Can you move the picture-in-picture window while you are recording?
Yes. The picture-in-picture inset can be dragged to a different corner, resized between the three available sizes, or given a different shape while a recording is already running, and none of that requires stopping the take. Layout itself — switching between picture-in-picture and split screen — is just as free to change. The reason this works without any special handling is that the preview on your screen and the frame being written to the file are the same rendered image, so whatever you set the window to becomes what shows up in the recording from that point forward.
It is easy to treat the corner as something you have to get right before you press record, the same way you would frame a tripod shot before rolling. That instinct is not wrong exactly, but it undersells the control: the window is not a one-time setup step, it is a live one, available for as long as the take is running.
What actually happens when you drag it
Touch the small window and drag it toward a different corner, and it moves to sit in that corner — one of four positions the frame is built around. Size works the same way: pick one of the three available sizes and the inset resizes accordingly, and shape switches cleanly between portrait, square and circle. None of these are separate from the recording in progress; there is no preview mode you drop into and no confirmation step. Whatever the window looks like at any given moment is what the camera is compositing into that frame right then.
That immediacy is also why there is nothing to wait for. A change you make settles into the next frame the way any other change in the shot would — the light shifting, the camera panning, someone stepping into frame. Moving the inset is just one more thing that can differ from one frame to the next, handled the same way everything else in a live composite is.
When dragging it mid-take actually earns its keep
The most common reason to move it while rolling is that the shot changed after you started, not before. In a two-person interview, if the person the main camera favors switches partway through, sliding the inset to sit closer to whoever is now the focus can read better than leaving it where it was set up for the first half. In a walking vlog, panning the back camera onto something new can put the inset directly over the new subject even though it was sitting in an empty corner a minute earlier — dragging it to the corner that is actually empty now is a live fix, not something you have to plan for in advance.
The same applies whenever someone unexpectedly moves into the corner the window occupies. Rather than losing the rest of the take to a face hidden behind the inset, dragging the window to whichever corner is currently clear keeps the shot usable from that point on. It will not fix what already happened, but it stops the same problem from continuing for the rest of the recording.
What moving it does not fix
Dragging the window changes where it sits starting on the next frame — it does not reach backward and change frames that have already been recorded. If the inset spent the first twenty seconds of a take covering someone’s face, moving it at second twenty-one clears the shot from that moment on, but the first twenty seconds are unchanged, because the composite is written into the file permanently as each frame happens, not layered on top of something that can be edited apart later.
That makes moving the window a way to stop a problem from getting worse, not a way to undo one. If a take is being ruined from the very first frame, the more reliable fix is to stop, reposition, and start a fresh take rather than trying to drag your way out of it mid-shot.
What stays locked no matter when you try
Corner, size, shape, full layout and the colour filter are all free to change for as long as a take is running. Resolution, whether the recording has an audio track, and the countdown timer are not — those are fixed the instant you press record and cannot be touched again until the file is finished. The difference is not arbitrary: one group only decides what gets drawn into a frame that is already a fixed size, and the other group decides the structure of the file itself, which is not something a video file can renegotiate partway through being written.
In practice this means the picture-in-picture window deserves less advance planning than it might seem to, and resolution deserves more. Settle resolution, mute and the countdown before you start, since there is no second chance at any of them — then treat the window’s corner, size and shape as something you are free to keep adjusting for as long as the camera is rolling.
Common follow-up questions
Does moving the picture-in-picture window mid-recording cause a skip or a dropped frame?
No. The preview and the recorded frame are the same rendered output, so moving the window is handled the same way any other change in the shot is — redrawn fresh on the next frame, with no pause, black flash or gap in the file.
Can I resize it or change its shape while recording, or only move it between corners?
All three — corner, size and shape — can be changed while a take is running, along with the overall layout and the colour filter. The only settings locked once you press record are resolution, mute and the countdown timer.
If the window covered someone’s face earlier in the take, does moving it later uncover that part of the video?
No. Each frame is composited and written permanently as it happens, so moving the window only changes what gets covered from that moment forward. The part of the take where the face was already covered stays that way — the fix has to happen at the time, not afterwards.
Related guides
Want to just do this?
DualCam records the iPhone front and back cameras at the same time and writes one finished MP4 while you shoot. Free, no account, no ads, nothing leaves the phone.