AES67 vs. Ravenna: Choosing the Right Audio‑over‑IP Protocol for Your Network

Professional audio networking has moved decisively from analog snakes and point‑to‑point digital connections to packet‑switched Ethernet. Two protocols dominate the conversation: AES67, the open interoperability standard, and Ravenna, a high‑performance solution built by ALC NetworX. Both deliver reliable, low‑latency audio over IP, but they differ in architecture, flexibility, and intended use. Understanding these differences is essential when designing a system that must scale, integrate with existing gear, and meet the latency demands of live sound, broadcast, or installed sound environments.

This article provides a deep technical comparison of AES67 and Ravenna, covering their transport mechanisms, synchronization methods, latency profiles, ecosystem support, and practical deployment scenarios. By the end, you will have a clear framework for deciding which protocol – or combination – best suits your professional audio network.

What Is AES67?

AES67 is a standard published by the Audio Engineering Society in 2013. It is not a complete networked audio system but a set of interoperability requirements that allow different audio‑over‑IP (AoIP) systems to communicate with one another. The standard specifies layer‑3 transport using RTP (Real‑time Transport Protocol) over IP networks, synchronization via IEEE 1588‑2008 (Precision Time Protocol, PTP), and a common media clock of 48 kHz or 96 kHz.

Because AES67 is an open standard, any manufacturer can implement it without licensing fees. This has led to broad adoption across the industry: Dante, Ravenna, Q‑LAN, Livewire+, and other protocols all offer AES67 compatibility modes. In practice, AES67 acts as a “universal translator” – a channel format that diverse AoIP devices can share, even if their native protocols differ.

Core Technical Specifications of AES67

  • Transport: RTP payloads over unicast or multicast UDP. The standard defines a specific RTP profile for compact audio data (PCM, typically 16 or 24 bits per sample).
  • Synchronization: IEEE 1588‑2008 PTPv2. The standard requires grandmaster clock support, with frequency accuracy of ±0.002% and phase alignment within 1 ms.
  • Sample Rates: 48 kHz is mandatory; 96 kHz is optional. Bit depths of 16 or 24 are supported.
  • Latency: The standard allows packet times from 0.125 ms (4 samples at 48 kHz) up to 4 ms. Typical implementations use 1 ms packet times, yielding end‑to‑end latency of 1‑2 ms.
  • Redundancy: No mandatory redundancy scheme. Implementations may use multicast redundancy or separate networks, but AES67 does not define a standard failover mechanism.

What Is Ravenna?

Ravenna is a proprietary technology developed by ALC NetworX (a subsidiary of the German audio company X42). It was designed from the ground up for demanding professional applications: live sound reinforcement, broadcast consoles, and large‑scale installed sound. Ravenna is a complete networked audio solution that includes transport, synchronization, device discovery, and connection management.

Unlike AES67, which is a minimum interoperability profile, Ravenna is a full‑featured protocol stack. It uses RTP for audio transport and IEEE 1588 for synchronization, but adds proprietary extensions for lower latency, higher channel density, and advanced redundancy. Ravenna devices can operate in “native” Ravenna mode for maximum performance, or in AES67 mode for cross‑vendor interoperability.

Core Technical Specifications of Ravenna

  • Transport: RTP with proprietary payload format optimized for low latency. Supports both unicast and multicast.
  • Synchronization: IEEE 1588‑2008 with enhanced clock recovery algorithms. Offers submicrosecond jitter performance.
  • Sample Rates: 44.1 kHz, 48 kHz, 96 kHz, 192 kHz (DXD rates in some variants). Bit depth up to 32 bits.
  • Latency: Native Ravenna can achieve sub‑millisecond latencies (0.125 ms packet time possible). Typical end‑to‑end latency in live systems is under 500 µs.
  • Redundancy: Built‑in support for redundant streams (SMPTE 2022‑7 style seamless protection), redundant network interfaces, and multicast stream replication.

Key Differences Between AES67 and Ravenna

While both protocols share the same underlying Ethernet and IP infrastructure, they diverge in five critical areas: interoperability, latency, channel density, redundancy, and ecosystem.

Interoperability

AES67’s primary raison d’être is interoperability. Any device that implements an AES67 stream can send audio to any other AES67‑compliant device, regardless of the vendor. This has made AES67 a bridge between formerly incompatible ecosystems – for example, a Dante‑based Yamaha console can send an AES67 stream to a Ravenna‑based Lawo mixing system.

Ravenna, being proprietary, does not guarantee interoperability with non‑Ravenna devices in its native mode. However, most Ravenna devices include an AES67 mode, so they can participate in mixed networks. The trade‑off is that when operating in AES67 mode, Ravenna devices revert to the lower performance profile of AES67 (e.g., minimum sample rate 48 kHz, maximum bit depth 24 bits, and higher minimum latency).

