Sign language interpretation in video is the rendering of spoken dialogue and relevant sound information into a natural sign language, such as American Sign Language or British Sign Language, by a qualified interpreter who appears on screen alongside or within the program. In video accessibility, it serves Deaf viewers whose first language is a sign language, for whom written captions in a second language may be less accessible.
Sign language interpretation is not a visual form of captions. Sign languages are complete languages with their own grammar, vocabulary and use of space, facial expression and movement, so interpreting involves translation between languages, not transcription.
Why captions do not replace sign language
Many Deaf people learn a sign language as their primary language and a written language as a second language. For them, reading fast captions can demand considerably more effort than watching a fluent signer, especially with dense or technical speech.
The following table compares sign language interpretation with neighboring accessibility services.
| Service | Audience | Modality | What it conveys |
|---|---|---|---|
| Sign language interpretation | Deaf sign language users | Visual, signed | Speech, tone and key sounds in sign language |
| Closed captions | Deaf and hard-of-hearing viewers | Visual, written | Speech and sound cues as text |
| SDH | Viewers who cannot hear audio | Visual, written | Subtitles plus speaker and sound information |
| Audio description | Blind and low-vision viewers | Auditory | Visual information as narration |
| Video transcript | Many audiences, including deafblind users | Text document | Full text of speech and descriptions |
Under WCAG 2.x, sign language for prerecorded audio is success criterion 1.2.6 at Level AAA, while captions are required at Level A. Broadcasters in several countries face separate regulatory quotas for signed programming.
Production formats for signed video
Signed content is delivered in several formats, which differ in how the interpreter is combined with the program and whether viewers can switch it off.
| Format | How it works | Trade-off |
|---|---|---|
| Open signing | Interpreter burned into the picture, usually in a corner | Always visible, cannot be turned off |
| Closed signing | Separate video stream composited by the player | User choice, needs player support |
| Signed edition | Separate full version of the program | Simple playback, double the files |
| Live in-vision interpreter | Interpreter beside a speaker in the camera shot | Natural for briefings, depends on framing |
| Native signed content | Presenters sign directly, no interpreter | Best fluency, requires signing talent |
Closed signing is technically the most flexible option. Streaming players and hybrid broadcast standards can carry the signer as a second video track, so viewers choose its position and size, but support across devices is still uneven.
Technical requirements for a legible signer
Sign language relies on handshapes, movement, facial grammar and body position, so image quality matters more than for an ordinary inset. A signer who is too small or blurred becomes unintelligible even when the interpretation is excellent.
| Parameter | Recommended practice | Reason |
|---|---|---|
| Size | Large enough to show face and full signing space | Facial expressions carry grammar |
| Framing | From above the head to below the waist, with elbow room | Signs extend beyond the torso |
| Frame rate | 50 or 60 fps where possible | Fast finger movements stay crisp |
| Background and clothing | Plain, contrasting, no patterns | Hands remain clearly separated from the background |
| Lighting | Even, soft light without hand shadows | Handshapes stay readable |
| Encoding | Sufficient bitrate for the signer region | Compression smears rapid motion |
Broadcast guidelines in several countries set minimum signer sizes, often expressed as a share of picture height or area. The signer should never be covered by graphics, logos or captions, and the position should stay consistent throughout a program.
Interpreting practice and quality
Quality depends first on qualified interpreters who are fluent in the specific sign language. There is no universal sign language: ASL, BSL, Auslan, LSF and DGS are distinct languages, and ASL and BSL are not mutually intelligible even though both regions use English.
- Team interpreting is standard for live events, with interpreters rotating about every 15 to 30 minutes to limit fatigue errors.
- Deaf interpreters, working from a hearing interpreter or a script, often deliver the most natural target language for broadcast.
- Preparation materials such as scripts, names and terminology improve accuracy, especially for specialist topics.
- Lag of a few seconds behind speech is normal, because interpreters need context before rendering meaning.
- Sound information such as music, alarms or laughter should be conveyed when it matters to the content.
Automated signing avatars are being researched and piloted, but Deaf community organizations have raised concerns about their naturalness, facial grammar and accuracy. For important information, human interpretation remains the accepted standard.
Common mistakes in signed video
Most failures in signed video are production decisions, not interpreter errors. Cropping the interpreter out of live briefings, using a tiny inset or placing captions over the signer all remove access that the organization intended to provide.
Another frequent mistake is treating interpretation as a substitute for captions. The two services reach different groups of people, so accessible video usually provides both captions and signing where sign language access is a stated goal.
The table below lists recurring problems, their effect on Deaf viewers and a practical fix.
| Problem | Effect on viewers | Fix |
|---|---|---|
| Interpreter cropped in live shots | Access lost during key moments | Brief camera operators to keep the interpreter in frame |
| Inset too small | Handshapes and faces unreadable | Enlarge the panel or use a signed edition |
| Captions placed over the signer | Signs hidden behind text | Reserve separate screen areas for each service |
| Wrong sign language | Content unintelligible to local viewers | Book interpreters for the audience’s national sign language |
| Low bitrate streaming | Blurred finger movement | Raise bitrate or prioritize the signer region |
Involving Deaf reviewers before release catches most of these issues early, because they notice legibility and language problems that hearing staff easily overlook.
In everyday language
For many Deaf people, sign language is their mother tongue, and written English or another written language is learned later, like a foreign language. Subtitles help, but they are not the same as hearing someone speak your own language.
Sign language interpretation in a video puts a person signing on screen, translating what is said into sign language as it happens. It works best when the signer is large, clearly lit and never covered by other graphics.
Seen in the real world
A national health agency publishes a weekly 20-minute video update about seasonal vaccinations. It already provides closed captions, but Deaf community organizations report that many sign language users still miss key details.
The agency books two qualified interpreters, records them against a plain dark background at 50 frames per second and offers a separate signed edition in which the interpreter occupies a large panel at the right of the frame. Scripts are shared a day in advance so that drug names and dates are prepared. Feedback from Deaf viewers improves, and the agency adopts a rule that live press briefings must always keep the interpreter in shot.