How to make a digital wedding invitation video

Last updated 2026-09-01 · 8 min read

What is the short version?

A digital invitation is not a card with music over it. It is a short sequence of physical actions — something pressed, something opened, something untied — and the whole charm is that each one looks like it really happened.

1. Go to Vizard Agent and describe the opening, action by action.
2. Give the wording that appears on the card.
3. Say the feel — warm and traditional, or modern and spare.

What do you need before you start?

The sequence and the wording. Vizard Agent generates the objects, so you do not need photographs of stationery you have not bought — but the order of the actions is the script, and it needs to be in the order a person would actually open it.

• The actions. Seal, envelope, ribbon, card, in order.
• The wording. Names, date, time, place.
• The materials. Wax colour, paper, ribbon.
• The feel. Formal, warm, minimal.
• The length. Fifteen to thirty seconds.

What do you type into Vizard Agent?

Write the brief as a sequence of physical actions rather than as a description. Vizard Agent generates one shot per action, so listing them in the order they happen gets you a sequence — whereas asking for "an elegant wedding invitation" gets you a picture of one.

Prompt

Variants worth knowing:

• List the actions in order. That is the whole structure.
• Name the materials. Wax colour and paper change everything.
• Ask for matched sound. The seal and the paper should be heard.

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real invitation. Steps seventeen to twenty-three are the important ones: every shot is stepped through half a second at a time to check the action reads before it is kept.

1. Checks the available tools and effect recipes.
2. Checks what the stock search and video generation can do.
3. Searches stock footage for wax seals, envelopes, ribbons and invitations.
4. Searches again with a script when the first passes come up short.
5. Checks the image and video generation options.
6. Checks the music and sound-effect options.
7. Generates the invitation shots and searches for romantic music.
8. Extracts and tiles preview frames for inspection.
9. Reviews the contact sheet of the wax stamp, the envelope and the ribbon.
10. Generates the music and synchronised sound effects.
11. Generates a high-detail image of the invitation card itself.
12. Reviews that card at full size.
13. Probes every asset's duration and format.
14. Extracts frames at half-second intervals across each shot.
15. Reviews the progression of shot one, then two, then three — the press, the opening, the untying.
16. Regenerates the wax stamp shot when the first version does not convince.
17. Generates a final slow zoom onto the invitation card.

Step fifteen is the check that decides whether it works. Generated video is very good at making a wax seal exist and quite bad at making it press — so each shot is watched frame by frame, and the one that failed here was made again.

What does the result look like?

The invitation this page is written from is a short sequence: a wax seal pressed, an envelope opening, a ribbon untying, and a slow move onto the card, with romantic music and a sound effect matched to each physical action.

The sound is doing more work than it seems. The soft thud of the seal and the paper sliding are what convince the eye that these are objects rather than pictures of objects.

When does this not work well?

Generated physical actions are by far the hard part of this, and the small hands-on movements that make an invitation feel real are the hardest of all to produce convincingly. These are the places where it will need more attempts, or a different approach.

• Hands in shot. Fingers are where generation fails first.
• Long sequences. Four actions is about the limit.
• Exact stationery. It will resemble yours, not match it.
• Text on the card. Check every character yourself.
• Cultural specifics. Describe the customs; it will not assume them.

How do you fix a result that came back wrong?

Name the action that is wrong rather than the shot number. Vizard Agent keeps every generated shot along with its frame-by-frame checks, so a regenerated moment drops straight back into the same sequence and the same sound design around it.

• "The seal does not press." That shot regenerated and re-checked.
• "The ribbon unties backwards." Re-generated with the motion described.
• "The date is wrong on the card." Card re-rendered with the corrected text.

How does Vizard Agent compare to doing it yourself?

By hand this is either a stationery photoshoot or a template with your names typed into it, which is what most digital invitations are. Vizard Agent generates the actions and then checks each one actually reads as that action.

Common questions

Do I need any footage? No. Vizard Agent generates every shot from your description.

How long should it be? Fifteen to thirty seconds. Vizard Agent uses three or four actions in that.

Can it match our colours? Yes. Vizard Agent takes the wax colour, the paper and the ribbon from your description.

Will the card text be right? Check it yourself. Generated lettering needs a human to read every character.

Can it show our actual stationery? Only if you photograph it. Vizard Agent can animate a photograph of the real thing instead.

Does it add sound effects? Yes, matched per action. Vizard Agent generates the seal, the paper and the ribbon separately.

What if a shot looks wrong? It regenerates it. Vizard Agent frame-checks each one before keeping it.

Can we send it on a phone? Yes. Vizard Agent renders vertical for messaging apps.

Can it do other invitations? Yes. Vizard Agent handles birthdays, launches and openings the same way.

Can we have two language versions? Yes. Vizard Agent re-renders the card and keeps the same sequence.

Why check every shot frame by frame? Because a wax seal that hovers a millimetre above the paper reads as wrong to everyone and as explainable by nobody. The action either happens or it does not, and that is only visible frame by frame.

Can we add our photo? Yes. Vizard Agent can end on a photograph of the couple instead of the card.

Does it check the result? Yes. Vizard Agent reviews each shot's frame-by-frame progression before assembling the sequence at all.

Can we include a map or directions? Yes. Vizard Agent adds a card at the end with the venue details rather than crowding the invitation itself.