Latency and Jitter

For live sound reinforcement, latency must be kept below the threshold of human perception – typically under 1–2 ms for monitor mixing or large‑scale arrays. Ravenna excels here: its native mode allows packet times as short as 0.125 ms (4 samples at 48 kHz), and its proprietary clock recovery can maintain phase alignment within nanoseconds. This makes Ravenna a strong choice for systems where cue wedge monitoring or IEMs are used.

AES67, by contrast, often operates at 1 ms packet times. While 1 ms is generally acceptable for front‑of‑house (FOH) and studio applications, it can become problematic in complex routed networks where multiple hops add latency. A properly designed AES67 network can still achieve 1‑2 ms end‑to‑end latency, but it cannot match Ravenna’s sub‑millisecond performance.

Channel Density and Sample Rates

Both protocols support high channel counts over 1 GbE. AES67’s profile caps sample rates at 96 kHz, which is sufficient for most professional work. Ravenna can operate at 192 kHz and 32‑bit depth, useful for high‑resolution recording, archiving, or scientific audio applications. For typical live or broadcast workflows, this difference matters little – but for mastering studios or research facilities, Ravenna’s higher ceiling is valuable.

Redundancy and Fault Tolerance

Redundancy is a major differentiator. AES67 does not specify a redundancy scheme; manufacturers must implement their own (e.g., Dante offers redundant networks via separate switches, or ST 2022‑7 for seamless protection). Ravenna includes a robust redundant streaming mechanism: you can send two identical multicast streams across separate networks, and the receiver seamlessly switches to the healthy stream if one fails. This is critical for broadcast or mission‑critical live events where audio dropouts are unacceptable.

Ecosystem and Vendor Support

AES67’s strength is its ubiquity. Every major AoIP vendor supports it: Audinate (Dante), ALC NetworX (Ravenna), Bosch (OMNEO), Wheatstone (Livewire+), and many others. This makes AES67 the safest choice for a multi‑vendor environment.

Ravenna has a focused but powerful ecosystem. Key partners include Lawo, DiGiCo, Neumann, K&F, and Merging Technologies. The Ravenna ecosystem is particularly strong in broadcast (e.g., Lawo consoles and Neumann monitoring) and high‑end recording (Merging Pyramix). For projects that already use Ravenna gear, staying native simplifies configuration and maximizes performance.

Use‑Case Analysis: When to Choose AES67

Multi‑vendor interoperability projects: If your network combines gear from different manufacturers – for example, a Yamaha console (Dante), a Wheatstone router (Livewire+), and a Lawo stagebox (Ravenna) – then AES67 is the universal glue. You can configure each device to output/input an AES67 stream and avoid vendor lock‑in.

Cost‑sensitive installations: AES67 hardware tends to be more affordable because the standard is open and licensed royalty‑free. For education, houses of worship, or small studios, AES67 provides professional AoIP without premium pricing.

Systems with moderate latency requirements: For recording, post‑production, or FOH mixing where 1–2 ms latency is acceptable, AES67 performs perfectly.

Use‑Case Analysis: When to Choose Ravenna

Low‑latency live sound reinforcement: Large‑scale tours, festival stages, or theater productions where monitoring latency must stay under 1 ms benefit from Ravenna’s native performance. The ability to use 0.125 ms packet times in a mixed IEM/FOH setup is a real advantage.

Broadcast environments requiring redundancy: TV, radio, and OB van deployments demand seamless failover. Ravenna’s built‑in redundant streaming (SMPTE 2022‑7 compliant) means no single switch failure will stop the audio. AES67 systems must rely on external redundancy schemes that may not be as tightly integrated.

High‑sample‑rate recording: Studios capturing at 96 kHz or 192 kHz with 32‑bit float can benefit from Ravenna’s native support. While AES67 can do 96 kHz, Ravenna’s lower jitter and higher bit depth maintain signal integrity over long distances.

Can You Use Both? Hybrid Networks

Absolutely. Many professional installations operate hybrid networks where Ravenna devices communicate with one another at native performance while also providing AES67 streams to non‑Ravenna devices. For example, a Lawo console (native Ravenna) can send an AES67 stream to a Dante‑compatible loudspeaker processor. The console still uses Ravenna for its stageboxes and monitor mixes, achieving low latency where it matters, while the output to the loudspeaker processor uses AES67 for compatibility.

To build a hybrid network, you need to plan the clock domain carefully. A single PTP grandmaster can serve both AES67 and Ravenna devices, provided all endpoints support IEEE 1588‑2008. The network switches must also be PTP‑aware and configured for proper boundary clock or transparent clock operation.

