What is RAVENNA?

An open architecture built on standard IP protocols

RAVENNA is an open technology platform for transmitting real-time audio over IP networks. It was designed to meet the exacting requirements of the broadcast industry and operates at Layer 3 of the network stack, which means RAVENNA streams are ordinary routable IP traffic that can cross LANs, VLANs, and WANs instead of being confined to a single switched segment. Media transport uses RTP, synchronization relies on the Precision Time Protocol defined by IEEE 1588-2008, and sessions follow SIP/SDP-style description conventions.

Because RAVENNA is an open specification rather than a proprietary protocol, implementations come from many independent vendors, and no licensing fees are attached to the technology itself. A RAVENNA system runs on standard Ethernet switches and routers with no specialist media hardware in the core, and audio traffic can coexist with other converged-network services when QoS and multicast are planned properly.

That openness also sets expectations. RAVENNA performance is directly related to the quality of the underlying IP infrastructure, so switch buffering, traffic prioritization, IGMP behavior, and PTP planning remain part of the audio system design. Teams that already operate a managed corporate network can usually carry RAVENNA audio on it without building a parallel media network.

• Layer 3 operation: streams are routable IP traffic, not limited to one switch fabric.
• Standard protocol stack: RTP for media, PTP (IEEE 1588-2008) for time, SDP for session description.
• Multi-vendor and license-free: any manufacturer can implement RAVENNA without royalty obligations.
• Coexists with existing network services when QoS, VLANs, and multicast are engineered properly.

How RAVENNA aligns with AES67

RAVENNA and AES67 were shaped by the same engineering problem — moving professional audio reliably over standard IP networks — and they draw on the same building blocks: RTP media transport, PTPv2 synchronization, and SDP session description. RAVENNA's design philosophy makes its implementations inherently capable of AES67 operation, and most RAVENNA devices expose an AES67 interoperability mode or AES67-compatible profiles alongside their native capabilities.

In practice, interoperability comes down to configuration. An AES67 exchange requires both sides to agree on the PTP media profile and domain, a compatible packet time (typically 1 ms, or 48 samples at 48 kHz), linear PCM payload, and matching multicast behavior. When those parameters line up, a RAVENNA transmitter can feed a pure AES67 receiver and vice versa, which is why RAVENNA is often described as a superset that includes AES67 rather than a competitor to it.

It is worth being precise about compliance claims. AES67 has no formal certification body, so statements such as "AES67 compatible" are declared by the manufacturer rather than independently audited, and products differ in which optional parts they implement. For procurement, the reliable approach is to verify the interop mode details in each product's documentation and confirm the exact stream parameters during acceptance testing.

"AES67-compatible" is a manufacturer declaration, not a certificate. Verify profiles, packet times, and PTP settings against the product documentation before finalizing a design.

Broadcast workflows with SMPTE ST 2110

In broadcast facilities, RAVENNA's closest standards context is SMPTE ST 2110, the suite that transports video, audio, and auxiliary data as separate essence streams over IP. ST 2110-30 defines PCM digital audio building directly on AES67, with conformance levels A, B, and C constraining packet time and channel count. ST 2110-31 transports AES3-encoded audio transparently using the AM824 payload. RAVENNA implementations support these payload types together with the ST 2059-2 PTP profile that broadcast timing plants expect.

Broadcast reliability requirements go beyond any single stream. RAVENNA supports network redundancy, and many implementations offer ST 2022-7-style seamless protection switching, duplicating each flow across two parallel networks so a receiver reassembles a hitless signal even when one path fails entirely. For mission-critical program paths, this class of redundancy is usually a baseline requirement rather than an option.

Gateways complete the broadcast picture. RAVENNA-capable gateways and interface converters bridge legacy AES3, MADI, and analog equipment into the IP domain, letting facilities migrate plant infrastructure incrementally instead of all at once. Combined with the ecosystem's management tools, a broadcast plant can run RAVENNA audio essences alongside ST 2110 video while sharing one timing reference.

