How to film a prank video with both iPhone cameras
A prank video runs on one thing a scripted video does not have to worry about: the reaction you are actually trying to capture belongs to someone else, and you do not get to decide when it happens. Recording the front and back camera at once still gives you the same core benefit it gives a solo reaction video — the setup and the reaction land in the same file, at the same second, with nothing to line up afterwards — but a target who does not know a shot is coming changes almost every practical decision around that benefit.
The general mechanics of simultaneous recording — what locks once you press record, how layout and audio actually work — are covered elsewhere on this site. This page is about what changes when the person on the other side of the shot is not you: which camera should lead, why you want to be rolling well before you think the moment is coming, why the one microphone in a dual-camera recording is not actually pointed at the person you are trying to catch, and where to put the small window so it does not end up over the reaction itself.
Decide whose reaction is the point, before you decide who is where
A dual-camera prank usually has two candidates for the main frame: the target, or your own face watching them. Both are legitimate choices, and they are not the same video. Putting the target’s reaction in the main frame and your own face in the small window reads as a video about what happened to them; the reverse — your face large, the target small — reads as a video about how you reacted to setting it up, which is a different joke.
Because which camera fills the main frame can be changed while the camera is rolling, and the file follows that swap immediately, you do not have to fully commit to either read before you start. Roll with whichever read makes sense for the setup, and swap to the other the moment the prank actually lands — the same technique that works for swapping to a face at the reveal of a package works here, except a prank’s reveal is rarely as scheduled as an unboxing, which is the reason to treat the swap as something you are ready for rather than something you plan to the second.
There is no pause, so roll before you think you need to
A dual-camera recording runs as one continuous pass from record to stop, with no button that freezes it mid-take. That is a real constraint for any kind of shoot, but it costs the most in exactly this genre: a scripted moment like an unboxing has a reveal you can count into, while a prank’s trigger is decided by someone who does not know they are on camera, which means it can happen four seconds earlier or a full minute later than you expected.
The practical answer is to start recording well before the moment you think is coming, and accept the dead air at the front of the file as the cost of not missing the one take that matters. A prank only truly happens once — a version staged a second time for the camera is a different video, not a rescued version of the real one. If the setup itself takes a while to build, that is what stop and start between takes is for: get the physical prank ready first, and only start rolling once you are actually waiting for it to go off.
One microphone, and it is probably closer to you than to them
Dual-camera recording writes a single audio track from whichever input iOS has selected, not a separate feed per camera and not a directional pickup aimed at whichever camera you have decided is the main frame. Distance to that one microphone is what decides how clearly a sound registers, and the microphone sits on the phone in your hand — which usually puts you, the person holding it, closer to it than the person you are recording a few feet away.
That is the reason a prank video so often has the filmer’s own laugh dominating the track over the target’s actual reaction, even when the target’s face fills the main frame. If their reaction is the sound you actually want, closing the physical distance between the phone and them matters more than which camera is pointed at them — hand the phone off, or close the gap the moment the reaction starts, rather than assuming the framing alone will carry the audio the way it carries the picture.
The two feeds expose independently, and you do not get to fix that later
The front and back cameras each set their own exposure from what they can individually see, which matters more in a prank than in most dual-camera shots because you and your target are rarely standing in the same light — you might be waiting in a dim hallway while they are lit by a bright kitchen, or the other way around. Tap the frame that matters more, usually the one with the target’s face, so that region drives its own exposure instead of whatever the lens landed on by default.
Decide which side to prioritize before the reaction happens, not after, because a dual-camera recording writes one flattened frame per moment with no separate layer left behind to re-expose afterward. A dark reaction shot in an otherwise fine take cannot be brightened back to what a properly exposed one would have looked like — it can only be edited the way any finished video can, which does not include recovering detail that was never captured.
Keep the small window out of the way a person actually moves
A box on a table stays where you put it, which is why picture-in-picture works well for an unboxing. A person reacting to a prank does not — they can flinch sideways, step back, or throw their hands up, and none of that is scripted enough to frame around in advance the way a scheduled reveal is. A picture-in-picture inset parked in a corner is more likely to end up sitting over exactly the part of the frame someone jumps into.
A top-and-bottom split is the more forgiving choice here for the same reason it works well for framing that needs the full width: it keeps the entire horizontal range of both cameras in the recording, so a sideways flinch stays inside the frame instead of running into an inset that was never built to move with the shot. Picture-in-picture still works fine for a prank staged in a tight space where a target genuinely cannot move far — the wider the space they can react into, the more a split earns its keep over an inset.
Common follow-up questions
Should the target’s reaction or my own face be the main frame?
Either works, and DualCam lets you swap which camera fills the main frame while recording, with the file following immediately. Putting the target’s reaction in the main frame and your face in the small window tells a video about what happened to them; the reverse tells one about your own reaction to pulling it off. Decide going in, but do not treat it as locked — you can swap the moment the prank lands.
What if I do not know exactly when the prank is going to happen?
Start recording earlier than you think you need to. A dual-camera recording cannot be paused and resumed inside one file, so there is no way to catch a moment that happened before you pressed record. Rolling early and accepting some unused footage at the start costs far less than missing the one reaction you were actually trying to catch.
Why does my own laughing drown out the target’s reaction in the audio?
Dual-camera recording writes one audio track from the phone’s microphone, not a feed tied to whichever camera is pointed at whom. The microphone sits on the phone in your hand, which usually makes it closer to you than to the person a few feet away, regardless of which camera’s frame is the main one. Closing the physical distance to the target matters more than which camera faces them.
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.