A radio quoted at twenty milliseconds can sit inside a system that takes four hundred. The figure an operator experiences is glass to glass — photons at the sensor to photons at the display — and the radio is rarely the largest term in it. Quoting the radio alone is the standard way this specification goes wrong.
The image has to come off the sensor. A rolling shutter reads rows sequentially, adding both delay and skew on a moving platform.
Compression is where the largest and least-discussed delay sits. B-frames reference future frames, so the encoder must hold frames before emitting them — excellent for bitrate, poor for latency. Low-latency profiles give them up deliberately.
Framing, error correction and the radio itself. This is the number vendors quote, and it is frequently the smallest term in the chain.
Effectively free at these ranges. Radio covers a kilometre in about three microseconds; the distance is not the problem.
The receiver holds a jitter buffer so playback stays smooth when packets arrive unevenly. Every millisecond of buffer is a millisecond of latency, and it is a deliberate trade rather than an accident.
Decoder pipeline plus whatever the display adds. A consumer monitor can contribute tens of milliseconds that nobody counted.
Lower latency means less buffering, which means jitter shows as stutter. Higher quality means more bitrate, which means less link margin and earlier failure at range. More robust modulation means better range and less throughput for the same spectrum.
There is no configuration that wins all three. What there is, is a correct answer for a given operation. An operator flying a survey grid wants picture quality and does not care about a quarter second. Someone landing a platform manually, or directing a response from a live feed, needs latency and will accept a soft picture.
So the useful question is not what latency the link achieves, but what the operator has to do with the picture. That question decides the encoder profile, the buffer depth and the adaptation policy, and it should be asked before any hardware is chosen.
The total delay from photons hitting the camera sensor to photons leaving the display the operator is watching. It includes sensor readout, encoding, transmission, buffering, decoding and display. A figure quoted for the radio alone can be a small fraction of it.
Because efficient compression references future frames. B-frames mean the encoder holds frames before emitting them, which is excellent for bitrate and poor for latency. Low-latency encoder profiles give up that efficiency deliberately, and accept a higher bitrate for the same picture.
Not at the same link margin. Lower latency means less buffering and less tolerance of jitter; higher quality means more bitrate and less margin at range. The right balance depends on what the operator is doing with the picture — flying a survey grid and landing manually want opposite settings.
Because they share an antenna and a power budget but not a requirement. Command carries a few kilobits and must never drop; video carries megabits and can degrade. Command gets the link margin and the robust modulation, because losing it means losing the aircraft.
Because the airframe banks and the antenna pattern moves with it. Nulls in the pattern point at the ground station during manoeuvres, so dropouts correlate with flight dynamics rather than with distance — which is why bench range testing does not predict flight behaviour.
Written for engineers. Share it with one.
The hardware.
TECHNOLOGYCoding gain on a power-limited link.
TECHNOLOGYLink budget and antenna.
PRODUCTShared SDR lineage.
INDUSTRYApproval and lifecycle.
SOLUTIONThe commercial offering.
Tell us the specification, the constraint and the deadline. Programmes that cross silicon, radio, embedded and AI are where Faststream is strongest.