• ST 2110-30: PCM audio essence with conformance levels A, B, and C.
• ST 2110-31: transparent AES3 transport using the AM824 payload format.
• ST 2059-2 PTP profile for alignment with broadcast timing.
• ST 2022-7-class seamless redundancy where implemented by the product.

RAVENNA in ProAV deployments

Outside broadcast, RAVENNA has become a practical foundation for professional AV systems: multi-room installations, large venues, stadiums, and campus-scale developments. Instead of point-to-point analog or AES3 wiring, audio routes over the IP network, so a source connected once can be distributed to any zone, and routing changes become configuration rather than re-cabling. Well-designed systems can carry several hundred or even thousands of concurrent streams.

The platform's format flexibility matters at this scale. A single RAVENNA network can carry streams with different sample rates and data formats simultaneously, and multiple media clocks can operate at the same time, so a critical live path and a background-music distribution can coexist without forcing one compromise setting on every device. Synchronization stays phase-accurate across all of it.

ProAV deployments also benefit from RAVENNA's fit with converged corporate networks. Audio can run on the same managed infrastructure as booking systems, digital signage, and office traffic once VLANs, QoS, and multicast behavior are agreed with the IT department, which keeps installation and operating costs closer to standard network projects than to bespoke media systems.

• Multi-room and campus-scale routing without point-to-point re-wiring.
• Hundreds to thousands of concurrent audio streams on commodity switches.
• Mixed sample rates, formats, and media clocks on one network.
• Coexistence with corporate IT traffic under a planned QoS and multicast policy.

Latency classes and synchronization

RAVENNA's latency model is deliberately flexible: latency is configured per stream rather than fixed system-wide. On a well-engineered LAN, latencies in the sub-millisecond region are achievable for interactive and stage-adjacent paths, while the same platform can be configured with substantially larger buffers for streams that cross WAN distances. Each stream receives the budget its application actually needs.

The numbers relate directly to packetization. RAVENNA supports packet sizes from 1 sample per packet up to 192. Smaller packets shorten the packetization contribution to latency at the cost of more packets per second of audio; larger packets do the reverse. Total latency is the sum of packetization time, network transit, and the receiver's de-jitter buffer, so sample rate, packet size, and network design together set the achievable floor.

Synchronization is handled by PTPv2 (IEEE 1588-2008), which distributes a common time base with nanosecond-level precision. RAVENNA achieves sub-sample-accurate alignment across nodes, and multi-site installations can discipline the clock from GPS so that distant plants share phase as well as frequency. Media clock distribution is phase-accurate, which is what keeps large distributed systems truly time-aligned rather than merely running at the same rate.

• Per-stream latency: no forced system-wide default.
• Sub-millisecond operation is possible on well-designed LANs; larger budgets suit WAN paths.
• 1 to 192 samples per packet trades latency against packet overhead.
• PTPv2 (IEEE 1588-2008) time base with sub-sample-accurate, phase-aligned media clocks.

The RAVENNA device ecosystem

Because the technology is open, the RAVENNA ecosystem spans the device categories a real installation needs: mixing consoles and processing cores, stage boxes and microphone preamp interfaces, format converters and gateways for AES3, MADI, and analog, playout and recording systems, and monitoring tools. Products implement RAVENNA natively and interoperate at the stream level rather than by cloning one vendor's control stack.

Representative vendors illustrate the breadth: Lawo builds RAVENNA-capable consoles and I/O infrastructure, Merging Technologies offers interfaces and playout systems, and DirectOut specializes in converters and gateways. These are factual examples of ecosystem members rather than an endorsement list, and procurement should evaluate each product on its own documentation.

Above the streams sits the management layer. RAVENNA implementations commonly use Bonjour/mDNS for discovery with RTSP-style connection management, and larger installations coordinate devices through orchestration and control systems — NMOS and Ember+ integrations and management platforms from the major ecosystem vendors are all used in the field. As with devices, control-layer choices are per-product and should be confirmed early in design.

