Vizard Agent

How to add a chapter progress bar to a demo video

Last updated 2026-09-01 · 8 min read

Upload the screen recording and name the sections. Vizard Agent builds a progress bar that advances with the video and labels each chapter along it, taking the bar's colour directly from your own interface screenshot so it reads as part of the product.

What is the short version?

A five-minute product demo loses people for a mundane reason: they cannot see how much of it is left, or what is still to come. A labelled bar running along the bottom answers both of those questions continuously and costs almost nothing in screen space.

  1. Go to Vizard Agent and upload the recording and its narration.
  2. Name the sections the demo covers, in order.
  3. Give it a screenshot of your interface so the colours match.

What do you need before you start?

The recording, the narration and a screenshot. Vizard Agent samples your brand colour out of the screenshot rather than asking you for a hex value, which is usually quicker than finding the value and always more accurate than guessing it.

What do you type into Vizard Agent?

Say what the sections are and let it handle the placement. Vizard Agent works out the timing from the narration, so the useful instruction is the list of chapters rather than where on the frame the bar should sit.

Prompt

Make a feature introduction video from this screen recording and voiceover. Requirements: 1) The narration and the screen must line up — show whatever is being talked about. Where the screen sits still on one view for a long time, add related supporting shots, each no more than [2–4] seconds. 2) Crop sensibly so viewers can see what matters and the irrelevant parts of the screen are gone. 3) Add a progress bar labelled with each section's topic, advancing as the video plays.

Variants worth knowing:

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real feature-introduction video built from a screen capture and a separate narration track. A third of the session goes on one question: where exactly does the application window sit in the frame, and does it stay in the same place throughout?

  1. Gets the durations of the audio and the video and transcribes the narration with timestamps.
  2. Runs a shot transcription of the picture to see what happens on screen.
  3. Extracts key frames to see the interface at each point.
  4. Reviews a dashboard frame to find the unnecessary edges.
  5. Detects the browser window's coordinates and saves a cropped test.
  6. Checks the coordinates again at sixty seconds to see whether the window moves.
  7. Reviews the loading and preview frames to map the desktop background across the whole video.
  8. Detects the window region in every extracted frame to see how it changes with zoom.
  9. Crops all the frames to a fixed region and checks nothing important is lost.
  10. Verifies the crop at 48, 60 and 85 seconds against the tools each frame needs to show.
  11. Searches for supporting stock footage for the long static stretches and downloads it.
  12. Extracts the brand's colour from the interface screenshot, finding the highly saturated pixels.
  13. Lists the installed fonts and locates one that renders the section labels.
  14. Downloads the inputs locally and checks every clip's frame rate for consistency.
  15. Builds the timeline slice by slice — cropping, re-timing, scaling, muting and inserting the supporting shots.

Step twelve is the small thing that makes the bar look native. Vizard Agent samples the product's own colour out of a screenshot rather than approximating it, so the progress bar reads as part of the interface instead of a graphic laid over it.

What does the result look like?

The recording this page is written from measured 1910x1080 and ran 131.88 seconds. The finished video crops to the application window, cuts away from static screens with short supporting shots, and carries a labelled bar advancing along the bottom in the product's own colour.

The effect on watching is specific: you always know which section you are in and roughly how much is left, which is exactly the information a long demo normally withholds.

When does this not work well?

A progress bar takes up screen space and makes a promise about structure, and both of those can go wrong. Vizard Agent will tell you when your sections do not divide the runtime cleanly, rather than putting arbitrary marks along the bar to make the design work.

How do you fix a result that came back wrong?

Say which section is wrong, or which part of the crop. Vizard Agent keeps the detected window coordinates, all the section timings and the colours it sampled, so a renamed chapter or a slightly wider crop is a re-render rather than a rebuild of the timeline.

How does Vizard Agent compare to doing it yourself?

By hand this is a bar animated across five minutes with labels hand-keyed to the narration, and every re-record means re-timing all of them. Vizard Agent derives the timings from the narration and the colour from your own interface.

By hand Vizard Agent
Section timings Hand-keyed to the audio Derived from the narration
The crop One box, set once Verified at several timestamps
Brand colour Typed from a style guide Sampled from your screenshot
Static stretches Left, or sped up Covered with short supporting shots

Common questions

How many sections work? Five or six. Vizard Agent will say when the labels stop fitting.

Where does the bar sit? Along the bottom, usually. Vizard Agent keeps it clear of the interface and the captions.

Will it match my brand colour? Yes. Vizard Agent samples it from a screenshot rather than approximating.

Can it crop to just my app? Yes. Vizard Agent detects the window and checks the crop at several points.

What if the window moves? It checks for that. Vizard Agent detects the coordinates at multiple timestamps.

What happens on static screens? Short supporting shots. Vizard Agent caps each one so they do not take over.

Does it need a separate narration file? No, but it helps. Vizard Agent uses whatever audio you supply.

Can the labels be in my language? Yes. Vizard Agent finds a typeface that renders the script.

Does the bar work vertically? Less well. Vizard Agent has far less width for labels in a vertical frame.

Can it add chapter markers instead? Yes. Vizard Agent can output the timestamps for your platform's chapters as well.

Why does the bar help at all? Because a demo asks for five minutes without saying what for. Showing the sections and the position turns an open-ended commitment into a visible one, which is why people stay.

Can I reuse it for the next release? Yes. Vizard Agent keeps the styling and rebuilds it around the new recording.

Does it check the result? Yes. Vizard Agent reviews the crop and the frames before rendering the final cut.