Introduction

Deploying an AES67 network for mission-critical audio applications demands far more than simply plugging in equipment and turning it on. Audio-over-IP (AoIP) protocols such as AES67 enable pristine, low-latency streaming across different devices and facilities, but they also introduce dependencies on network infrastructure that can make or break a production. Whether you are building a broadcast studio, a live-sound touring rig, or a fixed-installation performance venue, these ten essential tips will help you plan, implement, and maintain a rock-solid AES67 network.

The following guidance draws on real-world experience and best practices from the Audio Engineering Society, leading AoIP manufacturers, and integrators who have deployed thousands of AES67 channels in demanding environments. Each tip expands on common challenges and provides actionable steps to ensure your network delivers reliable, high-quality audio at all times.

1. Conduct a Thorough Network Assessment

Before you purchase a single cable or switch, take a hard look at your current network infrastructure. Many audio professionals assume that a standard office LAN can handle AES67 traffic without issues, but that is rarely the case. A comprehensive network assessment should identify potential bottlenecks, bandwidth limitations, and single points of failure.

Walk through every link in the chain: switches, routers, cabling, and endpoints. Use network analysis tools to measure baseline utilization and latency. Pay special attention to existing traffic that might compete with audio streams, such as large file transfers or video conferencing. Document the topology, including which devices support IEEE 1588v2 (Precision Time Protocol) and which do not. This assessment will form the blueprint for your AES67 deployment and help you decide whether you need a dedicated network or can share infrastructure with careful segmentation.

For a deeper dive, refer to the Ravenna technical papers from ALC NetworX, which offer detailed guidance on planning AoIP networks that meet AES67 compliance.

2. Ensure Adequate Bandwidth

AES67 audio streams are typically uncompressed, so bandwidth requirements are higher than many people expect. A single 48-kHz, 24-bit, 8-channel stream consumes roughly 12–15 Mbps depending on packet overhead and redundancy. When you multiply that by the number of channels in a large production—often 64, 128, or even more—the total can exceed 1 Gbps if not designed carefully.

Allocate at least double the projected bandwidth to handle peak usage and allow for future expansion. Use 10 Gbps uplinks between core switches whenever possible, even if your endpoints are only 1 Gbps. Also consider that multicast traffic duplicates across switch ports, so the load on each link must be calculated for the worst-case scenario. Don’t forget that other network services (PTP, control protocols, management traffic) also consume bandwidth. Oversubscription is a common cause of audio dropouts and increased jitter.

3. Use Managed Switches with QoS Support

Unmanaged switches have no way to prioritize audio packets over data packets. In a congested network, a large file download or a burst of broadcast traffic can delay AES67 packets, leading to audible glitches or total loss of sync. Managed switches with Quality of Service (QoS) capabilities allow you to tag audio traffic with high-priority DSCP values (typically EF or CS4 for AES67) and shape other traffic to stay within limits.

Configure strict priority queuing for the audio class, and ensure that all switches in the path honor the same QoS settings. Many manufacturers provide pre-built switch profiles for AES67 networks; these can save hours of manual configuration and reduce human error. Also enable IGMP snooping to confine multicast audio streams only to ports that have subscribed receivers, preventing unnecessary bandwidth waste.

4. Implement Proper Network Segmentation

One of the most effective ways to ensure AES67 performance is to isolate audio traffic from general-purpose data traffic. Segmentation can be achieved through VLANs (Virtual Local Area Networks) or by using physically separate switches and cables. A dedicated audio VLAN keeps PTP timing intact and prevents non-audio broadcasts from flooding the audio subnet.

When using VLANs, make sure the inter-VLAN routing is designed with sufficient bandwidth and low latency. Avoid routing audio through a firewall or a router that performs deep packet inspection, as that can introduce unacceptable delays. In mission-critical applications, consider a fully dedicated network: a separate set of switches and cabling for audio only, with no connection to the corporate LAN except for remote control or monitoring via a highly controlled gateway.

5. Synchronize Clocks Accurately

AES67 relies on a shared clock domain to keep all devices sample-accurate. The standard mandates Precision Time Protocol (PTPv2/IEEE 1588-2008) for synchronization. A poorly implemented PTP domain will cause drift, pops, clicks, and eventually loss of audio lock.

