A display and a sensor in one place
A touchscreen is really two devices sharing one surface. One is the display, which produces the picture. The other is a sensor, which reports where something is touching or hovering near the surface. The sensor does not draw anything. It simply sends coordinates to the device, and software decides what those coordinates mean: a tap on a button, the start of a scroll, or a line in a drawing.
Because the two parts are separate, the same panel types can be paired with different sensors. Understanding the screen itself is covered in the guide to panel types, while this article is about the sensing layer on top of or around it.
The basic job: turning contact into coordinates
Every touch technology has to do three things. It has to notice that something touched or approached the surface, work out where, and report it quickly enough to feel natural. Different technologies solve this in different physical ways, and each choice brings trade-offs in cost, durability, size, accuracy and what kind of object can be detected.
The main families are resistive, capacitive, infrared, optical (camera-based) and a few less common acoustic or electromagnetic approaches.
Resistive touch
A resistive sensor is a sandwich. Two thin, flexible conductive layers are separated by a tiny gap. When you press, the top layer bends and touches the bottom one, and the electronics measure the electrical values at the contact point to estimate its position.
Because it responds to pressure rather than to the electrical properties of what is pressing, a resistive screen can work with a gloved finger, a fingernail or a plastic stick. The trade-offs are that the extra layers can reduce clarity, the flexible surface can wear, and detecting several fingers at once is difficult. This method was common on older devices and is still found where low cost and glove use matter.
Capacitive touch
Capacitive sensing detects the small electrical disturbance a conductive object, such as a finger, causes in an electric field. A grid of tiny electrodes under the glass is scanned repeatedly. When a finger comes near, the stored charge at nearby crossing points changes, and the controller calculates the touch position from those changes.
This is the method behind most smartphones and tablets. It handles light touches, supports many simultaneous contacts and works under a hard glass surface. Its weaknesses are that ordinary gloves and non-conductive objects may not register, and water or electrical noise can confuse it. A detailed comparison with the resistive approach is in the article on capacitive touch.
Infrared and optical touch
Large screens often use a different idea. An infrared frame places light emitters and receivers around the edge of the screen, creating an invisible grid of beams just above the surface. A finger or object interrupts some beams, and the intersection of the blocked beams gives the position. Camera-based optical systems instead use sensors, usually in the corners, that watch the surface and calculate position by triangulation.
These methods scale to big displays at reasonable cost and work with almost any object, including gloves and pens. They do not need a special layer on the glass, though they can be sensitive to strong ambient light, dust or objects resting against the frame. Details are in the article on infrared touch.
Other approaches
Some systems use sound waves that travel across the glass and are absorbed where a finger touches. Others use an electromagnetic digitizer layer behind the display that tracks a special pen very precisely. Pen systems are their own topic, covered under stylus input, because tracking a pen tip differs a good deal from tracking a fingertip.
What happens after the sensor reads a touch
A touch controller is a small chip that scans the sensor, filters out noise, decides which readings are real contacts and assigns each one an identity so it can be followed as it moves. The device’s operating system then turns the stream of positions into events. When several contact points are tracked together, the software can recognize pinches, swipes and rotations, which is the subject of multi-touch gestures.
Along the way there are several places where delay can creep in: scanning the sensor, processing the signals, running the software and finally redrawing the screen. This is why touch latency is a design concern, especially for drawing.
Choosing between methods
No single method fits every situation. Rough guidelines include:
- Phones and tablets typically favor capacitive sensing for its light touch and multi-touch support.
- Rugged or low-cost equipment sometimes favors resistive sensing, particularly where gloves are common.
- Large public and classroom screens often favor infrared or optical frames, which scale up well.
- Devices intended for drawing often add a dedicated pen layer.
Where a screen will be used matters as much as the sensor itself. A machine in a bright, dusty public area faces different challenges from a phone in a pocket. Self-service kiosks, for example, have to cope with constant use by many people, and outdoor machines add sunlight and rain to the list.
What touch cannot do
Touch input has limits that are worth remembering. Fingers are large compared with small interface elements, so precision is lower than with a mouse or pen. Surfaces get smudged, and screen cleaning matters more than people expect on shared machines. Touch also provides little physical feedback on its own, which is the reason designers add vibration cues, discussed under haptic feedback.