Skip to content

Audio Description Changes the Commercial’s Timing. Recut It—or Make Another Version?

Advertising

Audio Description Changes the Commercial’s Timing. Recut It—or Make Another Version?

The problem arrives as a timing note. The commercial is locked. The visual payoff needs information that isn’t in the soundtrack. Someone suggests speaking faster over the dialogue. That is a rate adjustment, not an editorial decision.

Audio description changes what the listener knows and when they know it. When the existing cut has no room, you have to compare three different changes: integrating the information into the film, recutting to create room, and making a separately timed described version. Each changes something different. The choice depends on the actual delivery context, not on a general preference for one method.

Account for information before counting available seconds

Start with the indispensable visual facts. Map them against dialogue, effects, music, and silence. Ask what the existing sound already communicates. Ask what a listener still needs. Then mark when each fact becomes necessary.

Describing a reveal too early changes the film. Describing it after its consequence can leave the consequence unexplained. This is not the same as narrating every visible action. Audio description for a commercial is an access task: it carries the information without which the story does not work.

Take an invented 30-second spot. Call it “The Second Ticket.” The timecodes are placeholders. No soundtrack, timing trial, placement approval, or listener review exists for it.

On a train platform, Mara waits with a small envelope. A stranger sits on a bench, looking at the departure board. Mara opens the envelope, takes out a ticket, and offers it to the stranger. The stranger hesitates. Dialogue: “I can’t.” Mara: “Take it.” The stranger takes the ticket. Then Mara opens her hand to reveal a second ticket. The final image shows both tickets. They read the same destination: Ardmore. The apparent sacrifice becomes an invitation.

Here is what a listener needs, and when:

Time Picture Sound Necessary description
0:00–0:04 Platform. Mara with envelope. Stranger on bench. Music, station ambience. “A woman with an envelope waits on a train platform. A stranger sits on a bench nearby.”
0:04–0:08 Stranger looks at departure board, then down, worried. Music. “The stranger checks the departure board, then looks down, worried.”
0:08–0:12 Mara opens envelope, takes out a ticket. Music. “She takes a ticket from the envelope.”
0:12–0:16 Mara offers ticket to stranger. Music. “She holds it out to the stranger.”
0:16–0:20 Stranger hesitates. “I can’t.” / “Take it.” None. Dialogue carries hesitation and insistence.
0:20–0:24 Stranger takes ticket. Mara reveals second ticket. Music. “He takes it. She shows a second ticket.”
0:24–0:30 Close-up: both tickets read Ardmore. Music. “The destination on her ticket matches his. Both read Ardmore.”

The second line names an emotion rather than an action, and it could be written as a slack shoulder or a hand on the back of the neck instead. Either way the cue belongs there. The shot exists to show that the stranger needs the ticket; without it, the gift in the next shot arrives without a reason.

The total description is about 60 words. At a comfortable 160 words per minute, that is roughly 22 seconds of speech. The cut has 26 seconds of non-dialogue space. In total, it fits. The problem is where.

Four of the lines carry the setup, and all four have to land before the dialogue begins at 0:16. They total 41 words: about 15.4 seconds of speech in a 16-second window. Under a second of margin, and none of it in the shot that needs it most.

The 0:00–0:04 shot needs 17 words to establish the woman, the envelope, the platform, and the stranger. That is about 6.4 seconds. The shot is four seconds. The opening beat is already more than two seconds over. Let the description spill into the next shot, which is what the arithmetic forces, and the whole stretch before the dialogue runs behind the picture. Spoken end to end, the platform line reaches 0:06.4. The stranger’s board check—ten words, 3.75 seconds—then runs from 0:06.4 to 0:10.1, so the worry is described while Mara is already opening the envelope. The envelope line runs to 0:12.8, three-quarters of a second past its own shot. The offer ends at 0:15.4. By the time the dialogue begins at 0:16, the description has just caught up. Six-tenths of a second remain.

That is the shape of the problem. The four lines fit only if they run back to back, with no pause to let a shot land before it is named. And the deficit sits at the front, where four seconds have to carry the whole setup.

