Both SDI and IP facilities require accurate timing to process, route and play-out audio and video media.
Some engineers may be confused about key differences between a Precision Time Protocol (PTP) l transparent clock and a PTP boundary clock. This article will help clear up any confusion and suggest which of the two may be more suited to an IP broadcast center.
The term PTP is defined by the specification IEEE-1588, which is an algorithm that synchronizes the peripherals on the network with a common time reference. Before PTP, there was the network time protocol (NTP), which allowed systems to be synchronized within a few milliseconds of coordinated universal time (UTC). While that was sufficient for basic time of day information, it is not enough satisfactory precise or real-time applications like audio and video. To improve accuracy, thus enabling real-time applications to utilize the Ethernet network, the Institute of Electrical and Electronics Engineers (IEEE) defined PTP.
The method is simple. A high precision clock, the master, transmits PTP sync messages using User Datagram Protocol (UDP). Slaves then receive the sync messages with the master time (t1). If hardware timestamping is not provided by the master clock, a Follow_Up message will be sent out to provide the time at which the initial sync message was transmitted (t1). The slave stores the time at which it receives the Sync message (t2). After reception of the Sync message, the slave sends out a Delay_Request to the master clock (t3). The master finally answers with a Delay_Response message (t4).
Figure 1. Highly accurate time references are required in both SDI and IP facilities. This drawing shows how those clock pulses move within a media facility. Click to enlarge. (Image courtesy of Michel Proulx).
Armed with these four timestamps, the slave device can estimate the propagation delay and calculate its own offset from the master. The following formulas are used in the slave to establish the time at the master:
Equation 1: PTP Delay and Offset calculation:
Delay = [(t2 – t1) + t4 – t3)]/2
Offset = (t2 – t1) - Delay
The following image shows the low-level PTP algorithm.
Figure 2: This illustration of a PTP algorithm illustrates how a slave can accurately determine the proper time, taking into account any system propagation delays. Click to enlarge.
The transparent clock will relay the master’s PTP Sync, Follow_Up and Delay_Resp messages to all of the Embrionix’s IP SFPs and will transfer PTP Request_Delay messages from all the SFPs back to its master. The PTP master usually can answer only a limited number of slaves. When this number becomes too large, the master starts to be over solicited by the number of messages.
As a network gets more congested, packet scheduling across the network can add more delays. The result can cause an inaccuracy in time synchronization because PTP messages have different delays that are uncompensated. The transparent clock adjusts the PTP messages to remove the delays of its own packet processing, and thus compensates for any delays in PTP messaging.
In a typical facility spine-leaf architecture (Embrionix resource document), the top of rack switches can adjust the delay inside the PTP to ensure it is viewed as transparently as possible.
Figure 3. In this example of a transparent clock system, the Top of Rack (TOR) calculates the correct master clock time and then adds a correction factor in the PTP message. Click to enlarge. (Image courtesy of Michel Proulx).
The boundary clock switches already possess a built-in PTP master clock. The switch acts as the master clock for any endpoint devices attached to it. To provide extra stability, the switch will be a slave to another PTP master clock.
In this scenario, the PTP master clock in the switch will communicate to a limited number of slaves. This way the boundary clock method ensures that PTP masters are not over solicited, which greatly improves the accuracy of the PTP time and the system scalability. The following figure shows a boundary clock system.
Figure 4. In this boundary clock system design, the PTP master clock synchronizes downstream slave clocks, which then drive other, downstream, master clocks. Click to enlarge. (Image courtesy of Michel Proulx).
Timing Signals Generation in IP Media
Once the slave has iteratively computed its time difference with its master, it is synchronized within sub-microsecond accuracy of UTC. But time synchronization is just the first step in facility timing. The second step is to use precise time to derive the timing reference signals needed by ST 2110 audio/video devices.
The SMPTE created two standards, ST 2059-1 and ST 2059 that support PTP. The ST 2059-1 defined a method of deriving phase aligned audio and video sync from PTP. The ST 2059-2 defined a profile of IEEE-1588 suitable for audio and video requirements.
From the PTP ticks count, slaves can accurately find the vertical and the horizontal pulses and align its video to the video reference. It is also possible to extract the audio clocks and then generate the Digital Audio Reference (DARS) signal, if required. The following image shows the process of re-creating the clock for synchronization of signals.
Figure 5. This drawing illustrates how timing references could be processed throughout an IP media center. Click to enlarge. (Image courtesy of Michel Proulx).
IP Broadcast Environment Clocking
In an SDI broadcast environment, timing is crucial. It is no different in an IP broadcast environment; timing is important to synchronize endpoints. PTP was already used by various industries, so it was an obvious choice to be reused for broadcast.
From an industry’s perspective, PTP is mandatory and therefore is now defined in the ST 2110-10 standard for Audio, Video and Metadata Transport. To be fully compliant with the standards, any ST 2110 endpoint (device) also needs to be PTP/ST 2059-1 and ST 2059-2 compliant.
Timing is Critical
Finally, Embrionix endpoints such as the emSFP gateways, emQUAD, emVIEW and emFUSION will work in both types of PTP designs, transparent clock and boundary clock. But for scalability and accuracy, the boundary clock system should be implemented whenever possible over the transparent clock.
Renaud Lavoie, President and CEO, Embrionix.
You might also like...
We call them hard disks to distinguish them from floppy disks. As the latter have practically died out as a result of progress in solid-state storage such as flash memory, it is probably not necessary to specify that disks are…
Immersive audio has the great potential to transform our human listening experience, captivate our imagination, and inspire our inventiveness.
In part 8 of the series “Data transmission and storage”, consultant John Watkinson looks at some of the intricacies of RF transmission.
Part one of this four-part series introduces immersive audio, the terminology used, the standards adopted, and the key principles that make it work.
John Watkinson introduces the idea of channel coding to convert the uncontrolled characteristics of data into something that works within the limitations of real media.