The actor leaves the kitchen, but the camera does not follow. A drop of water remains on the rim of a glass. A chair settles back into place. The refrigerator hum returns to the foreground of the soundtrack. No person is visible and the product is no longer facing the camera, yet the audience knows the moment is not over.
Another film shows an empty kitchen too: polished light, a spotless counter, and a slow lateral move. It may be useful as a transition, but it does not change how we understand the story.
Both are empty shots. The difference is not what the frame contains. It is whether the shot has a relationship with the images before and after it. A useful empty shot is not simply atmospheric coverage captured at the end of the day. It allows a person, product, or event to keep shaping the viewer's interpretation after leaving the frame.
An empty shot is not empty when an action is still present
Adobe's guide to B-roll describes supporting footage as a way to establish a scene, smooth a transition, or add meaning. That is a practical starting point. A brand film needs one more question: when the principal subject is absent, what job does this particular image complete for the story?
A 2025 study in Humanities and Social Sciences Communications analysed 2,464 films released between 1905 and 2019 and found meaningful differences in the distribution and use of empty shots across genres. The authors also note that the term itself is contested. Some definitions require a frame with no people, while others include images in which people occupy only a very small part of the composition. For commercial filmmakers, the terminology matters less than the function. Does the shot move time, space, emotion, or causality forward?
Think of a narrative empty shot as a sentence without an explicit subject. The person is absent, but the image still has tense: something is about to happen, has just happened, has changed, or has failed to happen.
Four empty shots answer four different questions
Before an entrance: who is about to enter this space?
An empty shot before an entrance creates expectation. A moving shadow appears beneath a door. A standby light changes state. A lift chime sounds in the distance. The audience has not seen the person, but it has already begun to wait for the arrival.
The point is not to show the whole location. It is to plant a cue that the next shot can fulfil. If the following image ignores that cue, the empty shot is merely a location marker. If the shadow, sound, or device state receives an answer, the space has begun to participate in the story.
After an exit: what did the action leave behind?
An empty shot after an exit shows consequence. The runner leaves frame while the treadmill slows down. A child walks away while a robot vacuum circles the crumbs just dropped. A meeting ends, leaving one question circled on the whiteboard.
The product does not need to remain at the centre, but the change it caused must be visible. For brands, this is a useful way to express the state after use. Instead of repeating a feature demonstration, the film lets the result remain for a beat.
No person, but a trace: who still exists off screen?
A half-finished glass of water, a creased bed sheet, a screen that has just gone dark, or an umbrella left by the door can keep a character present outside the frame. The more specific the trace, the easier it is for the viewer to infer a human action. The more generic the trace, the closer the image becomes to stock footage.
This is where atmospheric prop styling often goes wrong. An open book, permanent steam above a coffee cup, and an unexplained shaft of side light do not automatically create lived experience. The trace should belong to the person in the script. Why were they in a hurry? What did they just finish? What will they return to collect?
A changed space: how do we know time has passed?
When the same location appears twice, change matters more than camera movement. Morning light reaches the other side of a wall. A sealed package has been opened. The water level beside a plant has dropped. Loose components have become a finished product. The film does not need a caption saying “later”; viewers read time through the difference in state.
Magliano and Zacks' 2011 study of continuity editing and event segmentation found that a discontinuity in action had a greater effect on how viewers segmented a narrative than a change in time or space alone. The useful production lesson is modest but important: a film can jump in time or location, but the viewer still needs a bridge between the earlier action and its present consequence.
Use one deletion test to judge whether the shot is doing real work
Remove the empty shot from the rough cut, then ask three questions:
- Is the character's decision harder to understand?
- Has the consequence of the product disappeared?
- Has the transition in time, place, or emotion become abrupt?
If the answer is no to all three, the shot is probably functioning mainly as edit coverage. That is not inherently a problem. Interviews and documentary-style corporate films need cutaways to remove a stumble, compress a pause, or bridge two takes. But coverage should not be mistaken for narrative achievement. When a narrative empty shot is removed, the story loses a relationship, not merely a beautiful image.
That is also why restraint requires planning. Saying less does not mean making fewer decisions, and removing a person from the frame does not give the camera permission to linger without purpose.
