How to Convert Horizontal Videos to Vertical

What the crop actually is

Converting a horizontal video to vertical is usually described as reframing, which makes it sound like a composition decision. It is not. It is a subtraction, and the amount is much larger than people expect.

Take a 16:9 frame and cut a 9:16 window out of it at full height. The window is nine sixteenths as wide as it is tall, sitting in a frame sixteen ninths as wide as it is tall, so what you keep is nine sixteenths squared: 81/256, or 31.6% of the picture. Just over two thirds of every frame is deleted.

The trap is thinking there is a way out of that number. There is not – there is only a choice about what it applies to. Refuse to crop, drop the whole 16:9 clip into a vertical frame, and the same fraction reappears on the other side: the picture now fills 31.6% of the screen and the rest is empty. Delete two thirds of the image or waste two thirds of the display. Every technique below is a way of answering that honestly rather than a way of escaping it.

Four ways to spend the frame

A 9:16 frame is taller than a cropped 16:9 shot needs to be. That leftover height is the actual design problem, and there are four things worth doing with it.

One subject, tracked. The crop follows the person, full bleed, top to bottom. This is right for a single speaker and it is the case automatic reframing handles well, because there is exactly one thing to follow.

Two crops, stacked. Two separate 16:9 crops, one above the other, each about half the height. This is the honest answer to an interview, and it is already the native look of gaming clips, where gameplay sits above a facecam. Both subjects stay full size and nobody is cut in half.

A crop plus a title zone. Put the video in the middle and give the top third to a line of text: the hook, the question being answered, the name of the guest. This is the option people forget, and it is the one that turns the tall frame from a cost into the reason the clip works. It pairs directly with hooks.

Fit on a blurred fill. The whole horizontal clip, centred, with a blown-up blurred copy of itself behind. It has a place – below – but it is the weakest of the four and it is the one used most.

Try it: drag the window

The abstraction becomes obvious the moment you move a real crop across a real shot. Here is an ordinary two-person setup with a whiteboard between them – the kind of thing every recorded conversation looks like. Drag the orange window, or use the arrow keys, and watch the list underneath.

Two things show up immediately. The first is that no position holds both people. That is not a failure of the tool or of your taste; the two faces are further apart than the window is wide, and the only fixes are the ones that change the shape of the output – stack two crops, or cut between them.

The second is that a centre crop, which is what every tool does by default and what most people accept, is frequently the worst available choice. On this shot it lands on the whiteboard and clips both speakers. A default is not a decision, and on a conversation it is usually the wrong one.

Crop the shot, not the clip

Here is the habit that separates a converted clip from a clip that looks like it was made vertical. Almost everyone sets one crop for the whole clip. The crop belongs to the shot.

A forty-second clip might contain a wide two-shot, a push in on one speaker, and a moment where they both look at a screen. Those are three different framings with three different centres of interest, and one fixed window cannot serve them. Set the crop at each cut, the way you would set exposure.

This has a consequence worth being explicit about, because it surprises people: converting to vertical changes the edit itself. A continuous wide take where two people talk over each other works horizontally and cannot work vertically as one shot. Vertically it becomes cuts – to whoever is speaking, back again on the reply. You are not reframing the edit, you are re-editing it, and the vertical version will usually have more cuts than the original had.

Two practical rules follow. Move the crop on a cut, never during a shot, unless you are tracking a subject who is genuinely moving – a window that slides sideways while the picture is still reads as a camera wobble. And keep the subject in the same part of the frame across cuts, so the viewer’s eye does not have to relocate every two seconds.

The blurred background problem

The blurred fill deserves its own section because it is the default in most tools and it is chosen far more often than it is right.

What it does: shrinks the whole 16:9 clip to fit the width, then fills the empty third above and below with an enlarged, blurred copy of the same frame. Nothing is deleted, which is the appeal. But the picture is now small, the text in it is smaller, faces are distant, and roughly a third of the screen is showing a smeared version of what the middle of the screen is already showing.

It earns its place in one situation: footage that is genuinely wide by design and would be destroyed by a crop. Landscapes where the width is the subject, as on travel clips. Slides and screen shares, where a crop makes the content unreadable – webinar clips covers why those often should not go vertical at all.

Everywhere else, the same space can carry something. A title. The question being answered. A caption block that is not covering anyone’s face. A second camera angle. The test is simple: if the top and bottom of your frame are not telling the viewer anything, you have given away a third of the screen for nothing.

Why the crop goes soft

A converted clip that looks slightly mushy usually has nothing wrong with the export settings. It is arithmetic that happened before the export.

A 1920×1080 source cropped to 9:16 at full height leaves you 608 pixels across. That is the entire horizontal resolution you now have. If the clip is then delivered at any width greater than 608, every one of those pixels is being stretched, and no bitrate or sharpening pass puts back detail that was never captured.

The fix is upstream. A 3840×2160 source cropped the same way leaves 1,215 pixels across, which is comfortably more than vertical delivery normally needs, so the crop is a downscale rather than an upscale – and downscaling actually looks better than not scaling at all. If you know a recording will be clipped vertically, record it at 4K even if the horizontal version never needs it.

Check what your destination actually asks for in your own account rather than trusting a figure in an article, since those change. The arithmetic going in does not: fixing blurry video clips covers the rest of the chain, but this step is the one that does most of the damage.

Shoot so the crop is free

Every difficulty above is created at the moment of recording, and a handful of habits make the conversion almost automatic. None of them costs anything on the day:

• Leave headroom and side roomFrame a little wider than you want the horizontal version to be. A centred subject with air around them survives any crop; a tight, beautifully composed 16:9 frame survives none.
• Put the subject on the centre lineRule-of-thirds framing is the single most expensive habit if the clip is going vertical. Off-centre looks better wide and gets cut in half tall.
• Shoot a second angle, tighterOne wide for the horizontal edit, one closer for the crop. Two cameras is the cheap version of a reframe you never have to do.
• Keep the action in the middle thirdWhatever the clip is really about – hands, a screen, a face – should live in the 31.6% of the width you are going to keep.
• Record 4K if you will cropA 1080p source leaves 608 pixels across after a vertical crop. A 4K source leaves 1,215, which is room to spare.
• Do not put anything at the edgesLower thirds, logos, scores and tickers at the frame edge are gone. If it matters, move it inward before you record.

The first two matter most. A subject placed on the centre line with room around them can be cropped to any shape by anyone in a few seconds; a subject framed hard to one side by the rule of thirds cannot be, and the beautiful horizontal composition is exactly what makes the vertical version impossible. If the clip is going into a feed, compose for the crop and let the horizontal version be the slightly loose one.

FAQ

Related: captions for video clips, long videos to TikTok and what video clipping is.