Vizard Agent

How to make a trade show booth video that works with the sound off

Last updated 2026-08-31 · 8 min read

Upload the product videos you already have and say what must go. Vizard Agent removes the presenters, intros and manufacturer end screens, finds the close-ups of the equipment operating, and captions them at a size that reads from across an exhibition hall.

What is the short version?

A booth screen has no sound, no captive audience and about four seconds to stop someone walking past. That makes it a genuinely different format from a social video, and most companies play a product video that was never built for it.

  1. Go to Vizard Agent and upload your existing product videos.
  2. Say what to cut: presenters, long explanations, title screens, end cards.
  3. Say it has to be understandable with the sound off.

What do you need before you start?

Existing footage is usually enough, which is the point. Vizard Agent pulls the close-ups out of videos you already published, so a booth loop can be assembled from material that was shot for a completely different purpose without filming anything new.

What do you type into Vizard Agent?

Write it as a specification, because that is what it is. Vizard Agent responds well to explicit prohibitions here — no fictional equipment, no unsupported claims, no social-media caption style — since a booth video is a technical document as much as a piece of marketing.

Prompt

Make one [60] second [16:9] exhibition video from the uploaded footage only. Introduce [our product line]. Use only authentic equipment footage. No fictional equipment or unsupported technical claims. Prioritise close-ups of the instruments, controls and equipment operating. Remove presenters, long explanations, intros, existing title screens and manufacturer end screens. Understandable without sound. Large concise captions readable from a distance. No emojis or social-media caption styles.

Variants worth knowing:

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real laboratory equipment exhibition video built from five source videos. Steps eight to eleven are the ones that make it possible: timestamped contact sheets of every source, so clips can be chosen by looking rather than by scrubbing.

  1. Checks the workspace and finds your logo, downloading the source videos.
  2. Reads the website for the correct logo and brand details.
  3. Probes the metadata of all five downloads.
  4. Extracts timestamped contact sheets from every source video.
  5. Inspects the frames from each product's footage in turn.
  6. Extracts high-resolution detailed contact sheets and reviews each one.
  7. Searches for scientific and technology-style music, downloads candidates and analyses the fit.
  8. Extracts a clean transparent version of the logo from a flat image file.
  9. Hunts the opening hook — the most arresting few seconds available.
  10. Extracts every frame of the lead product's footage and reviews the full breakdown.
  11. Cuts candidate clips, tiles them and builds a master preview grid of everything selected.
  12. Refines the trim points to land exactly on the equipment operating.
  13. Renders the titles, the corner logo bug and high-visibility caption cards.
  14. Assembles the timeline and renders the composite with music, overlays and graphics.
  15. Extracts quality-control frames and the audio waveform, then reviews every scene.

Step twelve deserves attention. Vizard Agent refines each trim so the clip starts on the machine actually running rather than a second before, because on a booth screen a clip that opens on a static object has already lost the person walking past.

What does the result look like?

The brief this page is written from asked for a single 60-second, 16:9 exhibition video assembled from five existing product videos, with a title card, a corner logo bug, large caption cards and a scientific-sounding underscore beneath it. Every clip in it came from footage the company had already published somewhere else.

What it delivers is the opposite of a social edit: no talking head, no emoji captions, no manufacturer outro, and every clip chosen for showing an instrument in operation. It reads perfectly with the sound off, which is the only way it will ever be seen.

When does this not work well?

An exhibition video is constrained twice over: by what your existing footage actually contains, and by the claims you are entitled to make about your own equipment. Vizard Agent reports the gaps it finds rather than filling them with generated machinery that does not exist anywhere outside the video.

How do you fix a result that came back wrong?

Name the clip or the caption. Vizard Agent keeps the contact sheets, the selected clips and the caption cards, so a trim point or a wording change is a re-render of one segment rather than a rebuild of the whole timeline.

How does Vizard Agent compare to doing it yourself?

By hand this is a day of scrubbing five long videos for the twenty seconds worth using, then discovering the manufacturer's end screen is baked in, then re-typing captions because they are unreadable at three metres. Vizard Agent selects by looking at contact sheets and sizes captions for the distance.

By hand Vizard Agent
Finding usable clips Scrub five videos Timestamped contact sheets
Removing presenters Cut around them Excluded at selection
Caption size Guessed at desk size Sized to be read across a hall
The logo Whatever file you have Extracted clean and transparent

Common questions

Can I use existing product videos? That is the intended route. Vizard Agent extracts what is useful and discards the rest.

Will it remove the presenter? Yes. Vizard Agent selects clips of the equipment rather than of anyone talking about it.

What about manufacturer end screens? Cut. Vizard Agent excludes existing title cards and outros at the selection stage.

How big should the captions be? Bigger than feels right. Vizard Agent sizes them to be read from several metres away.

Should it have music? Yes, but nobody will hear it. Vizard Agent chooses something that suits the hall if the sound is on.

Can it use our logo? Yes. Vizard Agent extracts a clean transparent version even from a flat image file.

Does it loop? Build it to. Vizard Agent can match the closing frame to the opening one for a seamless loop.

How many products fit? About four in a minute. Vizard Agent gives each one a few seconds of genuine operation.

Will it invent specifications? No. Vizard Agent captions only what the footage shows and what you have supplied.

What aspect ratio should I use? Whatever the booth screen is. Vizard Agent builds landscape, vertical or square from the same sources.

Can I get a social version too? Yes, and you should ask separately. A booth cut and a feed cut want different pacing and different captions, and Vizard Agent builds each properly rather than resizing one into the other.

What if my footage is low resolution? It will show on a large panel. Vizard Agent tells you which sources are too soft before it builds around them.

Does it check the finished video? Yes. Vizard Agent extracts quality-control frames and reviews every scene before delivery.