Latency is the time delay between an audio signal entering a system and emerging from it. In live sound, latency matters because performers need to hear themselves clearly to stay in time. If the delay between a singer's voice reaching their ears naturally and the amplified sound from their monitor speakers is too large, they'll sing out of sync with the track or other musicians. Understanding latency sources and acceptable limits helps you design and operate systems that sound natural.
Sources of System Latency
Every component in a digital audio signal chain adds latency. Analog-to-digital converters (ADCs) introduce typically 0.25-1.5ms of latency depending on the conversion technology and anti-aliasing filter design. Digital signal processing (EQ, dynamics, routing) adds variable latency depending on algorithm design — some DSP processes signal in small blocks, each block adding samples of delay.
Digital-to-analog converters (DACs) add similar latency to ADCs. Network audio transmission (Dante, AVB, AES67) adds variable latency depending on network congestion, clock synchronization, and buffer settings — typically 0.5-5ms for professionally designed networks. Wireless microphone systems add 1.5-4ms of codec and transmission latency depending on the system.
The cumulative latency of a signal chain determines total system delay. A typical digital mixing console might have 0.5-2ms total latency from input to output. Adding wireless mic systems (up to 4ms), network audio distribution (1-3ms), and signal processing plugins (variable, 0.1-10ms) can push total latency to levels where performers notice the delay.
Buffer Sizes and Sample Rates
Buffer size in audio interfaces and digital consoles determines how many audio samples are processed in each batch. Smaller buffers reduce latency but increase CPU processing load; larger buffers reduce CPU load but increase latency. At 48kHz sample rate, a 128-sample buffer equals approximately 2.67ms of latency; a 256-sample buffer equals 5.33ms; a 512-sample buffer equals 10.67ms.
Higher sample rates reduce latency for the same buffer size because more samples per second means each sample represents less time. At 96kHz, a 128-sample buffer is only 1.33ms. However, higher sample rates increase CPU load and network bandwidth requirements. Professional live systems often run at 48kHz with appropriately sized buffers to balance latency against processing capacity.
Fixed versus adaptive buffer systems behave differently. Most audio interfaces use fixed buffer sizes that only change when manually adjusted. Some systems (primarily consumer devices and software DAWs) use adaptive buffers that automatically adjust based on CPU load — this can cause audible glitches during CPU spikes, which is unacceptable in live environments where stability trumps minimal latency.
Acceptable Latency Limits
Research into human perception of audio-visual sync suggests that most people notice AV sync errors around 40-80ms for video with audio, but audio-only latency detection thresholds are much lower. Musicians typically notice acoustic feedback delays above approximately 5-10ms — beyond this, they begin to adjust their performance to compensate, which can degrade timing.
For stage monitor applications, under 5ms total latency from performer to monitor speaker is generally acceptable for most musicians. Singers with in-ear monitors can typically tolerate slightly higher latency because the IEM system itself adds some delay and performers adapt to it more easily. Drummers often require very low latency because they rely heavily on hearing their kit to maintain time.
For front-of-house systems, latency is less critical because the audience is receiving processed sound, not using it for timing feedback. Total system latencies up to 15-20ms are generally acceptable for FOH, though lower is always better for natural sound reproduction. The exception is when FOH sound reinforces the stage directly (drums bleeding into FOH, for example) — in these cases, lower latency reduces phase issues between direct acoustic sound and amplified sound.
Measuring and Managing Latency
Most professional audio equipment and software includes latency measurement capabilities. Dante network software shows latency between devices. Digital consoles display input-to-output latency. Audio measurement tools can measure round-trip latency by playing a test tone through the system and measuring the return time. Use these tools to verify your system is performing within acceptable parameters.
When connecting multiple digital systems (FOH console to monitor console, main PA to delay speakers), delays must be matched so that sounds arriving from different sources arrive at the audience position simultaneously. This is called time-alignment. Use the Audio Delay Calculator to estimate delay times based on distance, and use measurement tools to fine-tune alignment precisely.
Delay speakers for large venues add their own latency — typically 1-5ms for the signal processing, network transport, and conversion chain. When adding delay speakers at the back of a venue, account for this additional latency plus the propagation time of sound through air. The goal is that sound from the delay speaker arrives at the same time as sound from the main speakers.
Use our Audio Latency Calculator to estimate system latency and determine acceptable buffer settings for your application.