How to make a team line-up video with the right name on every face

Last updated 2026-09-22 · 8 min read

What is the short version?

The work is not the animation. It is that eleven photos and eleven names have to stay paired through every step, and the one thing your colleagues will notice is the person whose card says somebody else's name.

1. Tell Vizard Agent which photo is which person, in order.
2. Give it a reference video for the layout, transitions and music.
3. Say whether more members will be added later.

What do you need before you start?

The photos, the names in the same order, and something to match. Vizard Agent can invent a layout, but a team video usually has to look like the last one, and a reference video settles a dozen small decisions at once.

• The photos. One per person, numbered.
• The names. In the same order, spelled as they should appear.
• Any titles. Team lead, department, nickname.
• A reference video. The style you want matched.
• Whether it will grow. It always does.

What do you type into Vizard Agent?

Write the list plainly, one line per person. The pairing is the fragile part of this job, and a numbered list is far more robust than a description — especially when the names are nicknames that no amount of face-reading would guess.

Prompt

Variants worth knowing:

• "Photo 1 is…" Numbered pairing, not description.
• "Same style as this reference." Layout, transitions and music in one go.
• "The team lead first." Order carries meaning in a line-up.

What does Vizard Agent actually do?

Here is the Vizard Agent sequence on an eleven-person line-up built to match an existing team video, with a twelfth member added to it afterwards. The reference gets studied in detail before a single card is designed, because almost every decision in a video like this is already settled by the old one.

1. Looks at each of the eleven photos in turn.
2. Measures the reference video and extracts frames to read its layout.
3. Pulls the key timestamps to see how each slide is composed.
4. Checks the reference's audio and its transition points.
5. Finds a clean background frame to match the reference's own backdrop.
6. Designs thirteen cards in that style — the members plus the opening and closing.
7. Fixes the hexagonal frame treatment and exports every slide.
8. Reviews all the slides together before assembling anything.
9. Builds the video with the reference's transitions and music.
10. Checks the render on a contact sheet and a waveform.
11. Reviews the upload for transitions and audio sync.
12. Swaps in a new photo at position five and rebuilds only what changed.

Step eight is the check that protects the pairing. Looking at every card laid out together, before the video exists, is the moment a wrong name is cheap to fix — after the render it costs another pass and after publication it costs an apology.

Step two matters because "the same style" hides a lot: the crop shape, where the name sits relative to the face, how long each card holds, whether the music lands on the change. Vizard Agent reads those off the reference rather than approximating them.

Step twelve is the reason to keep the project. Teams change, and a replacement card built from the same design drops into the existing video without anyone being able to tell which one is new.

What does the result look like?

A short video that steps through the team in the order you gave, each face with its own name card in the reference's layout, transitions landing with the music, and a closing card. Every pairing is checked on a sheet before assembly, and adding a new member later produces a card that matches the others.

When does this not work well?

Some photo sets cannot be made to match. Wildly different crops, lighting and resolutions between eleven phone photos will show in a grid of identical frames, and Vizard Agent will suggest a consistent treatment rather than pretending the set is uniform.

• Very mixed photo quality. One blurry photo stands out in a grid.
• Different crops and angles. Head sizes vary card to card.
• Names you have not supplied. Faces cannot be identified for you.
• A reference you do not have. Style has to be described instead.
• Very large teams. Past a couple of dozen the format stops working.

How do you fix a result that came back wrong?

Name the position and the person in the same breath. Because Vizard Agent builds the video from numbered cards rather than one long timeline, it can correct a single pairing, restyle one card or replace one photo without disturbing the ten cards that are already right.

• "Card 5 has the wrong name." That pairing is corrected and only that card rebuilt.
• "Use this photo for number 3." The new photo is dropped into that position.
• "His title is wrong." The subtitle line on that card is changed.
• "Add a twelfth member." A matching card is built and placed where you say.

How does Vizard Agent compare to doing it yourself?

By hand this is a template and a lot of copy-paste, which is exactly the process that produces a wrong pairing. You duplicate a slide, change the photo, forget the name, and the mistake survives every review because everyone reads the face and not the text.

Common questions

Can it read the names off the photos? No, and it should not guess. Give Vizard Agent the list.

What if two people have the same name? Use the numbers. The pairing follows the order, not the name.

Can it add job titles? Yes, as a second line on each card if the reference has one.

Does every photo need the same size? No. Vizard Agent crops them to a consistent frame, though very different photos still show as different.

Can it match our old team video exactly? That is what the reference is for — Vizard Agent measures it rather than eyeballing it.

What about the music? It can follow the reference's track or Vizard Agent can generate one that fits the cuts.

Can I reorder them later? Yes. Vizard Agent keeps the cards separate, so order changes are cheap.

How long should each card hold? Long enough to read the name; Vizard Agent takes the timing from the reference.

Can it do a vertical version? Yes, from the same cards, though the layout is rebuilt for the frame.

What if someone leaves? Vizard Agent removes that card and closes the gap, keeping the music in step.

Can it animate the cards? Yes, and Vizard Agent matches the reference's motion as part of the design step.

Will the name spelling be exactly mine? Yes. Vizard Agent uses your list verbatim, diacritics included.

Can it build a version with photos only? Yes, though names are usually the point of a line-up.

How do I check it before sending? Read the names, not the faces — that is where the one error will be.