How to Reframe Horizontal Video for Vertical Shorts

Decide what must stay visible

A wide recording has room for details that a vertical crop can remove. Before moving the frame, decide which details explain the moment. In an interview, that might be the person speaking and the object they hold. In a stream, it might be an opponent near the edge of the gameplay and the creator's reaction.

Pause at the key moment and name what the viewer needs to see. Then play the whole passage. A crop that works on one still image may lose the subject when they move or hide the result of an action a few seconds later.

Choose the composition for the footage

A close vertical crop can work when one subject stays in a small area. When important details span the full image, preserving the wide frame may be clearer, even if the subject appears smaller. A blurred background can fill the remaining canvas, but it does not make small source details easier to read.

For gameplay with a facecam, a stacked layout can keep the reaction and game visible together. Check the balance at phone size. Giving the facecam more space also gives the gameplay less space, which can make distant opponents, menus, or subtitles harder to follow.

Treat automatic framing as a draft

intoshort can propose a gameplay and facecam composition for supported vertical processing. Detection can be uncertain, and an inset picture is not necessarily a human webcam. When no sufficiently stable region is available, or the framing service fails, the result can use a centered crop.

Watch for missing faces, overlays mistaken for the facecam, and action outside the selected region. An exported file can be technically valid while showing the wrong part of the scene. If the suggested composition hides why the moment matters, revise it before sharing.

Adjust the frame scene by scene in intoshort

Open Reframing in the short's editor. Manual framing controls require multiple existing scenes. If scene detection is missing, use Detect scenes where offered and review the proposed boundaries before editing. Detection itself does not replace the rendered export.

Select the scene and adjust the available source rectangles. The central preview shows the draft composition. Review the beginning and end of that scene, then check the transition into the next one. A scene crop should not be assumed to track every movement inside the scene.

Apply layout to all scenes can reuse the active composition, but review the other scenes before saving. A facecam can move, disappear, or be replaced by a different screen. Reusing one layout is convenient only when it still fits the footage.

Check captions and the finished export together

After changing a layout, review the caption position again. Keep the text clear of the subject and any game information the audience needs. Account for the controls that the destination platform places over the video.

Save the short, wait for the new render, and watch the finished file on a phone. Check the action, the reaction, and every scene transition. Keep a composition because it makes the moment understandable, rather than because it fills every part of the vertical canvas.