AUD-04

Audio Data Rate & Wavelength

SMPTE ST 2110-30 / AES67 · IEC 60268 · ANSI/AVIXA A104.01 · ANSI/AVIXA A102.01

01 · Audio format

Channels per stream

Sample rate

Bit depth

Packet timeⓘ

02 · Network

Streams per device

Devices on the link

Link speed

03 · Acoustic wavelength

Frequency

Air temperatureⓘ

°C

Awaiting input9 fields left: Channels per stream · Sample rate · Bit depth · Packet time · Streams per device · Devices on the link · Link speed · Frequency · Air temperature

Capacity and layout

The drawing appears when the inputs are complete

01
Work out PCM, Dante and AES67 stream bandwidth including packet overhead and link utilisation, and convert frequency to wavelength for loudspeaker spacing and boundary effects.
02
Tip. Packet time trades latency against overhead. Shorter packets cut latency but multiply the per-packet header count, so a 125 µs stream carries roughly eight times the overhead of a 1 ms one for exactly the same audio.
03
48 kHz — Standard sample rate
04
1 ms — AES67 default packet time
05
1500 B — Standard Ethernet payload
06
343 m/s — Speed of sound at 20 °C

Audio data rates and acoustics

Two unrelated questions, both answered from the same handful of constants. Networked audio bandwidth follows directly from sample rate, bit depth and channel count, with protocol overhead on top. Acoustic wavelength follows from frequency and temperature, and governs loudspeaker spacing and boundary effects.

Network bandwidth

Payload is fixed arithmetic; what varies is the packet rate and therefore the header overhead. Shorter packet times cut latency and multiply overhead.

Use cases

48 kHz / 24-bit is the professional default
1 ms packet time is the AES67 baseline
125 µs for low latency, at eight times the overhead
Keep links at or below about 70 % utilisation

Formula

Payload (b/s) = sample rate × bit depth × channels

AES67 · SMPTE ST 2110-30

Wavelength and spacing

Wavelength sets the distances at which sources reinforce or cancel. It is what makes boundary placement and loudspeaker spacing acoustic decisions rather than aesthetic ones.

Use cases

Quarter wavelength — deepest boundary cancellation
Half wavelength — worst comb filtering between sources
Speed of sound rises about 0.6 m/s per °C
A room at 30 °C has wavelengths ~2 % longer than at 20 °C

Formula

λ = (331.3 + 0.606 × T) ÷ frequency

343 m/s at 20 °C

A payload above 1500 bytes will not fit a standard Ethernet frame. Jumbo frames must then be enabled on every switch and endpoint in the path — one device left at the default drops the packets, and the failure looks like an audio fault rather than a configuration one.