Latency is important in any real-time simulation, especially in simulated flight. Low or no latency is the engineering objective. In real-time flight simulation, long latency causes difficulties for the trainee; queasiness induced by mismatched visual cues, and pilot induced oscillations from over-steering due to sluggish response in the flying controls. An example to help your imagination is the sinking feeling you get when the brakes of the car you are driving don't answer immediately when required.

In sequential digital real-time simulation there is a finite lag between the moment when the control inputs change and the moment when the simulation model produces an output in response to the change. You can think of it as the reaction time between input and output; it is the same as reaction time found in humans, the time between recognition in the offing of your mother-in-law and when you dodge round the corner to make your escape.

It has been said that analogue simulators have no latency when in fact they do, though it is very small in size. Latency in analogue simulators is caused by friction and stiction in the various electromechanical systems. Even so, it can be adjusted downwards by tuning, until it is insignificant to the training experience.

No amount of tuning will eliminate latency from a sequential digital real-time simulation. Advent of new technology has enabled engineers to make steady reductions in latency. For many years the desirable objective was to demonstrate a real-time simulation latency of 100 milliseconds or less, the measured time between a step-input to the elevator channel of the flying controls and corresponding movement of the visual display scene in the red-channel. Most recently, latency of real-time flight simulation has been reduced to around 70 milliseconds in the primary flight training devices currently being deployed to the United States Navy and Air Force.

Latency is the reaction time of a real-time simulation; the quicker the better.