Version for platforms

What this is about

A horizontal frame cropped to vertical loses roughly half its width, and the half it loses is usually where the second person, the product or the graphic was standing. Automatic reframing tools follow motion and lose composition; manual reframing keeps composition and costs time. Either way the decision belongs in the edit, not in an upload dialogue. On top of the geometry sit the platform habits: interface elements covering the lower edge, playback starting muted, and viewers deciding within the first moments whether to keep watching. Versioning is what turns one film into something each of those environments can actually carry.

The order matters more than the tools. Skip a step and you usually notice two steps later — by which point the fix costs several times as much.

What runs differently here

Where this differs from the general case:

• Shoot with the target frames in mind: leave headroom and side room so a vertical crop still contains the subject
• Keep essential action and text inside a safe area away from the edges where interface elements appear
• Assume sound is off at start — the opening has to work as pictures and text alone
• The hook belongs at the very front in feed formats and can breathe in a player embedded on your own site
• Check the current duration and format spec of each channel rather than trusting a number you memorised

Step by step

This order gets you there fastest:

1. List the channels and the frame each one needs before the edit starts. Vertical, square and wide are three different compositions, and knowing that early changes how scenes are shot.
2. Build the primary edit in the format that carries the most content, usually the widescreen version, and treat the others as deliberate adaptations rather than exports.
3. Reframe shot by shot for each aspect ratio. Automatic tracking is a starting point; a person deciding where the subject sits in every shot is what keeps compositions intact.
4. Redesign text and graphics per format instead of scaling them. A lower third built for wide becomes unreadable when squeezed, and a vertical frame has room for larger type.
5. Move the hook forward in feed formats. What can be a slow establishing beat in an embedded player has to be the first visible image in a scrolling environment.
6. Add burned-in subtitles wherever silent autoplay is likely, and keep them clear of the lower band where platform controls and captions of the platform itself appear.
7. Check each version against the channel's current specification for length, aspect ratio, file size and frame rate, then export from the same graded and mixed source.
8. Watch every version on the device it is made for before delivery. A vertical cut judged on a desktop monitor looks fine and still fails on the phone it was built for.

Common pitfalls

What most often goes wrong in practice:

• Centre-cropping a wide edit to vertical and cutting the second speaker out of every dialogue
• Leaving titles in the lower band where the platform draws its own interface over them
• Building one version and hoping the platform's own scaling handles the rest
• Judging a vertical version on a desktop monitor and never opening it on a phone

With TillyGen

When the target formats are part of the TillyGen brief, the shot list asks for the framing each one needs — so vertical versions come from shots with room at the sides instead of from a crop that removes half the scene.

Change one constraint and the consequences travel through the whole plan: affected shots are flagged, the call sheet is regenerated, and nobody keeps working from yesterday's version.