Recently we've been asked about lip-sync analysis, a problem that affects traditional Video over IP applications and more recently WebRTC. These applications use transmit audio and video data using separate RTP streams. This combined with the differing audio/video encode/decode, dejitter buffer sizes and passing through transcoders quickly gives rise to different packet arrival times that manifest as lip-sync problems.
Synchronization should be performed by receivers through the NTP timestamps in the RTCP sender reports and the RTP timestamp in the RTP packets. It is possible to use exactly the same method to passively measure the lip-sync error introduced at the network level. Measuring the degree of lip-sync, in frames or milliseconds, is possible in real-time using various characteristics of RTP and RTCP.
For each stream, the analyzer must keep track of the RTP timestamp-RTCP NTP timestamp relationship using the sender report mapping; it is then able to measure the lip-sync delay by calculating the NTP timestamp for each individual RTP packet. Thus the arrival differential can be measured between a video and audio stream in the same session; this equates to an indication lip-sync that can be express in milliseconds or frames.
As part of Valid8's comprehensive testing toolkit are a number of test call generators for testing wireline and wireless networks, including 3G, 4G, 5G, IoT, WebRTC, VoIP, POTS, and more. These include detailed media analysis capabilities and reporting. An example CDR output from a test on the Valid8 platform is shown below:
For small companies that are looking to use test call generators as a part of their overall testing strategy, the Valid8 platform is a cost-effective way to ensure QoS. For more information on Test Call Generation or Valid8, please visit our website at www.valid8.com.