This is why “speak faster over the dialogue” is misdiagnosed. The dialogue is at 0:16–0:20. The missing space is at 0:00–0:04. Speaking faster over the dialogue does not help the opening. It only makes the lines compete with the description. The first step is to see the problem as a distribution problem, not a single rate problem.

Try integration without making the scene explain itself unnaturally

Integration means changing the film so the necessary information is carried by dialogue, sound, or staging. In the invented scene, you could rewrite the exchange at 0:16. The stranger says, “Is this your only ticket?” Mara says, “I have another.” That moves the second ticket into dialogue. It solves the 0:20–0:24 gap.

But it changes the reveal. The audience now knows about the second ticket before the final image. The final image becomes confirmation, not surprise. The performance changes: Mara’s quiet action becomes a line. The point of view shifts: we are told rather than shown. The creative cost is the visual reveal.

What remains unavailable: even with the line, the matching destinations are not conveyed. You would need another line—“Ardmore, same as yours”—which is expository and unnatural. So integration can solve access, but it may solve it by making a different scene.

A motivated sound is another integration route. At 0:20, the platform announcement could say, “Now boarding, Ardmore.” That conveys the destination without dialogue. But it does not convey that the stranger’s ticket also says Ardmore. The listener might infer, but it is not certain. The announcement also changes the sound design. It might feel like a patch.

Staging changes are possible too. Mara could hold both tickets from the start, but that spoils the reveal. She could show the second ticket earlier, but that moves the payoff. Each integration route buys space by changing the film’s method.

The rule is not “never integrate.” It is “do not insert implausible dialogue merely to avoid another version.” If the line feels like a caption spoken aloud, it will weaken the commercial. Integration is a real option. Identify its cost and what remains unavailable after the change.

Compare a recut with a separately timed version

A recut changes the edit. It may hold a shot longer, shorten another, or redistribute dialogue—which is what this cut needs, because the room before 0:16 is spent and the room after it is not. In the invented scene, the deficit is at the front: the opening needs about 6.4 seconds and has four. A recut can hunt for those two and a half seconds further along, taking a second from the board check at 0:04–0:08 and another from the close-up at 0:24–0:30, and holding the platform shot to six seconds.

The seconds are not free. The 0:04–0:08 shot establishes the stranger’s worry, which is what makes the gift motivated rather than arbitrary. The close-up is the reveal the spot is built toward. Take time from either and the scene pays for its description: the need may not register before the gift, or the match may feel rushed. A recut can create room, but you have to re-examine the rhythm and the reveal. Does the scene still work? If the recut makes the description fit but weakens the story, you have traded one problem for another.

A separately timed described version is a different deliverable. It can be longer, or it can pause the video to allow description. W3C’s Understanding Success Criterion 1.2.7: Extended Audio Description (Prerecorded), checked in September 2026, explains extended description when soundtrack pauses are insufficient, including paused playback and separately described versions. That is WCAG 2.2 Level AAA guidance for web content. It is not an advertising-slot specification. It does not establish that a particular placement will accept a paused or longer version.

So the question is not whether W3C permits it. The question is whether the actual destination permits it. If the commercial is contracted for a 30-second linear slot, a 35-second version is not a version of that placement. If the destination is a web page or an app where a separately timed described version can be offered, then it may be possible. The decision depends on the delivery specification, not on the accessibility technique alone.

Keep the decisions distinct. A recut keeps the original duration and changes the edit. A separately timed version changes the duration or playback. Integration changes the film’s construction. They are not interchangeable.

Choose using the actual destination and a listening review

For each route, state what changes, which essential facts survive, and what delivery restriction remains.

  • Integration changes dialogue or staging. Essential facts can survive if the scene is rewritten. The restriction is the creative cost: the reveal may become verbal, and the visual surprise may be lost. It fits the original duration only if the rewrite works.
  • Recut changes shot lengths. Essential facts survive if the recut preserves the setup and the reveal. The restriction is rhythm and emotional logic. It fits the original duration only if the recut yields enough space without weakening the scene.
  • Separately timed version changes duration or playback. Essential facts survive with comfortable description. The restriction is the delivery context. It is only possible where the destination accepts a different duration or a paused presentation.

