Thee Network Transport Foundation of AES67

AES67 nie wynalaza networking. It operates as a profile that selects specific, proven technologies frem the standard Internat protocol approve. The protocol stack is designate tte to priorititize low latency andd real-time delivery over reliability mechanisms that implemente delay. Thii fundamental desin choice dicats thee entire packet structure.

Dlaczego RTP / UDP / IP?

AES67 mandates the use of the Real- time Transport Protocol (RTP) over User Datagram Protocol (UDP) over Internet Protocol (IP)Thile combination is deliberately chosele over TCP. While TCP provides equivable delivery them exceptiar them exceptiar them exception through transmissionon, these retransmissions introdule variable latency andd jitter that are unacceptable for live audio. RTP providees thee necesary sequencing andd timing information for real-time streams, while UDP providepence a fast, connectionless transport layer. AES67 specially utilizes IP multict, allowent a single audio source o be amended te o multiple receivers neemping datation thel, making hity expercente effect effectifön a single infön.

Anatomy of an AES67 Packet

An AES67 packet consists of a sequence of headers wrapping a payload of audio samples. Understanding thee functiong of each headder field is critial for troubleshooting packet loss, latency issues, and syncization errors. The total packet size is limited thee Ethernet Maximum Transmissivon Unit (MTU), typically 1500 bytes, which limits the number of audio channels or samples per packet.

Ethernet Frame (Layer 2)

Ten czas zaczyna się od tego fizyka. Adresaci MAC. In a multicast stream, the destination MAC addios is a multicast addios derived from thee IP multicast group, typically in thee range of - Nie. for IPv4. Switches use this adresses to forward traffic efficiently. The headder also contains an optional 802.1Q tag VLAN, which is highly recommended for segregating audio traffic frem tell network data and for carrying Priority Code Point (PCP) bits for Quality of Service (QoS).

The EtherType field (0x0800 for IPv4 or 0x86DD for IPv6) identifies the network layer protocol.

IP Headder (Layer 3)

Te IP headder manages logical adressing and routing. AES67 typically useses IPv4 multicasto adresses in then 239. 0. 0 to 239. 255. 255. 255 range (administratively scoped). Differentiated Services Code Point (DSCP) Field is perhaps the mott critical byte in then IP headder for audio quality. AES67 zaleca specjalne wartości DSCP (typically) EF - Expedited Forwarding, or AF41) to ensure that audio packets receive priority treatment across changes androuters. Without proper DSCP marking, audio packets can be dropped during network congestion.

Nagłówek UDP (Layer 4)

Te UDP headder is minimal but essential. It providele source and destination port numbers. The destination port is used by by thee receiver to demultiplex thee incoming straim tam thee correct application. The 16- bit checksum provides basic error confidention for thee UDP datagram and its payload. Because UDP is connectionless, there is no handshake, allowing for expicate data flow.

Te RTP Headder

Te RTP headder is thee heart of thee AES67 packet, provising the mechanisms for ordering, timing, and payload identification. The fixed headder is 12 bytes long.

  • Version (V), Padding (P), Extension (X): AES67 typically does not t use thee headder extension.
  • CSRC Count (CC): Identifies the number of contribuing sources.
  • Marker (M): Used to signal boundaries, such as the start of a talk spurt or a video frame boundary in ST 211010- 30.
  • Payload Type (PT): For AES67, Consun values include L16 (Linear PCM 16- bit) or L24 (Linear PCM 24- bit), often using dynamic payload type numbers (np., 96 or 97) digitated via SDP.
  • Sequence Number: Increaments by one for each RTP packet sent. This is the primary tool for detelting packet loss, misordering, and for management the jitter buffer on thee receiving end. A wrap- around of thee sequence number is handled by thee jitter buffer logic.
  • Czas stampu: Reflects the sampling instant of the first sampe in the timestamp is derived directly frem the Klock PTPA 48kHz stream will increment this timestamp by 48 ticks per millisecond or 48 ticks per sampe period. The receiver uses this tio reconstruct thee original audio timing and clock domayn.
  • Synchronization Source (SSRC): A unique 32- bit identifier for the source of thee stream. This is used to differencish multiple streams from the same source.

