Vertical Video: Reframe the Meaning, Not Just the Canvas

A portrait version of a tutorial can look beautifully composed while leaving the viewer unable to follow the instruction. The presenter stays in frame, but the object she is pointing towards has disappeared. The video has a new shape; the explanation has lost a necessary part.

For independent creators, adapting a landscape recording deserves an editorial decision before an export decision. Ask what the viewer needs to see at each moment, then choose a composition that carries that meaning. The following worked example and review method are original editorial guidance, rather than findings from a platform test.

Start with the action the viewer must understand

Imagine a fictional artist demonstrating how to align a print inside a mount. The camera shows her hands, the print and a reference mark near the right edge of the work surface. She says, “Keep this corner level with that mark.” The relationship between two visible details is the instruction.

If a portrait adaptation shows only the hands, the words remain audible but their reference is missing. Keeping the most prominent subject centred would not solve that particular problem. The important unit is the relationship between the corner and the mark.

Before editing, write a short sentence for each essential moment: “The viewer must compare these edges”; “The viewer must see which fastener moves”; or “The viewer must read this label.” This turns the adaptation into a series of concrete decisions.

What the frame arithmetic tells us

Suppose the source image is 1920 pixels wide and 1080 pixels high, a 16:9 frame. For this calculation, assume a 9:16 portrait crop that keeps the full source height, fills the portrait frame and does not stretch the picture.

At a height of 1080, the crop’s width is 1080 × 9 ÷ 16 = 607.5 source pixels. A perfectly centred mathematical crop would run from horizontal position 656.25 to 1263.75. These fractional boundaries describe the geometry; an actual editor must implement its own pixel sampling.

Now place one essential detail at horizontal position 250 and the other at 1650. Their separation is 1400 source pixels, wider than that crop. Moving the same narrow window left or right cannot include both simultaneously. The arithmetic was checked for this article; no actual video was exported or assessed.

The example establishes a constraint, not a universal editing rule. A different scale, composition or shot can change what fits. It simply shows why finding a better crop position cannot always preserve a wide visual relationship.

Choose how to preserve the explanation

For the fictional mount demonstration, several editorial options are available:

• Retain the wider view within the portrait composition. Leave space around it and use that space deliberately, while checking whether the essential details remain large enough to recognise.
• Reshoot the critical step. Arrange the objects so the corner and reference mark can be understood together in the intended composition.
• Use a sequence of closer views. Establish the relationship first, then show the relevant detail, provided the edit makes the connection clear.
• Revise the instruction. Replace an ambiguous reference with wording and imagery that agree about what the viewer should do.

These are proposed creative choices, not promises about features in a particular application. Select the option that preserves the lesson with the resources available. A short replacement shot may be more useful than repeatedly adjusting a crop that cannot contain the required relationship.

Give AI-assisted editing a meaning-based brief

If an assistant is helping prepare an edit plan, give it the purpose of the scene rather than only the requested aspect ratio. A useful brief would say: “The viewer must understand how the print corner aligns with the reference mark. Identify moments where both are needed. Propose a portrait composition or replacement shot for those moments, and flag anything you cannot verify from the supplied material.”

Treat the response as a proposal. If the assistant has only a transcript, it cannot establish where the objects actually sit in the picture. Supply the relevant visual material through a supported route, or make the visual decisions yourself. Do not turn a text-only planning exercise into a claim that the finished frames were inspected.

Ask for a time-based decision list: the instruction, the essential visual detail, the proposed composition and the point requiring review. That list can guide an editor without pretending that a completed export already exists.

Review transitions as well as individual frames

A still frame can show the right object while the sequence remains confusing. In the fictional tutorial, moving from a view of the left corner to a view of the right mark needs enough context for the viewer to connect them. Decide whether a wide establishing view, a pause or revised narration would make the relationship clearer.

I would review the adaptation once for the action alone: can someone follow what moves, what remains fixed and what the comparison means? Then review picture and narration together. Check words such as “here”, “this” and “that” against what is visible at the moment they are spoken.

Keep a note of the moments that need a different shot instead of accepting a vague feeling that the video looks awkward. A precise note gives the next edit a useful target.

Recheck text after changing the composition

Captions and other on-screen text need their own pass in the portrait version. Check that they remain readable and do not cover the detail needed for the instruction. If text is already part of the source picture, inspect what the new composition has removed before deciding the adaptation is complete.

MaryChuks’ guide to keeping an editable caption track with the final cut provides a related handover discipline. When the cut changes, keep the reviewed caption version connected to the video it actually accompanies.

For this tutorial, a label explaining the reference mark should agree with the shot and the spoken instruction. Adding a label cannot make an unseen relationship self-evident; it still needs a clear visual explanation.

Keep the source and the approved adaptation identifiable

A compact handover can contain the original recording, the approved portrait cut, its editable caption material and a short note identifying replacement shots. Give the versions clear names so a later correction starts from the intended edition.

Before approving the portrait cut, use four questions:

• Can the viewer see every detail required to understand the instruction?
• Do the picture, narration and text refer to the same thing at the same moment?
• Are changes of view clear enough to preserve the relationship between details?
• Has the actual finished export been watched through, with any remaining limitations recorded?

A creator can preserve the same lesson while changing its presentation considerably. The useful completion standard is whether the intended viewer can still understand the demonstration. Let that decision lead the reframing work, from the first crop proposal to the final reviewed cut.

Sources and example basis

• MaryChuks: Video Captions — Keep the Editable Track with the Final Cut, related editorial guidance; publication and link verified 30 October 2026.
• The artist, demonstration and coordinates in this article are fictional. Frame dimensions and crop boundaries are original worked calculations. The review method is proposed editorial practice, not a reported experiment or platform capability claim.