Technical Considerations for Deployment

Network Infrastructure

Both protocols run over standard IP networks, but they impose requirements on network switches. You need:

  • IGMP snooping for multicast traffic – both AES67 and Ravenna rely on multicast for efficient stream distribution.
  • PTP support – switches should be PTP‑aware (boundary clock or transparent clock) to avoid jitter accumulation. Unmanaged switches are not recommended.
  • Sufficient bandwidth – a single 1 GbE link can carry hundreds of channels, but oversubscription must be avoided. Plan for at least 600 Mb/s on the uplink to a core switch if carrying 256 channels at 48 kHz/24‑bit.

Configuration Complexity

AES67 can be simpler to configure because the standard intentionally limits options to ensure interoperability. However, the lack of automatic device discovery means you must manually set IP addresses, multicast groups, and stream parameters (e.g., packet time, sample rate). Many vendors provide software tools to simplify this – for example, Audinate’s Dante Controller can configure AES67 streams on Dante devices.

Ravenna includes a discovery and connection management protocol called RAVENNA Device Discovery (similar to mDNS). Devices advertise their capabilities, and you can route streams from a graphical interface. This makes setup faster when using only Ravenna gear, but when mixing protocols, you may need to use vendor‑specific tools.

Latency Budget Planning

Whether you choose AES67 or Ravenna, estimate your total latency budget:

  1. Packet time (the time over which samples are collected). 1 ms is common for AES67; 0.125 ms is possible for Ravenna.
  2. Transmission delay – negligible on a local network (<10 µs hop).
  3. Switch queuing delay – switches can add 10–50 µs per hop under light load.
  4. Receiver buffer – the destination device must buffer enough samples to absorb network jitter. This is typically 2‑4 packet times.

For AES67 with 1 ms packets, you should budget 3–4 ms end‑to‑end. For Ravenna with 0.125 ms packets, total latency can be under 1 ms – but only if the entire network chain is optimized and all devices operate in native mode.

Future of Both Protocols

The AoIP landscape continues to evolve. AES67 is being succeeded by SMPTE ST 2110 (for broadcast) and AES70 (for control), but AES67 remains the core transport layer for many systems. Ravenna is actively developed by ALC NetworX, with recent additions including support for AMWA NMOS (for IP flow management) and enhanced audio‑to‑video alignment.

Additionally, the rise of Milan – a deterministic profile of AVB (Audio Video Bridging) – offers another alternative, especially in environments that want guaranteed latency without the complexity of PTP. However, Milan requires special switches and endpoints, while AES67 and Ravenna work on standard commoditized hardware.

Given the industry momentum, both AES67 and Ravenna will remain relevant for the next decade. The choice often comes down to whether you prioritize interoperability (AES67) or peak performance in a single‑vendor environment (Ravenna).

Decision Matrix: AES67 vs. Ravenna

Criteria AES67 Ravenna
Interoperability Excellent – open standard, cross‑vendor Good – via AES67 mode, native limited to Ravenna ecosystem
Latency (typical) 1–4 ms 0.125–2 ms (native sub‑ms)
Redundancy Not defined (vendor‑specific) Built‑in redundant streams (ST 2022‑7)
Max sample rate 96 kHz (48 kHz mandatory) 192 kHz, up to 32‑bit
Cost Lower (no licensing) Higher (proprietary licensing)
Ecosystem size Very large (all major vendors) Medium (focused on broadcast/live)
Ease of setup (single‑vendor) Manual IP config (but vendor tools help) Automatic discovery, user‑friendly

Conclusion

AES67 and Ravenna are not adversaries; they are complementary tools in the audio engineer’s kit. AES67 provides the universal language for cross‑vendor communication, while Ravenna delivers premium performance in low‑latency, high‑reliability scenarios. For most professional networks, a hybrid approach yields the best of both worlds: use Ravenna natively for the high‑bandwidth, time‑sensitive paths (stageboxes to console, console to IEM transmitters) and deploy AES67 for outputs to third‑party devices or for integration with existing infrastructure.

Before making a final decision, evaluate your latency budget, network redundancy requirements, and the specific hardware you plan to use. Consult with your equipment vendors to confirm the AES67 implementation level (e.g., mandatory vs. optional features) and test Ravenna’s native performance in your specific switching environment. With careful planning, either protocol – or both – will deliver the reliable, high‑fidelity audio transport that professional applications demand.

For further reading, consult the official AES67 standard document and the Ravenna technology overview. Additional insight into AoIP network design can be found in the Dante learning hub’s AES67 guide and the SMPTE ST 2110 overview for broadcast convergence.