Video Quality Analytics (VQA) is crucial for seamless streaming experiences. A video that won’t load is just as frustrating as one that degrades in quality. While availability and scalability are easier to measure, video quality is more complex and demands real-time monitoring for accurate insights.To optimise performance, engineers need to monitor video quality in real-time, pulling data from encoders, networks, players, and other streaming components. However, traditional VQA tools are breaking under the pressure. They’re:
- Resource-heavy: real-time video stream analysis requires significant computing power
- Hardware hurdles: physical probes add cost and complexity to the infrastructure
- Scalability issues: as viewership grows, traditional monitoring becomes increasingly expensive
Thankfully, new industry standards are simplifying VQA, making it more accessible and efficient. Enter Common Media Server Data (CMSD)—offering a smarter, more scalable solution to address these challenges.
What is CMSD?
Unified Streaming has provided a great definition of CMSD: “CMSD refers to a standardised method defined by the Consumer Technology Association (CTA) where media servers, both origin and intermediate, can consistently communicate data with each media object response, ensuring all intermediary servers and players process the information uniformly; essentially, it’s a standard way for media servers to exchange information about media streams during delivery, improving efficiency and performance across different systems.”
At its core, CMSD provides a structured way to capture, transmit, and analyse the key metrics that determine Quality of Experience (QoE) and Quality of Service (QoS). This framework provides a single source of data on the network side about video quality. Combined with Common Media Client Data (CMCD), another CTA-WAVE standard, operations engineers can get a granular and powerful picture of video quality without the compute power needed to combine and crunch multiple data sets.
How CMSD Transforms VQA
So what about CMSD makes it such a powerful dataset for VQA? There are three main ways that CMSD transforms VQA:
- Real-time monitoring. Remember that understanding and optimising video quality must be done in real-time but, without CMSD, that can require a lot of different data sets and a lot of computational horsepower. But data gleaned through CMSD provides data about bitrate switches, buffering, frame drops, and latency. This can provide the background of real-time monitoring without the overhead of multiple data sources.
- Automated issue detection. When streaming is happening at scale, there can be billions of data points generated across the entire workflow. It’s not feasible for operational engineers, or anyone for that matter, to look at that data in real-time. But by leveraging CMSD in VQA systems, data anomalies or data that exceeds thresholds can be elevated automatically, greatly increasing the speed of mean-time-to-diagnose (MTTD).
- Data-driven decision making. With the kind of data CMSD can inject into VQA systems, optimisation and operational decision-making can be driven entirely by data that is relevant to video quality. Because CMSD encapsulates a core set of data points needed to analyse video quality, there is a lot less noise from other, less relevant data sources ensuring that operations remain hyper-focused on the data that matters most.
Industry Adoption: CTA WAVE Project and SVTA’s CMSD Initiative
The CTA Web Application Video Ecosystem (WAVE) project is where CMSD originated. Developed by leading video companies, the CMSD standard will ensure a consistent capture and delivery of data regardless of streaming operator, technology vendor, or network provider. This means that everyone will finally be “speaking the same language” which will significantly reduce the complexity in triangulation of video quality errors. In addition, the Streaming Video Technology Alliance (SVTA) is working on efforts to capture and publish best practices for the use of CMSD data. Together, these and other organisations, are pushing the industry towards a consistent and agnostic source of data for video quality analytics.
In addition, the CTA WAVE project and SVTA are collaborating on a joint standard addressing distributed tracing within the streaming workflow. Where distributed tracing has long existed in network observability, the fragmented nature of the streaming technology stack has prevented the application of this approach which can make it difficult to trace a single session through the entire workflow. But, with standards like CMSD and CMCD creating an agnostic, standards-based consistent data structure for video quality, the joint project, Distributed Media Tracing, will bring a level of observability to streaming operations which will enable streaming platforms to ensure a much better viewing experience.
Next Generation VQA Solution: Touchstream’s eVQA
The key to an end-to-end workflow VQA approach is connecting other workflow components, like encoders, to the standard data framework provided by CMSD. For this, Touchstream has created eVQA, which is an innovative encoder-based video quality analysis solution.
How does it work? First, eVQA improves QoS by providing granular visibility into encoding data. This is done by injecting Media Quality Assessment (MQA) data from the encoder into the CMSD header and passing it downstream into VQA tools which, leverage automation, can even switch encoding pipelines (such as to a redundant server or another server pair) when the data reveals quality issues below an acceptable threshold. Second, eVQA reduces the complexity of encoder quality monitoring. It does this by looking at all output channels without the need to decode the video or apply DRM. In doing so, eVQA acts as an “early detection system” and alerts operations (or automatically switches to a better quality version) before poor quality video is injected into the workflow and reaches the end-user. Finally, eVQA enhances the very backbone of modern streaming architectures–adaptive bit rate (ABR). By transmitting real-time KPIs about encoder output, the ABR ladder can be adjusted to remove bad quality bitrates so that there isn’t even the chance of player selection. All of these benefits ensure one thing–a better viewer experience because video quality issues are addressed upstream, at the encoder, rather than downstream, in the player (where the viewer can see it).
Conclusion: CMSD is Changing the Game
CMSD, as an industry standard, is transforming how video and operations engineers monitor, analyse, and optimise video quality. Because of a standardised approach to collecting data about the video experience, every streaming workflow component, from encoders to CDN caches, can communicate key data metrics which can be ingested by a VQA system to provide real-time observability of video quality. With industry-backed initiatives driving adoption and next-generation solutions like Touchstream’s eVQA leveraging its capabilities, CMSD is quickly becoming an essential tool in the modern video engineer’s toolbelt. The future of VQA is here—and it’s more cost-effective, more scalable, and more intelligent than anything we’ve seen before. The question isn’t whether you should be looking at CMSD-based solutions, but how quickly you can implement them to gain a competitive edge in delivering exceptional streaming experiences.
Ready to learn more about how Touchstream’s eVQA can transform your video quality monitoring? Contact our tech team today for a demo!