The Tools We Use to Set Caption Lines and Timing

Most of the timing work in this module happens in a subtitle editor that shows the audio as a waveform beneath the video. I ask learners to start there because the waveform makes speech visible. You can see where a sentence begins, where the speaker breathes, and where a long pause leaves room for a clean cut between two captions.

A plain text editor can hold the words, but it cannot tell you when they were said. With the waveform in front of you, setting a start time stops being a guess and becomes a matter of lining a marker up with the first syllable you can see and hear.

We work in short passes. First listen to a stretch of about thirty seconds, then mark the starts, then adjust the ends. Doing it in this order keeps the ear fresh and stops learners from polishing one caption while the rest of the clip drifts out of sync.

• Waveform view for locating the first and last syllable
• Looping playback for short, difficult phrases
• Frame-level nudging for fine adjustment

Alongside the editor, every learner keeps a one-page reading-speed guide. It sets out how many characters a viewer can comfortably read per second and how long a caption should stay up before it becomes a flicker. We use it as a reference, not a rule to obey blindly.

The guide earns its place when a speaker talks quickly. Learners often try to fit every word on screen and end up with captions that vanish before anyone can read them. The guide shows when to trim a filler word or split a thought across two captions instead.

Priya's habit, which I pass on, is to read each caption aloud at a normal pace while watching it appear. If you stumble reading it, a viewer will too, and that is the cue to change either the wording or the timing.

• Character limit per line, checked against the guide
• Minimum and maximum display time per caption
• A quick read-aloud test for every tricky caption

Line breaks are a method, not an afterthought. The material we rely on here is a short set of break rules: split at natural pauses, keep a name or a verb phrase together, and avoid leaving a single short word hanging alone on the second line.

Learners practise on a printed transcript first, drawing a slash wherever they would break the line. Only then do they move into the editor. Working on paper separates the question of where the thought divides from the question of how the software handles it.

We compare results as a group. Two people often break the same sentence differently, and both can be correct. The useful discussion is which version reads more easily at speed, and why, since that judgment is what the module is really teaching.

• Break at a pause or a clause, not mid-phrase
• Keep titles, names and numbers on one line
• Aim for two lines at most per caption

The sample recordings are chosen because they behave differently. One is a calm single-speaker talk with clean audio. Another is a fast interview with short answers. A third has pauses, false starts and a speaker who trails off mid-sentence.

Each clip exposes a different timing problem. The calm talk teaches the basic rhythm of start and end. The interview teaches how to handle quick exchanges without crowding the screen. The hesitant speaker teaches when to hold a caption through a pause and when to let it clear.

We deliberately keep the clips short, usually a few minutes, so learners can finish a full pass and then review it. A long recording early on tends to produce fatigue and sloppy timing, and a finished short clip teaches more than an abandoned long one.

• A steady talk for learning the basic rhythm
• A fast interview for tight turn-taking
• A hesitant speaker for pauses and false starts

Before a learner calls a clip finished, we play it back from the start with the captions on and nothing else open. This is the simplest tool we have and the one most often skipped. It shows drift that the editor view can hide.

We look for three things: captions that appear before the speech, captions that linger after the speaker has stopped, and places where a caption changes too close to a cut or a change of speaker. Each has a small fix, usually a few frames either way.

Learners keep a short note of the fixes they made on each clip. After a few clips the note shows a pattern, such as always starting captions slightly late, and that pattern is more useful to the learner than any general advice from me.

• Play the whole clip once with captions only
• Note early starts, late ends and crowded changes
• Record your own recurring habit and correct it

None of this requires special equipment. A laptop, a pair of closed headphones, the editor and the printed guide are enough. We have found that learners do better with a plain, repeatable setup than with a long list of extras they never learn to use well.

Headphones matter more than people expect. Laptop speakers blur the quiet start of a word, and that is exactly the moment you are trying to mark. Closed headphones make the first syllable clear and cut out the room.

If you work on your own recordings later, the same small kit travels with you. The method stays the same whether the clip is an interview, a talk or an event recording, and that portability is the main reason we teach it this way.