Audio Payload

Te payload contains thee raw or encoded audio samples. AES67 mandates support for Linear PCM at 48 kHz sample rate with 16, 20, or 24- bit depth. The payload can also carry AM824 (Audio over USB) format. The number of samples per packet is determinaed by the packet time (ptime). Common values included 1ms (48 samples), 125 microsebs (6 samples), or 4ms (192 samples).

A 1ms packet time offers lower latency but increates network overhead (more packets per second). The payload size size must be carefuly calculated to o fit the network MTU toavoid framentation, which is strictly forbidden in AoIP networks due to thee latency and comprecity it intomietes.

Precision Timing with PTPv2 (IEEE 1588)

Synchronization is the single most contribuing aspect of AoIP. AES67 mandates the use of thee IEEE 1588- 2008 Precision Time Protocol (PTPv2) This ensures that the transmitter 's sample clock ande thee receiver' s sample clock are perfectly synchronized, preventing buffer underruns or overruns that cause audible clicks andd pops.

The Best Master Clock Algorithm

Every PTP- capable device on thee network runs thee Bess Master Clock Algorithm (BMCA). The BMCA analyzes clock quality acquisites - such as clock class, clock clippeacy, and priority - to elect a single Grandmaster (GM) ClockAll tenor devices synchize te GM. In a property ly designed system, thee GM is often a decretated, highly stable device (np., a GPS- locked master clock generator). The BMCA also provides automatic failover; if thee GM failes, a new GM is elected from thee eling devices.

PTP Message Exchange

PTPv2 definiuje specjalny message sekwencji to correct clock drift. The GM periodically sends a Sync message. It impetately follows this with a Follow _ Up message containg the precise time thee Sync was sent. The slave device sends a Delay _ Req message andd records the time it is sent. The GM responds with a Delay _ Resp message, containg the precise time thee original Delay _ Req was received.

Bycomuting thee ronda-trip time, the slave can calculate it offset the GM andd adjuss its local clock. This process repects constantly, maintaing sub- microsecond closacy across thee network.

Session Discovey wigh SAP andSDP

How does a receiver find a n acceptable AES67 stream? The standard uses the Session Announcement Protocol (SAP) for discvery. AES67 transmitery periodically multicast a SAP zapowiada to dobrze znanych adresatów SAP. Thii zapowiada zawiera a Session Description Protocol (SDP) Payload.

Te SDP file is thee recipe for thee stream. It contains all thee metadata a receiver neds to connect, synchize, and decode thee audio.

  • Media Description: Xi1; Xi1; FLT: 0 Xi3; Xi3; definites the media type, port, and transport (RTP / AVP).
  • Connection Data: Xiv1; Xiv1; FLT: 1 Xiv3; Xiv3; specifies the e multicast group IP addios.
  • Attributes:
    • ptime: Packet time in milliseconds (np., 1,0).
    • rtpmap: Maps the payload type tu an encoding (np., Xi1; Xi1; FLT: 2 Xi3; Xi3; means 2 channels of 24- bit PCM at 48kHz).
    • Media Clock: Xiv1; Xiv1; FLT: 3 Xiv3; Xiv3; or a specific PTP timestamp reference.
    • / Xion1; Xion1; FLT: 4 Xion3; Xion3; links the stream clock to the PTP grandmaster.
    • Channel Order: While SMPTE ST 2110- 30 specifies channel ordering for AES67, many streams rely on a mutually concord channel map.

Mechanizmy pływowe Data

Te flow of audio data from source te sink is a continuous, tightly orchestrated process. understanding thee transmiter andd receiver operations is key tu diagnosing network or audio artifacts.

Operacje transmitter

Te audio interface digitalizas analoge audio samples at a constant rate (np., 48kHz). The transmiter 's digire reads these samples from the hardware and places them into a buffer. At every packet interval (np., every 1ms), thee dissor encapsulates a block of samples into an RTP packet. Thee sequence is incremented, and timestamp iset to thee PTPTPaligned plane count. Thee packet ithen wrapped UP and IP heaked, marked with correct, and sent, and sent thee contriburect.

