How to put your webcam on a screen recording

You have two files: a screen capture of the thing you are demonstrating, and a webcam take of you talking over it. Putting one on top of the other is five minutes of work. Making it not look like a template is a handful of decisions about size and placement, and those are where most tutorials go wrong.

Sync them first

Everything downstream is easier once the two files are locked together.

Drop the screen recording on the lower video lane and the webcam on a lane above it. If you clapped, snapped or clicked audibly at the start of the take, the two waveforms have a matching spike — line those up and you are synced to the frame. If you did not, find a moment where you say a hard consonant and something visibly happens on screen, and align there.

Do it before you trim anything. Trimming first means re-syncing after every cut, which is the same work repeated.

Next time: start both recordings, then clap once in front of the camera. It costs one second and saves this entire step.

Float the webcam

Lane order is draw order: whatever sits on a higher lane draws over the lanes beneath. Select the webcam clip, open Transform → Layout and choose Make overlay (PiP). It shrinks to a floating box that keeps its true aspect ratio — a portrait webcam take stays portrait rather than being stretched into a rectangle.

Then open Edit layout to get drag handles on the preview itself: move it, resize from a corner, or snap it to a corner preset. Size and position sliders sit in the same editor when you want exact numbers rather than a drag.

For the round webcam bubble, apply a circle or ellipse mask with a small edge feather. Feathered circle masks are the whole look, and the feather is what stops it reading as a sticker.

Size and placement, which is the actual craft

Three rules cover most of it.

The webcam is not the content. Fifteen to twenty-five percent of the frame width is plenty. If someone cannot read the UI you are demonstrating because your face is in front of it, the video has failed at its one job. The exception is the intro and the outro, where you are the content.

Put it where nothing happens. Watch the screen recording once and notice where the action lives — usually the left sidebar and the centre. Park the overlay in the quietest corner and leave it there. Moving it around mid-demo is distracting unless the movement is motivated by the thing being covered.

Respect the platform's furniture. On vertical platforms the right-hand third carries the interface, and the bottom fifth carries the caption and the account name. A bubble in the bottom-right of a vertical cut is behind a like button. Bottom-left or top-left survives everywhere.

If you want the overlay to enter rather than just exist, keyframes on position, scale and opacity turn a static PiP into a fly-in. Use it once, at the start. A bubble that animates every time it appears is a tic.

Better than a bubble, if you can

If you shot against a green screen, key it out and drop yourself directly onto the screen recording — no box, no border, just you standing beside the interface. Combine an overlay with chroma key and you have a floating cut-out subject. The background remover does the same job on footage shot without a screen, at the cost of a generation.

This reads far more expensive than a bubble and takes about the same time. It is also the single most recognisable "produced" cue in a tutorial.

Fix the audio, because that is what people leave over

The screen capture has audio you do not want — system sounds, a second copy of your voice through the laptop mic, keyboard noise. Mute that lane entirely and keep the webcam's audio, which was recorded closer to you.

Then, in order:

• Remove silences on the talking track — free, and it tightens the demo more than any other single step.
• Remove filler words afterwards, because that scan is billed on the clip's length and the clip is now shorter.
• Duck the music under the voice if there is a bed at all: flag the music clip as Duck and it dips under speech automatically.
• Check it on phone speakers before you call it done.

Make the screen legible

The most common problem with screen recordings is not the webcam, it is that a 1440p capture of a dense interface is unreadable on a phone.

Two fixes, both cheap. A Ken Burns move slowly zooms into the region that matters, so the button you are clicking fills a meaningful part of the frame. The spotlight dims everything except the area you are pointing at, which is the "look here" tool for a busy interface — size, feather and darkness are adjustable.

Use one or the other on any shot where you say "up here in the corner". If you are saying it, the viewer cannot see it.

Caption last

Captions are generated against the audio as it exists, so every cut you make afterwards puts them out of step. Lock the edit, then caption.

Keep the lines short — two or three words at a time — and check that they are not landing under the webcam bubble. That collision is the reason to keep the overlay out of the bottom of frame.

The order that works

1. Sync on the audio spike before trimming anything.
2. Make the webcam an overlay, and mask it if you want the bubble.
3. Size it at 15–25% and park it in the quietest corner, clear of platform furniture.
4. Mute the capture's audio, keep the webcam's.
5. Remove silences, then filler words — in that order, because the second is billed by length.
6. Zoom or spotlight anything you refer to verbally.
7. Caption last, and check nothing lands under the bubble.