Designate a highly stable grandmaster clock—ideally a dedicated PTP grandmaster with GNSS or IEEE 1588-2008 holdover capability. Configure all switches as transparent clocks (TC) or boundary clocks (BC) to correct for switch residence time. Avoid using endpoints as ordinary clocks if they lack high-quality oscillators. Regularly monitor PTP metrics such as offset, mean path delay, and sync interval. Any device reporting an offset greater than 1 microsecond should be investigated immediately.

6. Use Compatible Hardware and Software

Although AES67 is an interoperability standard, not every device claiming AES67 support implements every optional feature in the same way. When mixing equipment from different vendors, verify compatibility through hands-on testing or by consulting manufacturer interoperability lists. For example, some older Dante devices require firmware updates to support AES67, and some Ravenna devices may need specific session descriptions.

Always start with a small test bench that includes each make and model you plan to use. Test at the highest channel count and sample rate expected in production. Pay attention to redundant streams (e.g., ST 2022-7 seamless protection switching) — both the sending and receiving devices must support the same redundancy method for it to work. Use the same major firmware version across all switches and endpoints to reduce hidden incompatibilities.

7. Regularly Update Firmware and Software

AES67 devices and network switches are under continuous development. Manufacturers release updates to fix bugs, improve timing accuracy, and add support for newer PTP profiles or multicast management features. Ignoring firmware updates can leave you vulnerable to known issues that cause intermittent dropouts or even network-wide failures.

Establish a regular maintenance window for updating firmware on all network infrastructure and audio endpoints. Keep a staging environment where you can test new firmware before deploying it to the production network. Document version numbers and release notes for every component, and subscribe to vendor security advisories. A well-maintained network is far less likely to suffer from “mystery” problems that take days to troubleshoot.

8. Test and Validate the Network

Thorough testing before going live is non-negotiable. Simulate your worst-case production traffic with a test generator that sends the maximum number of AES67 streams at the highest bit depth and sample rate. Use network analyzers like ClockTamer, Audio Precision, or open-source tools to measure latency, jitter, packet loss, and drift over a 24-hour period.

Walk through every possible failure scenario: unplug a switch, restart a grandmaster, introduce a multicast storm. Document how the network recovers and whether any audio is dropped. Testing should also include a full load test where all devices simultaneously transmit and receive the maximum expected stream count. If you find jitter values above 1 ms or packet loss greater than 0.01%, investigate and remediate before proceeding. The Lincomatic AES67 testing guide provides practical steps for validating network health.

9. Implement Redundancy and Backup Plans

In critical audio applications, a single point of failure can cause a catastrophic loss of sound. Design your network with redundant paths, dual power supplies, and failover capabilities. Many AES67 systems support SMPTE ST 2022-7 seamless protection switching, which sends two identical streams over physically separate network paths. The receiver automatically switches to the clean stream if one path degrades.

Also consider capacity redundancy: have spare switches, cables, and media converters on hand, pre-configured and labeled for quick swap. If your venue uses a single redundant stream, test that the failover happens within the acceptable audio gap (generally less than 10 ms for live performance). Document every redundant path and practice a change-over drill with the operations team. Redundancy is only valuable if it works when you need it.

10. Provide Training and Documentation

The best-designed AES67 network will fail if the people operating and maintaining it do not understand its nuances. Invest time in training your audio engineers, IT staff, and technical crew on the fundamentals of AoIP, PTP, QoS, and multicast addressing. Hands-on workshops with your specific equipment are far more effective than generic online courses.

Maintain detailed, living documentation that includes network topology diagrams, IP address schemes, VLAN assignments, PTP configuration, QoS policies, and step-by-step troubleshooting procedures. Store this documentation in a shared repository that can be accessed from a mobile device during a crisis. Update it whenever a change is made. A well-trained team with clear documentation can resolve most issues in minutes rather than hours, keeping your critical audio on the air.

Conclusion

Deploying a successful AES67 network in critical audio applications is achievable when you combine careful planning, robust infrastructure, and continuous maintenance. The ten tips covered here—from network assessment and bandwidth allocation to redundancy and training—form a foundation that scales from a single room to a global broadcast backbone. Remember that AES67 is a means to an end: delivering flawless, low-latency audio that listeners and engineers can trust.

For further reading, consult the official AES67 standard document and the RAVENNA White Paper on AoIP Networking. These resources provide deeper technical details that will help you refine your deployment and stay ahead of evolving audio-over-IP standards.