Do not write “empty shot” in the storyboard; write the visible change
“Empty kitchen,” “empty office,” and “city atmosphere” do not give camera, art, and sound departments enough direction. A shootable description should state when the image appears in relation to an entrance or exit, which state changes, which sound supports that change, what ends the shot, and what the next image must fulfil.
For example, replace “three-second empty living room” with this: “After the father exits, hold the locked frame as the sofa cushion rises back into shape. The silent football broadcast continues on television. His keys remain on the table. Cut downstairs when the lift chime is heard outside the flat.” Camera knows the frame must be preserved. Art knows which props cannot be reset. Sound knows the lift has a narrative role. Edit knows that the cut point is not an arbitrary three seconds.
On set, capture three states of the same space: before the person arrives, during the action, and after the person leaves. Hold clean handles at both ends and record room tone plus the sounds of any object that changes state. ONCE's guide to shooting multiple TVC versions in one production treats empty scenes as useful version coverage. The next step is to label the dramatic function of each empty shot in the shot list, not merely mark it as a possible transition.
Sound can keep a person inside an empty frame
An empty shot can be quiet without being silent. Off-screen sound often provides the strongest evidence that a person still exists nearby: footsteps fading down a corridor, water stopping in a bathroom, a distant reply that cannot quite be understood, or a device chime after a task is complete.
Sound can confirm an action in sync, preserve an action after it has ended, or introduce the next scene before it appears. The delayed relationship is easily overlooked. The person has left, but cloth movement, a lock settling, or a motor slowing continues. The image gains an after-effect instead of being filled by music.
If every empty shot sits under the same music bed, the relationship between character and space becomes flat. Recording 30 to 60 seconds of stable room tone and isolating the sounds of objects that change state will often preserve the identity of a location better than a generic library effect.
The product can disappear while the brand value remains visible
An empty shot in a brand film does not need to hide a logo in the set. The brand can be felt through the kind of change the film chooses to observe.
For a product about efficiency, the space may move from disorder to readiness. For a service about companionship, an unoccupied place may remain available. For durable equipment, the environment can change while the order created by the product remains. Brand value is not an adjective written on an empty wall. It is the consequence the camera considers worth one more second.
Distribution still changes the decision. Google's official ABCD guidance recommends getting to the heart of a video story quickly, keeping the message focused, and making branding clear early. That does not mean an empty shot must be removed. It means the empty shot must itself help the film enter the story. A six-second version may keep only the strongest state change; a longer brand film can preserve both the anticipation before an entrance and the echo after an exit. The same atmospheric footage should not be scaled proportionally into every duration.
AI can generate an environment, but it cannot decide what that environment means
Generative video is useful for testing weather, season, light changes, and environmental layers that would be difficult to capture. Adobe's December 2025 Firefly tutorial explicitly demonstrates text-to-video for atmospheric B-roll and controls for shot size, angle, and motion.
But a prompt such as “cinematic premium empty room” is likely to produce generic footage. A narrative prompt needs state: who has just left, which object has changed, where the light came from in the preceding shot, and which sound the next shot must answer. Without that context, the image may be attractive but cannot prove that it belongs to this film.
When a scene depends on an exact product state, a character-specific trace, or physical continuity with adjacent shots, capture a real plate first and use generation for replaceable environmental extension. AI is good at expanding space. It should not invent a consequence that changes the facts of the story.
Leave a relationship, not merely a few seconds
A good empty shot should do at least one thing: make the audience wait for an entrance, see the result of a departure, feel a person through a trace, or read time through a changed environment.
Before the shoot, circle every storyboard panel labelled “empty,” “atmosphere,” or “transition.” Add the cause, visible change, sound, and next-shot payoff. Any image that cannot answer those questions may still be captured as insurance, but it should not occupy the place of a story beat.
If you are planning a brand film, product commercial, or hybrid live-action and AIGC production, share the script, references, and target versions with ONCE VISUAL. We can turn character actions, product consequences, and spatial states into a shootable shot map, then decide which empty shots deserve to remain in the final cut.
Source note: this article draws on Adobe's official B-roll guide and Firefly tutorial, Google's official ABCD video-ad guidance, Magliano and Zacks' 2011 study of continuity editing and event segmentation, and the 2025 Humanities and Social Sciences Communications study of empty-shot distribution across 2,464 films. These sources help explain creative mechanisms; they do not establish that any individual shot will produce a specific advertising result.
