How to turn a horizontal video into a vertical one
Last updated 2026-10-20
To convert a horizontal video to vertical, set the canvas to 9:16, then choose how the 16:9 clip sits inside it. Fill crops the sides and keeps about a third of the width. Fit shows the whole picture with empty bands above and below. If the subject moves, reframe at each cut or keyframe the position. If the shot is wide or has two people, reshoot vertical.
What a 16:9 clip becomes on a 9:16 canvas
Fill enlarges a 1920 by 1080 clip to about 178% so its height matches the 1920 pixel canvas. Only the middle 608 source pixels of its width stay on screen, roughly a third, and they are stretched, so 1080p footage looks softer. Fit does the opposite. The whole clip shrinks to the canvas width, 608 pixels tall, and covers less than a third of the canvas height.
If the subject and the space they move through fit inside a third of the width, fill. If not, use the table.
| Footage | Best approach | What you give up |
|---|---|---|
| One speaker, centered, little movement | Fill, and set the crop so the face is centered | The outer two thirds of the frame |
| Subject walks or gestures across the shot | Reframe at each cut, or keyframe a wider picture | Extra setup, and no tracking |
| Two people in one wide shot | Give each person half the canvas, or cut between them | Each person is cropped tight |
| Wide scene with text near the edges | Fit, with a blurred copy behind it | The picture shrinks to under a third of the height |
| Action across the whole width | Reshoot vertical | Time |
Keep a moving subject in frame
A crop is fixed, so a speaker who walks left slides out of it. The reliable fix is a cut. Split the clip where the subject moves to a new spot and give each piece its own crop. In Woven the crop holds still for a whole clip, so splitting is how you move it.
For a continuous follow, you need a picture bigger than the canvas that you can slide. Scale the whole picture up so it sticks out past the canvas, then keyframe its position from one side to the other. Add a key when the subject changes direction, not every second. Woven has no automatic tracking or auto-reframe. Its AI agent sees footage as still frames, so any crop it picks is a judgment call. Check the result frame by frame.
For two speakers filmed together, give each copy of the clip a half-height band of the canvas, 1080 by 960, and crop each to one person. Mute one copy so the voice plays once. A 16:9 copy scaled to 960 pixels tall keeps about 63% of its width, enough to hold one person.
Fit with a blurred background
Plain bars look unfinished, so many creators put an enlarged, blurred copy of the clip behind the sharp one. Use the bars on purpose: captions and titles can sit in them without covering the face.
Woven has no one-click version. You build it from two copies: the sharp one set to Fit, and a muted second one set to Fill with Blur raised on lane 0, the background lane. Lower lanes draw first, so the blurred copy sits behind and captions land on top.
When to reshoot vertical instead
Reframing rescues footage that already exists. It cannot fix a shot composed for a wide frame. Reshoot when:
- Two or more people share the frame. A 9:16 window holds one face comfortably.
- The action crosses the full width. You would need a key on nearly every second.
- Text sits near the edges. Fill cuts off screen recordings and titles at the sides.
- The subject is small in a 1080p frame. Fill already enlarges 1080p to about 178%, and a small subject needs more. A 4K source needs no enlargement for 1080 by 1920.
Reframe to 9:16 in Woven
Pick the vertical canvas before the first clip goes in, because a blank reel copies the first video's dimensions. Fill and Fit are set per clip, and each cut can have its own crop.
Ask in chat
- Split this clip wherever the speaker moves and give each piece its own crop.
- Put this wide shot on a 9:16 canvas, fit, with a blurred copy behind it.
- Give each of the two people in this wide shot their own half of the vertical frame.
- Render a frame at the 10 second mark so I can check the crop.
Or do it by hand
- 1Go vertical before placing anything. In the toolbar's Format menu pick Vertical 1080x1920, then drag your 16:9 file in from Finder.
- 2Choose Fill or Fit. Select the clip, open the Crop tab and choose Fill or Fit under Framing. To pick which strip shows, double-click the clip on the canvas and drag it.
- 3Reframe at each cut. Press ⌘B where the subject moves to a new spot, then give each piece its own crop.
- 4Follow a walk. Raise Scale under Dimensions until the picture overhangs the canvas. On the Animation tab press ◆ Add, move the playhead to where the subject turns, press ◆ Add again, then click that key and change Position X.
- 5Blurred bars. Ask in chat for a muted Fill copy with Blur raised on a lane behind the original, which stays on Fit.
Common questions
Set a 9:16 canvas, then fill or fit the clip inside it. Fill crops the sides and Fit adds bars. In Woven, Format sets the canvas and the Crop tab sets Fill or Fit.
Only with bars. Fit keeps the whole picture, so a 1920 by 1080 clip becomes 1080 by 608 on a 1080 by 1920 canvas, with empty space above and below.
Split the clip where they move and crop each piece, or keyframe a picture larger than the canvas. Woven has no automatic tracking, so check each result.
Filling does on 1080p footage, because it enlarges the picture to about 178%. A 4K source has enough pixels to fill 1080 by 1920 without enlargement.
YouTube says a Short has a square or vertical aspect ratio and runs up to three minutes.
When two people share the frame, the action crosses the full width, or text sits at the edges.
Sources
- Understand three-minute YouTube Shorts - YouTube Help, checked October 19, 2026
Try Woven free for 3 days
Then $25/mo billed yearly ($299/yr) or $39/month. Native Mac app for short-form video.
Paste into Claude Code to install and connect Woven.