Can you make the picture-in-picture window bigger in a dual camera video?

What sizing up the window actually does

The composited canvas — 1080 pixels wide, 1920 or 2340 tall depending on which aspect you picked — does not change size when you pick a bigger picture-in-picture window. What changes is how much of that fixed canvas the inset is allowed to claim, at the expense of the main frame in that corner.

The camera feeding the inset, meanwhile, is captured at 720p or 1080p — a setting fixed before recording starts and locked once it does. That source frame gets scaled down to fit whatever box the inset currently occupies. A small window throws away a lot of that captured detail in the scaling; a larger window throws away less of it, because there is less distance between the source size and the display size. That is the real reason a bigger inset looks sharper — not because it suddenly has more resolution to draw from, but because it wastes less of what it was already given.

This also means the size step interacts with the resolution setting you picked before you started. At 1080p there is real detail to preserve, so sizing the inset up is worth doing. At 720p the source itself is coarser, and stretching it across a bigger box on screen will not manufacture detail that was never captured — it just makes the softness more visible.

A bigger window also covers more of the main shot

The inset sits in one of four corners with a small margin, and whatever is under it in the main frame is hidden the moment the window is drawn. At the smallest size step this barely registers. At the largest, it is a real chunk of the frame — check what is actually in that corner of the main camera's shot before you commit to it, especially if the main frame is doing something other than sitting there as background.

Shape changes the calculation too. DualCam's inset comes in three shapes: a tall rounded rectangle that keeps the same proportions as the viewfinder and therefore crops the source the least, and a square and a circle, which are true 1:1 crops of the same frame and lose more of it. Two insets at the same size step are not showing the same amount of picture if one is tall and the other is a circle — the tall one is both keeping more detail and covering less of the main shot underneath, for roughly the same footprint.

When it is worth sizing up

• The inset is a face reacting to something, and the reaction is the point — go bigger. A face read from a tiny window loses exactly the thing you are recording it for.
• The inset is context nobody is meant to study closely — a location, a glance, a corner of the room — leave it at the smaller or middle step so it does not compete with the main shot.
• The take is at 1080p and the moment matters — size up, since the detail is actually there to show.
• The take is at 720p, or the phone has been recording for a while and may have stepped down — a larger window will not fix softness that was already set by the capture setting.

None of this has to be decided before you press record

Layout, the inset's corner, its size and its shape can all be changed while a DualCam recording is already running, and the file follows the change immediately, because the preview and the recording are the same rendered frame. The only settings locked the moment you start are resolution, mute and the countdown timer — size is not one of them.

That is worth remembering because it removes the pressure to guess correctly in advance. Start at whatever size looks reasonable in the viewfinder, and if a reaction gets bigger than expected or the small window turns out to matter more than you planned, resize it live instead of trying to predict the whole take before it happens.