The PROVANTAGE Handbook / Audio networks
Networked audio can carry many channels over shared physical links and make routing flexible. It also concentrates decisions that analogue patching makes visible.
A useful design maps audio signal flow and network topology together: devices, switches, subscriptions, clocking, bandwidth, failure domains and the person authorised to change them.
Four connected layers
EVERY LAYER
MUST
AGREE.
Physical
Switches, ports, copper or fibre links, link speed, cable paths, power and device interfaces.
Network
Addressing, topology, VLAN decisions, QoS where required, multicast handling and ownership boundaries.
Audio transport
Sample rate, latency, channel capacity, clocking, subscriptions and device compatibility.
Operation
Names, saved states, access control, monitoring, change process and recovery responsibility.
Before audio passes
NAME IT.
CLOCK IT.
SUBSCRIBE IT.
Device identity
Use stable, meaningful names and record the physical location and owner of each endpoint.
Addressing
Confirm how addresses are assigned and avoid uncontrolled changes or duplicate addressing.
Clocking
Establish a suitable clock leader and verify synchronisation, sample rate and device status.
Subscriptions
Patch intended transmit and receive channels, then save and document the working state.
Latency
Choose settings the network and device path can sustain with margin; lower is not automatically safer.
Monitoring
Observe links, errors, clock state, subscriptions and audio at meaningful points—not software colour alone.
The available-port mistake
AN ETHERNET PORT
IS NOT A
PRODUCTION NETWORK.
Before connecting to a venue or corporate network, establish ownership, switch capability, configuration, security, traffic, addressing and change control. Where those conditions cannot be guaranteed, use an appropriately designed production network.
Redundancy needs independent paths
SECONDARY
MUST MEAN
SEPARATE.
Map failure domains
Identify what one cable, switch, power source or configuration error can remove.
Respect device capability
Not every endpoint supports the same redundancy mode or recovery behaviour.
Separate infrastructure
Where redundancy is required, evaluate switches, links, power and routes—not just duplicate ports.
Keep fallback useful
Some critical signals may justify a separate analogue or digital route outside the network.
Test recovery
Remove representative links safely and confirm expected continuity and alarms before show day.
Control changes
Freeze the working configuration and route necessary edits through one network owner.
The practical takeaway
FLEXIBLE ROUTING.
VISIBLE
RESPONSIBILITY.
Networked audio is dependable when topology, configuration, physical infrastructure, monitoring and recovery are explicit. Flexibility without documentation becomes live-production uncertainty.
Review audio connections →