Direct monitoring (also called direct mixing or zero-latency monitoring) is a hardware feature built into most audio interfaces. Instead of routing the microphone signal through the computer and back — which introduces measurable delay — it routes the signal directly from the preamp to the headphone amplifier on the interface itself.
The resulting monitoring latency is effectively zero — the only delay is the speed of electrons through the circuit, which is imperceptible. This allows vocalists and instrumentalists to hear themselves in real time while recording.
Without direct monitoring, the alternative is software monitoring through the DAW. Modern computers can achieve low enough latency (under 10 ms with proper ASIO drivers and small buffer sizes) that software monitoring is usable. However, direct monitoring is simpler and always reliable regardless of CPU load.
Most interfaces with direct monitoring allow blending the direct mic signal with the playback from the computer — so the performer hears both themselves and the backing track simultaneously. The Focusrite Scarlett series has a Mix knob that blends input and playback continuously.