The Screen Has No Idea What's Touching It. Until It Does.

You're mid-sketch, hand resting flat on the glass, and the app ignores your palm completely while the stylus tip draws a clean line 0.4 millimetres wide. No smearing. No accidental zoom. The screen just... knows. There's no groove where the pen lives, no physical boundary separating pen zone from hand zone. The glass is just glass.

So how does it know?

The answer lives in the physics of what each object actually does to the screen's electric field. And the answer is less "clever software" than most people assume.

Two Completely Different Signals Walking Into the Same Bar

Most consumer touchscreens use projected capacitive technology: an invisible grid of electrodes behind the glass, constantly pulsing with a tiny electric field. Your finger disturbs that field because human skin conducts electricity reasonably well. The controller chip measures where the disturbance happened, triangulates across electrode intersections, and reports a coordinate.

A bare finger is a blunt instrument electrically. It creates a diffuse capacitive blob, roughly 8 to 10 millimetres across. Characteristic spread, characteristic magnitude.

A cheap rubber-nub passive stylus just mimics that blob. It pretends to be a fat finger, which is exactly why those styli are so imprecise and not worth your money.

Active styli do something entirely different. They contain a battery and a transmitter, and they broadcast their own signal into the screen rather than simply disturbing the screen's field. The Apple Pencil, for instance, transmits a high-frequency signal through its tip that the iPad's digitizer reads as a distinct, separate channel from the ambient capacitive sensing. The screen isn't detecting "something touched here." It's receiving a labelled broadcast: stylus tip, here, at this pressure level, at this tilt angle.

That's why the tip can be 0.4 millimetres wide and still register precisely. A finger could never achieve that resolution, because a finger doesn't transmit. It only disturbs.

The Other Technology: EMR and the Ghost in the Tablet

Wacom-based devices, including the S Pen in many Samsung tablets and a wide range of professional drawing tablets, use a completely separate technology called electromagnetic resonance (EMR). This one sounds improbable the first time you hear it.

The stylus contains no battery. The tablet's digitizer layer emits a pulsed electromagnetic field. Inside the pen, a small coil and a resonant circuit absorb that energy, then re-emit it at a resonant frequency. The tablet detects the re-emission and uses it to pinpoint the pen's location, pressure, and tilt.

Here's the critical detail. The EMR digitizer operates on a completely separate hardware layer from the capacitive touchscreen sitting above it. Two different grids, stacked behind the glass, tuned to entirely different signals. The EMR layer listens only for the resonant frequency of the pen's coil. Your finger emits nothing at that frequency. Nothing at all.

The EMR layer cannot see your hand. Not because software filters it out, but because your hand is invisible to that layer by design. It's like trying to pick up a radio broadcast on a frequency your receiver wasn't built for.

This is why a Samsung Note user can rest their entire palm on the screen and the S Pen still registers cleanly. The two layers aren't competing. They're not even speaking the same language.

What "Palm Rejection" Actually Is (And Isn't)

Most people picture palm rejection as software that recognises the shape of a hand and ignores it. Sometimes that's part of the picture. But the more elegant version is hardware-level separation: the stylus and the finger were never competing inputs in the first place.

With active Bluetooth styli like the Apple Pencil, the device knows the Pencil is connected and active. When the tip's transmitted signal appears, the system enters a mode where large capacitive contact areas near the tip get classified as palm rather than intentional touch. Software does do some shape-analysis here, comparing blob size and position relative to the pen tip's reported location. But it's working with clean, labelled data. Not guessing.

Consider two people who bought the same drawing app on the same tablet. One uses a passive rubber stylus, the other an active Apple Pencil. The first person has to toggle a palm rejection mode that blocks all touch input while drawing, which means no pinch-to-zoom mid-stroke, no two-finger gestures, nothing. The second person can pinch to zoom with two fingers while the Pencil is in their other hand, and the system handles both simultaneously without confusion.

That difference is entirely downstream of hardware signal separation. The software is just tidying up after the physics.

What People Assume That Isn't Quite Right

The common belief is that screens detect pressure to separate a stylus from a finger: the idea being that a pen tip presses harder and more precisely. Pressure is a real variable that active styli report (up to 4,096 levels in Wacom's professional pens), but pressure alone isn't how the separation happens.

A firm fingernail pressed into a screen creates a small, high-pressure contact point, yet screens don't mistake it for a stylus. A lightly hovering Apple Pencil, on the other hand, registers before it even touches the glass, because the transmitted signal is detectable at a short distance. Separation happens at the signal-type level, not the force level.

And here's something worth knowing before you spend money: not all screens support active styli. A standard capacitive touchscreen on a budget phone has no secondary digitizer layer and no protocol for receiving a stylus broadcast. No stylus, however it's marketed, will give you real precision on that hardware. The spec sheet matters more than the packaging.

The Practical Upshot

If you care about precise stylus input, the thing to check is whether the device has a dedicated digitizer layer for that stylus. Not whether the screen is "touch-sensitive." Every modern touchscreen is touch-sensitive. That tells you almost nothing useful.

Look for EMR digitizer support on the spec sheet, or check whether the manufacturer's own active stylus connects over Bluetooth with a proprietary protocol. Those two paths are the two clean solutions to the separation problem, and they both work by ensuring that the stylus and the finger are never, at the hardware level, the same kind of signal.

The glass really is just glass. The intelligence is in what's listening underneath it, and what it was built to hear.