Network Audio

OSI Model

The OSI Model groups different services, functions and applications of telecommunication systems into seven hierarchically arranged layers:

OSI Model





Application Layer

End user layer, HCI layer


Presentation Layer

data conversion, syntax


Session Layer

connection management, sockets


Transport Layer

end-to-end connections (TCP, UDP)


Network Layer

packet routing


Data Link Layer

data formats (bits to frames, MAC addresses)


Physical layer

bit stream transmission over medium/hardware (Ethernet, WiFi, ...)

Network based audio systems can be based on different layers. This affects their capabilities and application areas. A comprehensive list can be found here: comparison on Wikipedia

Layer 1 Solutions

Layer 1 solutions only rely on the hardware used in telecommunication systems and use their own routing mechanisms. As a consequence, they usually need specific routers and are often used for direct peer-to-peer connections. The most widespread solution is the open AES50 format, which is found in devices by Behringer and Midas.

Layer 2 Solutions

Layer 2 solutions use the standard Ethernet protocol for transmitting data. Standard routers and hardware can thus be used for routing. Among the well known formats are AVB and AES51, as well as several proprietary solutions.

Layer 3 Solutions

Layer 3 solutions feature an IP header in their packages. Example solutions are DANTE, AES67, RAVENNA and AVB.

Layer 4 Solutions

Some solutions are based on Layer 4 protocols like TCP or UDP 1. Since UDP is faster due to the missing handshake and error-correction. Although this makes it prone to package loss, it is the preferred method for achieving acceptable latency at the cost of dropouts, depending on the quality of the connection.

Examples for Layer 4 solutions can be found in the free and open software community, including NetJack2 2, Zita-njbridge 3 and JackTrip.


This needs more references, since it is not unambiguous on which layer they are working.