Within this ecosystem, SoundNet modules also join via the standard SAP stream-discovery mode, and offer small-channel, cost-effective options alongside a wide range of OEM development models.

• Consoles, stage boxes, converters, gateways, playout systems, and monitoring tools.
• Implementations from multiple vendors, including Lawo, Merging Technologies, DirectOut, and SoundNet.
• Bonjour/mDNS discovery with RTSP-style connection management.
• Control via NMOS, Ember+, and vendor orchestration platforms depending on the product.

SoundNet modules inside the RAVENNA ecosystem

RAVENNA devices offer an AES67 interoperability mode, and that is exactly where the SoundNet portfolio fits: the YW series modules implement the AES67 transport and PTP synchronization layer that RAVENNA aligns with, so manufacturers can give consoles, processors, and amplifiers a direct path into RAVENNA networks without licensing a proprietary stack. Channel counts scale from 4×4 to 128×128 at 48/96/192 kHz.

Complementing the modules, SoundNet ships tools that operate directly on live networks: AES67 Handy is a mobile controller that discovers RAVENNA devices on the LAN, monitors live channel levels, and adjusts volume from an iPhone, iPad, or Android device, while the open-source Stream Monitor keeps AES67 multicast streams in view from an engineering laptop. The AES67 LINK GUI software pairs DSP control with AES67, exposes its full API, and supports deep ODM/OEM development.

• AES67 endpoints for RAVENNA networks: the modules speak the same PTP/RTP layer that RAVENNA interoperability relies on
• Scalable from 4×4 edge devices to 128×128 Mini-PCI high-density transport
• Small-channel, cost-effective options: the economic 4×4 YWR04 module covers basic endpoints with full AES67 capability at the best value
• Standard SAP stream discovery: multicast flows are announced as standard SAP/SDP sessions, so SoundNet endpoints and RAVENNA devices can discover each other's streams
• AES67 Handy: discover, monitor, and control RAVENNA devices from a phone or tablet
• Various OEM development models: from module integration and branded production runs to custom development for your product line

What RAVENNA is not: boundaries and misconceptions

RAVENNA is a technology solution set, not a single closed standard with one compliance certificate. Devices vary in which parts they implement — payload formats, sample rates, channel counts, redundancy — so "RAVENNA capability" on a datasheet is the beginning of an evaluation, not the end. The comparison table on this page shows how much wider RAVENNA's optional space is than the constrained profiles of AES67 and ST 2110.

When RAVENNA equipment interconnects with pure AES67 devices, the boundaries land in discovery and control. AES67 defines neither, while RAVENNA implementations typically discover with Bonjour/mDNS and manage connections through their own tooling. Audio can flow perfectly once streams are established, but setting them up may require manual SDP entry or a third-party controller on the AES67 side, and the stream parameters themselves must stay inside the AES67 profile: 48 kHz, the agreed packet time, linear PCM, and limited channel counts.

The extras do not cross that boundary. DSD and DXD formats and very high per-stream channel counts have no AES67 counterpart, and packet times beyond the AES67-mandatory 1 ms — such as RAVENNA's 192-sample packets — are only optional in AES67, so these capabilities remain on the RAVENNA side of a hybrid system. Planning a deployment means deciding explicitly which streams live in the interoperability world and which stay native.

Choosing RAVENNA for your system

RAVENNA earns its place when a project needs what the full platform offers: dense multichannel distribution, mixed formats and sample rates, per-stream latency control, redundancy, and multi-vendor openness — at broadcast scale or ProAV scale. Facilities that expect channel counts to grow, rooms to be added, or systems to merge over time benefit from a platform whose ceiling sits well above today's requirements.

When a deployment only needs to exchange audio across a mixed-vendor ecosystem, AES67 interoperability may be all that is required — and RAVENNA devices serve that role too, because the AES67 subset is built in. The decision point is how much of the wider platform you intend to use. A system that will never exceed 48 kHz stereo distribution gains little from DSD support, while a broadcast or campus plant quickly reaches the limits of constrained profiles.