How to film an ASMR video with both iPhone cameras

Which camera should carry the trigger and which should carry your face

Most of an ASMR video is the trigger — tapping, scratching, brushing, whatever the object of the day is — not your face doing the tapping. Picture-in-picture with the trigger camera filling the frame and your face in a small window plays to that: the object stays central and easy to watch, and your face is there to be seen without competing with it for space.

For the moments you actually want to talk to the camera — introducing the trigger, a whispered aside, a thank-you at the end — DualCam lets you swap which camera fills the frame at any point during the recording, live, the same way a cooking or tutorial video cuts between the food and the face. Decide the emphasis in the moment rather than locking in one layout and living with it for the whole take.

One audio track, and why this will not give you true binaural sound

DualCam writes a single mono AAC track at 44.1 kHz alongside the composited video — one track for the entire file, no matter which camera or layout is on screen. There is no per-ear channel and no way to place a sound left or right afterwards. That is the real difference between an ordinary phone microphone and the specialised binaural mics ASMR channels are often built around, and no amount of picture-in-picture or split-screen framing changes what the microphone itself is doing.

The source is whatever iOS has set as the default audio input, so a wired or Bluetooth microphone plugged in and selected by the system gets recorded instead of the phone’s own mic — the app does not override that choice or offer its own picker. That lets you get a mono source right next to the trigger instead of a phone propped a foot away, which closes most of the gap that actually matters to someone listening on headphones, even without two independent channels.

If binaural stereo is genuinely the point of the video, the two-camera picture is still useful as a visual guide: record the binaural audio separately with a dedicated recorder, then layer that track over the original in a normal video editor afterwards. A dual-camera recording is an ordinary finished MP4 once you are done shooting, and adding a new audio track on top of the one already in the file is exactly the kind of edit a normal editor already does.

Getting the taps you want without also recording the stand

General advice for dual-camera audio says not to touch the surface the phone is resting on, because a knock travels straight through that surface into the microphone at the bottom edge. ASMR looks like an exception, since tapping and scratching sounds are the entire point — but the advice is really about the phone’s own contact point, not about triggers in general.

Keep the phone’s stand or mount mechanically separate from whatever you are tapping or scratching: a tripod that does not touch the same table the trigger object sits on, or a stand on a different surface entirely. Sound reaching the microphone through the air, at close range, is what you want. Sound reaching it by traveling through the phone’s own body from something it is physically touching just adds a dull thud underneath everything else, including the trigger sounds you were trying to capture.

Framing for a trigger that moves

If you prefer a split layout to picture-in-picture, the two ways to split are not equal here. A top-and-bottom split keeps the full width of both frames, which matters when a trigger object moves side to side or you reach off to the edge of the shot for the next one. A left-and-right split crops roughly a quarter off each side of both source frames to fit a vertical canvas, and that missing edge is exactly where a hand reaching sideways for the next object goes.

Picture-in-picture avoids the question entirely, and is the safer default for the same reason it is in a cooking video: the main frame keeps its full width, and only the small inset — your face — is cropped to fit its window.

Lighting for a low, warm room

ASMR is usually filmed dim and warm on purpose, and that works against a phone camera before it works against anything else: a small lens and a small sensor already struggle in low light, and a dual-camera session adds a constraint on top by sharing one hardware budget across both cameras — which is why it tops out at 1080p and cannot use an automatic “best quality” preset the way a single-camera shot could. In a dim room, that ceiling is the first thing that runs out of real detail to work with.

The two cameras also meter exposure independently, each reading only its own frame, so a warm lamp on the trigger table and whatever light happens to be on your face rarely agree. Tap the frame that needs help — tap-to-focus carries exposure with it — rather than expecting one adjustment to fix both sides.

What actually helps in a dim room

• A lamp facing whichever camera is struggling helps more than any setting. DualCam’s torch only lights whatever the back lens is pointed at, so it does nothing for a dark front-camera face.
• Do not zoom past a lens’s optical range in a dim scene — the digital crop enlarges noise along with whatever detail is left.
• If one side of the shot is noticeably dimmer than the other, put it in the small picture-in-picture window rather than the full frame. A noisy image reads cleaner scaled down.
• Mono or Film, two of the app’s six live filters, suit the low-contrast look a lot of ASMR already goes for, and quietly remove the color mismatch between two independently-exposed cameras along with it. Both apply live, in the shader, while you shoot — not something you add afterwards.

Long sessions run into heat and storage before the trigger list does

A full trigger set easily runs twenty or thirty minutes, well past what most other videos on this site need in one take, and a dual-camera recording is a continuous file with no way to pause it mid-take — any break, planned or not, means stopping and starting a new take. That actually suits ASMR better than most subjects: a trigger list is already segmented by object, so filming one take per trigger and stopping between them is the natural shape anyway, not a workaround.

Two live camera pipelines and a GPU composite running for that long draw on the same thermal budget the whole time. If the phone gets hot enough, DualCam drops a fresh 1080p recording to 720p before it starts, or warns you if you are already recording at that resolution; past that point it ends the recording and finalizes the file rather than pushing through. Starting a long session at 720p is the safer plan, not a downgrade forced on you partway through.

Storage works the same way: at roughly 10 Mbps, a recording uses about 75 MB a minute, so a gigabyte of free space is around thirteen minutes. Clear space before a long session rather than finding out mid-take, on a trigger you cannot easily reshoot in the same mood.