Latency is a measure of time delay experienced in a system. It refers to a short period of delay (usually measured in milliseconds) between when an audio signal enters and when it emerges from a system. Potential contributors to latency in an audio system include (AD) analog-to-digital conversion, buffering, digital signal processing, transmission time, digital-to-analog (DA) conversion and speed of sound.
When monitoring a record enabled track, there may be some amount of latency. Latency occurs with any Audio system that utilizes the computer's CPU for its signal processing. This is the small amount of time it takes the card to carry an input, process the information and get it back out to the outputs. This latency can be reduced to a level that is unnoticeable.
In the configure hardware driver you can change the Samples Per Buffer setting allowing you to reduce the latency to where it isn't even noticeable. Approximately 4ms will be the smallest latency. This makes the processor work harder so the number of tracks you are playing back, amount of effects you are running, and what machine you are using, will determine the minimum amount of samples per buffer you can afford to get away with while monitoring a record enabled track.
Latency in digital audio equipment is most noticeable when a singer's voice is transmitted through their microphone, through digital audio mixing, processing and routing paths, and then sent to their own ears via in ear monitors or headphones. In this case, the singer's vocal sound is conducted to their own ear through the bones of the head and then 1-5 milliseconds later through the digital pathway to their ears. This combination of bone conduction and digital latency is unsettling to some singers. Latency times under 2 ms and over 15-20 ms can reduce the annoyance.