How to trigger a transition on an action in the footage

Last updated 2026-09-05 · 8 min read

What is the short version?

A transition triggered by an action only works if it starts on the right frame. One frame early and it anticipates the movement; one frame late and it feels like a delay. Finding that frame is the job, and it is done by looking rather than by scrubbing.

1. Go to Vizard Agent with both clips.
2. Name the action the transition should start on.
3. Send a reference if you have one.

What do you need before you start?

The clips and a description of the moment. Naming the body part or the object that moves — the foot landing, the hand closing, the ball hitting — narrows the search enormously, because the exact frame can then be found on that part of the picture rather than on the whole of it.

• Both clips. The one that leaves and the one that arrives.
• The action. As specific as you can make it.
• What the transition does. Falls, slides, wipes, whips.
• A reference, if you have seen the effect somewhere.
• The format. Vertical changes the geometry.

What do you type into Vizard Agent?

Describe the mechanics, not the vibe. "A cool transition" and "the first video falls downward until it clears the frame, and the second is revealed underneath" produce very different results, and only the second one can be checked afterwards.

Prompt

Variants worth knowing:

• "Where her foot lands." A frame, not a timestamp.
• "Until it reaches the bottom." Says when it ends.
• "Underneath." Says what is revealed and how.

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real two-clip edit whose transition had to fire on a footfall. The reference gets examined twice — once for how the effect runs and once for what happens after it.

1. Gets motion and timing notes for both clips.
2. Builds contact sheets of the landing frames and the insert frames.
3. Inspects the landing frames to see the moment closely.
4. Refines the exact frame using a strip of the lower body only.
5. Views that refined frame strip to confirm the choice.
6. Analyses the reference transition's mechanics in detail.
7. Builds the falling transition on the chosen frame.
8. Renders the combined edit and extracts review frames.
9. Watches the transitions back for motion and sync.
10. Enlarges the reference's transition frames for exact layout matching.
11. Compares the reference's layout against the delivered edit's.
12. Inspects how the reference resolves after the effect finishes.
13. Rebuilds to match those mechanics and re-checks.

Step four is the technique worth stealing. Rather than examining the whole frame around the moment, only the part of the picture that carries the action is pulled out — a strip of the lower body for a footfall — which makes the landing frame obvious instead of arguable.

Step two is quietly doing a lot. Contact sheets are made of both the outgoing landing frames and the incoming clip's first frames, so the trigger and the thing it hands over to are chosen against each other rather than one at a time.

Step twelve is the half that gets skipped. Everybody studies what a reference transition looks like while it runs; what makes an edit feel like the reference is usually what happens in the second after it, when the incoming shot settles.

What does the result look like?

A transition that starts on the frame the action lands, moves the way you described, and hands over to the second clip the way your reference does — including how the new shot settles once the movement stops.

It reads as one continuous idea rather than two clips joined.

When does this not work well?

The action has to be visible and definite. A soft movement, a blurred frame, or a moment hidden behind something else gives no single frame to fire on, and the transition then has to be placed by feel like any other cut.

• No clear landing frame. Motion blur hides it.
• The action is off screen. Nothing to trigger on.
• A slow movement. There is no single moment.
• Mismatched formats. The geometry has to be rebuilt.
• A reference you cannot share. Describe the mechanics instead.

How do you fix a result that came back wrong?

Say whether the timing or the movement is wrong, because they are separate. Vizard Agent keeps the frame strip and the reference analysis, so shifting the trigger by a frame is a different fix from changing how the transition behaves.

• "It fires too early." Moved to the next frame in the strip.
• "It does not look like the reference." The layout re-matched.
• "The second clip arrives oddly." Its settle re-built to match.
• "Wrong direction." Rebuilt to fall the other way.

How does Vizard Agent compare to doing it yourself?

By hand the trigger frame gets found by scrubbing, which lands you within a few frames and rarely on the right one. References get watched at normal speed, where the mechanics are impossible to see and the layout is impossible to measure.

Common questions

Do I need a reference? No, but it helps. Vizard Agent will match its mechanics.

How exact is the trigger frame? To the frame. Vizard Agent narrows it on the moving part.

Can it work on a hit or an impact? Yes. Name the moment and Vizard Agent finds it.

What about music instead of an action? Yes, that is a beat grid; Vizard Agent can use either.

Can the transition go upwards? Yes, or sideways. Say which and Vizard Agent builds it.

Will it work in vertical? Yes, though Vizard Agent rebuilds the geometry for the format.

Can it stack more than two clips? Yes. Vizard Agent chains the transitions in order.

Will the audio jump at the join? Not if you say so. Vizard Agent carries it across.

Can it match a reference exactly? Vizard Agent matches the mechanics; an exact copy depends on your source.

What if the action is blurred? Vizard Agent will say so and place the trigger by feel.

Can I see the candidate frames? Yes. Ask Vizard Agent for the strip it used.

Does it slow down or speed up around it? Only if you ask. Say what you want at the moment of impact.

Can it add a sound on the hit? Yes, and Vizard Agent finds a hit sound usually sells the effect.

Why look at a strip rather than scrub? Because scrubbing shows you one frame at a time, and the right frame is only obvious next to the frames on either side of it.