home/ atoms/ jitter-vs-latency

Jitter is variation in latency over time, not latency itself

Latency is the time a single packet takes to travel from sender to receiver. Jitter is distinct: it measures how much that latency varies across successive packets. If Alice sends packets every 100ms and there is no jitter, Bob always receives them exactly 100ms apart regardless of the absolute latency. Jitter collapses this regularity — a packet may arrive 80ms after the previous one, or 110ms. This distinction matters because latency alone can be compensated by a constant playback offset, whereas jitter requires buffering to absorb the unpredictable timing variation. Bandwidth describes how much data can travel; latency describes how long one packet takes; jitter describes whether that time is consistent. All three are independently observable on a network link. Confusing jitter with latency leads to misdiagnosis: adding more bandwidth or reducing base latency does not eliminate jitter-induced audio glitches.

Examples

A VoIP call with 200ms latency but zero jitter sounds fine (predictably delayed). A call with 20ms latency but 30ms jitter stutters because packets arrive out of the expected rhythm.

Assessment

Given a scenario where packets are sent every 50ms but arrive at intervals of 30ms, 70ms, 45ms, 65ms — calculate the jitter (range or variance) and distinguish it from the average latency. Then explain which symptom each would produce in a real-time audio stream.

“Jitter on the other hand, is a measure of how much that latency _changes_ when many packets are sent over time.”
corpus · sounding-smooth-with-jitter-buffers-jacques-heunis · chunk 1