Live streaming is a rapidly expanding sector, enabling sports events, artists, news organizations and educational institutions to deliver content with minimal latency directly to consumers. The source of the broadcast is generally irrelevant, whether from a live camera stream, produced by OB vans, or a transcoding from an OTT provider. At the same time, consumer devices can be much more diverse: smartphones, laptops, tablets, smart TVs and many others. This imposes certain restrictions on both the broadcaster and the consumer. Typically, this means different bitrates, profile sets, or different codecs must be used. Since the encoder is a crucial part of any broadcasting facility, this requires fine-tuning all aspects of stream production for end-user target devices, while simultaneously ensuring stable coding performance is maintained, regardless of the type of content. Industry experts estimate that live video represents almost half of the entire broadcast market and forecasts indicate that the global video-streaming market will continually expand.
As an industry leader, MainConcept builds many unique features into its codecs. AutoLive, a special mode for streamlining live-encoding use cases, is one of them. AutoLive is part of MainConcept’s HEVC/H.265 and VVC/H.266 offerings, this article focuses on the high-performance implementation of the MainConcept VVC Video Encoder.
Technology is advancing all the time to enhance the efficiency and reliability of content contribution. Live streaming is a comprehensive term in the broadcast industry, ranging from the process of creating video content with the appropriate codec to distributing this content to end-user devices, so simplicity in elements of this chain is critical to achieve a smooth and responsive workflow. Simplifying the encoder configuration as a central workflow element ensures flexibility to react quickly, adapt to content and stay ahead by proactively preparing for "what's next". Since live content can vary significantly, ranging from very simple scenes to more complex and heterogeneous formats, achieving maximum video quality in a compressed stream while maintaining a specific target frame rate is particularly challenging.
Even with careful selection of settings for a broadcast encoder, there are situations where using a pre-configured encoder with a fixed set of parameters may not be best suited to the specific hardware or cause the CPU to become overloaded, leading to drops in video quality or playback drops for end users, resulting in a poor quality of experience (QoE).
This means one thing: a live encoder must be capable of adapting to the content being produced, while simultaneously considering the current settings specified for the broadcast.
MainConcept addresses these live encoding challenges via its high-performance implementation of the VVC/H.266 standard with a new encoding mode: AutoLive. This mode enables content creators to avoid having to fine-tune the encoder during the production pipeline. It is no longer necessary to define appropriate settings to maintain a stable encoding speed to prevent drops while processing the most complex scenes.
By leveraging broadcast environment parameters and the encoder’s internal state, AutoLive balances the coding complexity via heuristics without affecting the stream structure, which is important for further transmission of the stream over the broadcast infrastructure. Natively integrated into the MainConcept Live Encoder application, AutoLive allows seamless configuration and robust content production using the next generation VVC/H.266 broadcast codec up to 4K60p.
First introduced in HEVC/H.265, AutoLive for VVC offers superior flexibility and provides users with extensive configuration opportunities, because the algorithm is not restricted to any predefined performance settings. Speed-quality corrections can be applied around any customized settings set by the user. This allows the user to choose any performance level from the offline encoding performance curve as a starting point and to modify it continuously based on the complexity of the incoming video. This configuration determines the baseline state of the encoder and speed/quality adjustments are performed around this anchor point to achieve the highest possible quality for the given frame rate. The encoder will maintain maximum quality, while for more complex scenes, it will simplify the complexity of coding, ensuring the quality remains unaffected. This prevents dropped frames during encoding. In other words, AutoLive will automatically generate a performance curve for specified settings.
Offline and AutoLive base performance levels offer similar speed and quality, with some minor differences in SPS/PPS headers. This allows offline mode to serve as a reference for evaluating the base performance level of live mode. If there are no other user-imposed restrictions on the stream, users can easily avail of offline performance levels to achieve the desired result more easily than having to fully evaluate live encoding.
MainConcept AutoLive encoding for VVC primarily relies on statistics collected during the encoding process. Based on these statistical observations, the encoder decides when it can increase quality without imposing a negative impact on performance (frame time is much less than needed for the target encoding speed) or improve performance while maintaining almost the same quality. Finally, when statistics signal which target frame rate is closer to the optimal value, the encoder determines that the current feature toolset is optimal for the current scene and no further adjustments are required.
The MainConcept VVC/H.266 Video Encoder offers an option to set the maximum and minimum latency, and AutoLive is particularly sensitive to these settings. Higher latency allows AutoLive encoding to deliver better quality.
However, latency can pose a problem in certain cases. Answers to the following questions can help determine the most balanced solution for a specific use case:
After assessing these factors, users can configure latency settings accordingly. Latency directly affects quality: the lower the latency, the poorer the quality compared to when there are no latency restrictions.
MainConcept, as a leading provider of codec technologies, is committed to delivering exceptional quality and performance in its products. With solutions available in both the VVC/H.266 SDK MainConcept Live Encoder, MainConcept AutoLive is designed to support production professionals in overcoming everyday challenges. Live encoding remains a key focus for MainConcept. By adapting to an ever-evolving landscape, we provide flexible and efficient solutions that empower our customers to deliver superior live-streaming experiences.
Learn about other MainConcept codec enhancements:
Want to try AutoLive? Download our VVC and MainConcept Live Encoder trials today.