Odbiorca Operations and Jitter Management

Thee receiving device listens for SAP anvecements. Thee user selects a stream, andthee receiver subskrybent to thee multicast group using IGMP (Internet Group Management Protocol). Once subscribed, packets begin arriving. However, network changes introdule variable delays (jitter). The receiver 's first task is to place incoming packets into a jitter bufferThis buffer smooths out thee network delay variations.

  • Jitter Buffer Sizing: Te buffer must be large enough to absorb thee worst- case network jitter with out underflowing, but small enough to limit latency. Typical jitter buffers in AES67 devices range from 1ms to 20ms.
  • Sequence Number Check: Te receiver monitoruje sekwencje numbers for gaps, indicating lost packets. It can also use tio reorder out - of- order packets.
  • Recovery Clock (PLL): Thee receiver wykorzystuje thee RTP timestamps derived from thee PTP clock to drive a Phase- Locked Loop (PLL). This PLL reconstructs thee transmitter 's sample clock, ensuring that its internal DAC (Digital- to- Analog Converter) runs in perfect syncization with the source.

Handling Latency andloss

Strönch delay, packetization delay (ptime), network transit delay (uplinks, queuing) is sum of selial factors: sampling delay, packetization delay (ptime), network transit delay (uplinks, queuing) ite delicres, jitter buffer delay, and processing delay. A 1ms ptime with a 5ms total end end-to-end latency is evaluable in a well-tunetwork. While AES67 does not mandate specific Forward Error correction (FEC), it does support RTP Redancy (RFC 2198).

Network Infrastructure Requirements

AES67 places strong demands on network hardware. Standard office- grade changes are often unappropriable for high-channel- count AoIP. Professional managed changes are required.

  • IGMP Snooping: Switches must listen to IGMP messages to forward multicast traffic only too ports that have explamitly requested it. Without IGMP Snooping, multicast traffic floods to all ports, wasting bandwidth.
  • PTP- Aware Switching: For sub- microsecond closacy across large networks, changes should be support PTP Boundary Clock (BC) or Transparent Clock (TC) functiality. This corrects for thee latency the switch itself intro the PTP timing messages.
  • Quality of Service (QoS): Strict QoS queuing (np., Strict Priority) must be configured on all switch ports. PTP event messages should receive thee highest priority (CS7), followed by audio traffic (EF or AF41), and then best-effict data.
  • Bandwidth Planning: Inżynierowie mutt calculate total stream bandwidth. A single 24- bit, 48kHz, 8- channel stream consumes routly 9.2 Mbps. A network handling 64 such streams requires condicates controlly 600 Mbps of dedicated audio bandwidth. Proper planning prevents over- subscription of uplinks.

Interoperability in Practice

Te pierwsze goale of AES67 is to make different t decrers concerts; gear work together. Thii is typically enabled in devices through gh an quentig; AES67 Mode. Quenticule; For example, Dante devices can be configured to output or decuste AES67 streams, albeit often with a limition on channel count or packet time compared to native Dante flows. Livewire + and Q- LAN devices cain similarly interface via AES67.

Common sability issues often aris airs from mismatched configurations. A transmiter sending 1ms packets may not t compatible with a receiver that only supports 4ms packages. Clock domain alignment is anotherr dispects comparaters competione; all devices must be locked to thee same PTP granmaster. Network confilers mutt pascloce attention to the SDP parameters reklamowane by each device and ensure thee network infrastructure (PTP profiles, DSCP values, and VLAN configuribution) is conficuts conficuts the the the thee stintire stem. For further rett optung, reptung, ref, ref, refö@@ IETF specification on PTPv2 and thee IEEE 1588- 2008 standard.

Praktykal konfiguracyjny guides are access from Audinate 's AES67 resources.

Konkluzja

AES67 przedstawia Fundational standard for thee modern, incorporate audio-over- IP ecosystem. By breaking down the packet structure - frem the Ethernet frame andd RTP headder the PTP timestamp andd audio payload - we can see how every layer contributes to the robutt, low- latency transmissivoon of professional audio. Thee mechanics of data flow, consine by precise timing and careful buffer management, ensure thatt audio delis delid reliably acrossi netx work topopologies. For professials, moving beyond a black- box technologies ologi teses, ensions, exergens estinen.