How to put a horizontal clip inside a vertical video

Last updated 2026-08-30 · 7 min read

What is the short version?

Joining a clip you filmed sideways to one you filmed upright is two problems: what to do with the empty space above and below the horizontal one, and the fact that two files from different sources often refuse to join cleanly at all.

1. Go to Vizard Agent and upload both clips.
2. Say which one opens and which follows.
3. Say what should happen at the join.

What do you need before you start?

Both clips and a clear order. Vizard Agent will ask rather than guess if the order is ambiguous, which is worth knowing because "the second one first" is a sentence that reads two ways and costs a round when it is read the wrong one.

• Both clips. Whatever orientation each is.
• The order. Unambiguously stated.
• The transition. A cut, a glitch, a fade.
• The final shape. Vertical, usually.
• The audio. Which clip's sound survives.

What do you type into Vizard Agent?

State the order by describing the clips rather than numbering them. "The taxi clip opens, then the aquarium footage" cannot be misread, whereas "the second one first" is exactly the kind of instruction that produces a video in the wrong order.

Prompt

Variants worth knowing:

• A long, obvious transition. Rather than a quick cut.
• A sound on the join. A line or an effect at the switch.
• Filled or letterboxed. What sits above and below the horizontal clip.

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real two-clip join. The step that explains the first failed render is the sixteenth: the two files had different time bases, which is a mismatch no amount of editing fixes.

1. Checks both clips' pictures and durations.
2. Reviews the available glitch transition effects and looks at representative frames from each clip.
3. Checks the pacing, the endings and the levels of both.
4. Reviews the full rhythm of the first clip and the full action of the second.
5. Checks the opening and closing composition of the second clip, then the ending of the first.
6. Confirms which glitch effects the current tools support.
7. Measures the transition noise and writes the final edit script.
8. Locates a syntax problem in the filter chain, fixes it and re-renders.
9. Removes a stray empty separator, unifies both clips' time bases, re-renders, then checks the frames, the transition and the audio waveform.

Step 9's middle is the one worth remembering. Two clips recorded on different devices carry different time bases, and joining them without reconciling that produces either a failure or a video whose audio drifts away from the picture.

What does the result look like?

The horizontal source clip measures 1920x1080 at 30fps and runs 5 seconds. In the finished vertical video it sits centred and sharp at its own proportions, with the frame filled above and below rather than the picture being stretched, followed by a deliberately long glitch transition into the vertical footage.

Nothing is distorted. That is the whole reason to handle the join properly rather than scaling the horizontal clip up until it fills the frame, which is what most quick edits do and what every viewer notices.

When does this not work well?

Fitting one frame shape inside another always costs something somewhere, and joining files recorded on different devices introduces a second set of problems that have nothing to do with the edit at all. These are both worth understanding first.

• Something has to fill the gap. Bars, a blurred backdrop, or a crop.
• Cropping loses the sides. A wide shot cut to vertical loses most of itself.
• Low-resolution clips suffer. Enlarging to fill a vertical frame softens them.
• Mismatched frame rates need reconciling. So do time bases and audio rates.
• Ambiguous orders cost a round. Describe the clips rather than numbering them.

How do you fix a result that came back wrong?

Say which clip or which join is wrong and what it should be. Vizard Agent keeps both clips' analyses, all the composition checks and the edit script, so the fit or the transition can change without rebuilding the whole video.

• "The taxi clip looks squashed." Re-fitted at its own proportions.
• "The transition is too quick." Extended; the clips stay as they are.
• "They are in the wrong order." Rebuilt from the same analyses.

How does Vizard Agent compare to doing it yourself?

By hand this means scaling one clip to fit, deciding what goes in the empty space, and then finding that the two files will not concatenate — which presents as an error message about codecs rather than as anything you can act on.

Common questions

Will the horizontal clip be stretched? No. Vizard Agent fits it at its own proportions and fills the space around it.

What goes above and below? Your choice — bars, or a filled backdrop. Say which one and Vizard Agent builds it.

Can it crop instead? Yes, though Vizard Agent will warn you that a wide shot cropped to vertical loses most of itself.

Why did my own attempt fail to join them? Usually a time-base or frame-rate mismatch. Vizard Agent reconciles those before joining.

Can I have a specific transition? Yes. Vizard Agent checks which effects are available and builds the one you name.

How should I describe the order? By what is in each clip rather than by number. "The taxi one opens" cannot be misread.

Can sound play over the join? Yes. Vizard Agent can place a line or an effect at the switch.

Does this work with more than two clips? Yes. Vizard Agent applies the same reconciliation across as many clips as you have.

Which clip's audio survives? Whichever you say. Vizard Agent keeps both tracks if the levels allow it.

Is a filled backdrop better than black bars? Usually, on a phone. Vizard Agent can build a soft blurred version of the clip itself behind the sharp centre, which reads as deliberate where two black bands read as a mistake.

Can the whole thing be widescreen instead? Yes, and it is the easier direction. Vizard Agent fits the vertical clip inside a wide frame the same way, and a vertical clip in a wide frame usually needs a filled backdrop rather than bars too.

Should the transition be long or short? Whatever the join needs. Vizard Agent extended the glitch deliberately here because the two clips share nothing visually, and a hard cut between unrelated shots reads as a mistake rather than a choice.

Does Vizard Agent check the finished video? Yes. It reviews the frames, the transition and the waveform before delivery.