Get the actual delivery specification before comparing feasible editions. You need to know the exact duration, whether the placement accepts a different file, whether it can pause, whether there is a separate audio track, and who controls playback. Without that, you are comparing hypotheticals.

Then plan a listening review with an appropriate description practitioner and intended users. This is not optional. The review should check whether the listener understands the stranger’s need before the gift, whether the second ticket lands at the right moment, whether the match between destinations comes through, whether the description competes with the dialogue, and whether the pace is comfortable. The file ending on time is not the test. A compressed track that fits the slot but fails comprehension is not a successful route.

If the slot cannot accommodate the proposed route, return to the creative construction or establish an authorized alternative destination. Do not offer a longer version as though it fits the original placement. Do not label an unreadably compressed track successful just because its file ends on time.

The route that preserves the listener’s necessary understanding within a verified delivery context is the one to choose. In the invented scene, if the 30-second slot is fixed and the sixteen seconds before the dialogue are already committed, the recut or the integration may be the only on-air options. Each changes the film. The separately timed version may be the better access route if the destination allows it.

The wrong answer is to speak faster over the dialogue and call it solved. That treats a timing and information-design problem as a rate problem. The right answer is to decide what must be understood, when, and which version can carry it.

Frequently asked questions

Why isn't speaking faster over the dialogue a fix?

In the invented spot, the dialogue is at 0:16–0:20, but the missing space is at 0:00–0:04. The opening shot needs 17 words, about 6.4 seconds, and has four seconds. The four setup lines total 41 words, about 15.4 seconds of speech in a 16-second window, leaving under a second of margin and no pause to let a shot land before it is named. Speaking faster over the dialogue does not help the opening; it only makes the description compete with the lines. The first step is to see the problem as a distribution problem, not a single rate problem.

What does integration change in the invented 'Second Ticket' scene?

Integration changes dialogue, sound, or staging so the film itself carries the necessary information. Rewriting the exchange to 'Is this your only ticket?' and 'I have another' moves the second ticket into dialogue and solves the 0:20–0:24 gap. But the audience now knows about the second ticket before the final image, so the final image becomes confirmation rather than surprise; Mara's quiet action becomes a line, and the point of view shifts from shown to told. The matching destinations still are not conveyed without another expository line such as 'Ardmore, same as yours.' So integration can solve access by making a different scene. The rule is not never integrate; it is do not insert implausible dialogue merely to avoid another version.

What are the limits of a recut?

A recut changes the edit, holding some shots longer and shortening others or redistributing dialogue. In the invented scene the deficit is at the front, so a recut might take a second from the board check at 0:04–0:08 and another from the close-up at 0:24–0:30, and hold the platform shot longer. Those seconds are not free: the board check establishes the stranger's worry that makes the gift motivated rather than arbitrary, and the close-up is the reveal the spot is built toward. Take time from either and the scene pays for its description. A recut can create room, but you have to re-examine the rhythm and the reveal; if it makes the description fit but weakens the story, you have traded one problem for another.

When is a separately timed described version possible?

A separately timed version changes duration or playback, and it can be longer or pause the video to allow description. W3C's Understanding Success Criterion 1.2.7: Extended Audio Description (Prerecorded), checked in September 2026, explains extended description when soundtrack pauses are insufficient, including paused playback and separately described versions; that is WCAG 2.2 Level AAA guidance for web content, not an advertising-slot specification. It does not establish that a particular placement will accept a paused or longer version. The question is whether the actual destination permits it: a 35-second version is not a version of a fixed 30-second linear slot, while a web page or app might allow it. Decide from the delivery specification, not the accessibility technique alone.

What should a listening review check?

Plan a listening review with an appropriate description practitioner and intended users; it is not optional. The review should check whether the listener understands the stranger's need before the gift, whether the second ticket lands at the right moment, whether the match between destinations comes through, whether the description competes with the dialogue, and whether the pace is comfortable. The file ending on time is not the test. A compressed track that fits the slot but fails comprehension is not a successful route. If the slot cannot accommodate the proposed route, return to the creative construction or establish an authorized alternative destination; do not offer a longer version as though it fits the original placement.

More in Advertising Browse all articles