Sound Descriptions: A Checklist to Run Through Before You Decide

The first question on my list is the simplest one: if this sound were removed, would a viewer who cannot hear it miss something that matters? A door closing behind a speaker usually does not matter. A phone ringing in the middle of a recorded interview, which makes the guest stop and apologise, does. Without the description, the pause looks like hesitation.

When I teach this part of Sound Descriptions, I ask learners to watch a clip once with the sound off and note every moment where they feel lost. Those moments are the candidates. Everything else is background, and background belongs in the audio track rather than in your caption lines.

A useful habit is to separate three groups as you go: sounds that carry meaning, sounds that set a mood, and sounds that are only there. Only the first group is almost always worth describing. The second needs a judgment call, and the third can be left alone.

• Would the viewer be confused without it?
• Does a speaker react to it?
• Does it explain a pause or a cut?
• Is it a one-off or constant background?

Decisions about sound descriptions depend on the audience. A recorded conference talk watched by colleagues needs fewer notes than a training video where a warning tone signals a step. Before you decide, write down who will watch, on what kind of screen, and whether they are likely to have the sound off by choice or by necessity.

Viewers who cannot hear the audio rely on you for everything non-spoken. Viewers who simply muted the video in a waiting room want the gist and will skim. Both groups are served by short, plain descriptions, but the first group will notice every gap, so lean toward including more when access is the reason for the captions.

Also consider where the video lives. A file embedded in a web page may be watched at small size, where long bracketed lines are hard to read. An event recording shown on a large screen can carry slightly fuller notes without crowding the picture.

• Name the audience in one sentence before you start
• Decide whether access or convenience is the main reason
• Check the screen size the video will most often be seen on

Once a sound passes the first test, the next question is whether you can name it briefly. Good descriptions are short and concrete: applause, laughter, a chair scraping, a phone buzzing. If you need a full sentence to explain it, either the sound is more complicated than it seems or you are describing too much.

Avoid adjectives that tell the viewer how to feel. Writing that applause is warm or enthusiastic is your opinion, not the sound. Say what happens and let the picture do the rest. Where tone matters, such as a speaker sighing or speaking quietly, a single plain word beside the speaker label is usually enough.

Keep your wording consistent across the whole video. If you call it laughter in the first minute, do not switch to chuckling in the tenth unless the sound really differs. Consistency helps viewers learn your conventions quickly and trust them.

• Prefer nouns and simple verbs
• Avoid judgments about mood or quality
• Use the same word for the same sound each time

Placement is the part beginners often overlook. A description has to appear while the sound happens, stay on screen long enough to read, and not collide with speech. If a door slam lands during a spoken line, you may have to fold the note into the line before or after, or shorten it, rather than stacking two blocks of text.

Check reading time as you would for dialogue. A short note such as applause still needs a moment on screen, usually a second or two beyond the sound itself, unless the next caption starts immediately. If it flashes past, a viewer will see that something was there but not what.

Style matters too. Many teams put descriptions in square brackets, sometimes in lowercase or italics, so they stand apart from speech. Whatever convention you choose, apply it everywhere. Ask the person supplying the recording whether they already have a preferred style, because it is easier to follow one than to redo it later.

• Show the note while the sound is happening
• Give short notes enough time to be read
• Never let a description hide a speaker's words

Sometimes a speaker names the sound for you. If a presenter says that was the alarm, a bracketed alarm note right before it repeats what the viewer is about to read. Likewise, when an audience laughs at a joke that is plain from the words, a note may be unnecessary, though a brief one can still help viewers who want to know the room reacted.

The test here is redundancy. Read your captions with the sound off and ask whether the description adds anything the text and picture have not already given. If it does not, remove it. Fewer, better notes keep the reader's attention on the speech, which is the main content.

There is one caution. Do not assume the picture always explains the sound. A camera may be pointed at a slide while someone coughs repeatedly off screen, or music may start under a talk. When the visuals give no clue, the description is doing real work and should stay.

• Remove notes that repeat what a speaker just said
• Keep notes where the camera shows nothing relevant
• Re-read the captions with the sound muted

Music and ongoing noise need their own decision. A short musical sting at the start of an event video can be covered by one note. A long passage of background music under a talk should usually be mentioned once, when it begins and when it changes or ends, rather than repeated on every line.

If a song has lyrics that matter, that is a separate question: you may need to caption the words themselves, which affects your timing and your workload. Decide this before you start rather than halfway through, because it changes how you plan the whole file.

Finally, set a rule for yourself about volume and effort. Describing every rustle across a ninety-minute recording is not realistic or helpful. Pick a threshold, apply it evenly, and review the finished file once with fresh eyes, asking whether each note still earns its place.

• Mention background music once at start and end
• Decide early whether lyrics need captioning
• Set a threshold and apply it evenly