Picture-in-Picture Video Free: Facecam Over Gameplay in Your Browser

You can put your face in the corner of your gameplay in a browser tab, free, with no watermark. VideoCraft is a picture-in-picture video editor almost by accident of how it renders: it draws every video track at the same moment, stacked bottom to top, so a small clip on an upper track sits right on top of a full-screen clip below it. Drop the gameplay on one track, the facecam on the track above, shrink the facecam into a corner, and there's your picture in picture. This post walks the whole recipe, then the blend-mode and opacity tricks for looks beyond a plain box.

The picture-in-picture video editor trick is stacking tracks

Open VideoCraft and you get two video tracks and two audio tracks to start, and you can add more, up to sixteen of each. The rule that matters here is the stacking order. The bottom track, V1, renders first. Every track above it paints over it, V2 over V1, V3 over V2, on up to Vn. So whatever sits on your highest track wins the front of the frame. Put the facecam on V1 and the gameplay on V2 and the gameplay hides the face completely. Flip them, gameplay on V1 and facecam on V2, and the face becomes the layer that can sit on top.

One thing to notice when you drop a video file: VideoCraft lays the picture on a video track and a linked audio clip on a free audio track, tied together with a small chain badge. The preview video is muted, and all the sound plays through that audio clip, which you can slide around on its own. For a facecam that matters, because you usually want the game audio, not the room noise off your webcam. Right-click the clip and choose Unlink to move the two apart, or delete the audio clip entirely if you only want the picture.

The facecam-over-gameplay recipe

1. Make a free account and open VideoCraft. Nothing to install.
2. Drop your gameplay clip onto the first video track, V1. It becomes the full-frame base.
3. Drop the facecam clip onto the track above it, V2. Right now the facecam fills the whole frame and hides the gameplay, which is exactly what you'd expect from a top layer at full size.
4. Click the facecam clip to select it, then find Clip Settings on the right. Click the word Motion (the section title, not the little fold arrow next to it) to arm the transform box on the Player.
5. Drag a corner handle inward to shrink the facecam, or type a number into the Scale field. Then drag the middle of the box to slide the inset into whichever corner you like.

Here are the real numbers, because they help if you want the same corner on several clips. In a 1920 by 1080 sequence the Position field starts at 960, 540, which is dead center, and Scale starts at 100 percent, meaning the clip fills the frame. I took the facecam down to 32 percent and dragged it to the bottom right, and the Position readout landed at 1558, 852. Those are the anchor coordinates of the inset in sequence pixels, so instead of eyeballing the drag on the next clip you can type the same two numbers and the inset lands in exactly the same spot.

The Motion fields are the scrubby kind. Drag one up or down, or roll the scroll wheel over it, and the value changes live; click it once and you can type an exact figure. The box on the Player gives you eight handles for scaling (the corners hold the aspect ratio so a 16:9 facecam stays 16:9), a middle you drag to reposition, curved zones a little outside each corner for rotation, and a small anchor dot you can move. Press Esc once to drop the box, and again to clear the selection. If you ever want to be sure it's genuinely compositing and not faking it, start a free project and scrub the playhead while the facecam is shrunk. The gameplay keeps playing behind the inset the whole way.

Blend modes and opacity, for looks past a plain box

A picture-in-picture inset doesn't have to be an opaque rectangle. Every clip has an Opacity value and a Blend Mode, and both mix against the tracks underneath. Drop the facecam's opacity to around 50 percent and the gameplay reads straight through it, a ghosted double-exposure that suits a slow intro or an outro card.

Blend Mode gives you sixteen ways to mix the top clip into the one below instead of covering it. Set the facecam to Screen and the dark parts drop away while the bright parts stay, so the layer glows over the scene rather than blocking it. Normal is what you want for an actual facecam most of the time, but Screen or the lighten group earn their keep on overlays like a light leak or a logo sitting on black.

If your facecam already sits on a green background, skip the box entirely. Key the green out and the face floats over the gameplay with no rectangle around it at all. That's the green-screen keying workflow, and keyed transparency composites against the track below in the same stacking order, so the face on V2 reveals the gameplay on V1 wherever the green used to be.

What to know before you start

Two honest caveats. VideoCraft has no timeline export to MP4 yet and no project save yet, so build your picture-in-picture edit in one sitting and finish it before you close the tab. You can grab a single composited frame at the full sequence resolution with the camera button in the transport bar (every layer, transform and blend baked in), which is handy for thumbnails, but rendering the whole timeline to a video file is still on the roadmap. The editors themselves are free for everyone. Paid plans sell storage and AI compute, not the editing tools, so stacking tracks and shrinking a facecam never costs a token or a tool use.

Want the inset to do something other than sit still? Animate its Position and Scale with keyframes and the facecam can slide in from off-screen or pop bigger on a beat. And if this is your first time in the editor, the rest of the VideoCraft tourcovers the timeline, effects and the pop-out panels that let you drag the Player onto a second monitor. Open it, put a clip on V1 and another on V2, and shrink the top one. That's picture in picture, and it took about a minute.