear. Audio description (814) — conversely, describes the visual events by VOICE for blind users. These are needs of different groups; we will honestly say what you need and whether you need both.