Using per-title encoding for video Streaming

Why should you care?

‍

‍

If your ask is to deliver Netflix like video streaming quality, “Per-Title Encoding” is the answer. Back in Dec 2015, Netflix introduced this concept of per-title-encoding enabling a shift towards more efficient high quality online video delivery. The traditional static or step ladder encoding methods using AWS/Apple recommended settings for playback are now replaced, especially if your goal is growth while delivering exceptional video experiences.

‍

What does per-title encoding do?

In a nutshell, it adds an extra layer to the encoding process by analysing the video asset. The results are used to calculate the most suitable bitrate ladder for every single asset and thus elevating the video quality while maximising compression efficiency.

Per-title encoding leverages advanced machine learning algorithms to analyse the content of each video and determines the best encoding parameters for that specific title. These parameters include crucial aspects like bitrate, resolution, and compression methods. Through this customized approach, per-title encoding optimizes the utilization of bandwidth and storage, ultimately resulting in an enhanced video quality.

‍

Steps involved in per-title encoding process

1. Content analysis: The video content is analysed to assess its complexity, motion, and other characteristics. This analysis helps determine the encoding parameters that will yield the best results for that specific video.
2. Encoding: Once the optimal settings are determined, the video is encoded using those settings. This results in a video file that is tailored to the content's unique requirements.
3. Adaptive streaming: Per-title encoding is particularly valuable for adaptive streaming, where the video player can seamlessly switch between different quality levels based on the viewer's device and network conditions. This ensures that viewers always receive the best possible quality without buffering or interruptions.

‍

Replacing the traditional encoding models

The traditional encoding model or the “old way” (as we think) was using a static encoding ladder which enables the video playback quality on the viewer’s end to automatically (or dynamically) change. This takes in factors like network bandwidth and device resolution capacity.

With higher bandwidth, the video player would opt for a higher resolution/quality such as HD 720p to Full HD 1080p. And with a lower bandwidth it’s the vice versa – shifting down from 1080p to 720p or lower. The YouTube “Auto” video quality setting is a real example of what we are talking about.

‍

‍

Adaptive bitrate ladder

The process of auto-switching between video quality happens through an Adaptive Bitrate Ladder (ABR). It is a collection (manifest file) of video file segments with different resolutions that are available on the streaming server.

Here’s what apple recommends as an adaptive bitrate ladder to best suit their iOS devices while encoding a video:

‍

The challenge: It’s static!

Using a static encoding ladder is not the ideal approach as these ladders are fundamentally inefficient considering the new era. Every video content created today is different with different levels of complexity.

For instance, some videos have fast action scenes (sports, action genres) and some are more slow-moving. While animated (anime) videos have low-textural features and some are highly-detailed (Avatar).

Take, for example, the diversity of videos. On one hand, videos with rapid action sequences, like sports and action genres, and on the other hand, slower-paced ones. Additionally, when considering animated content (anime), videos have minimal textural intricacies, while others are rich in intricate details (Avatar Movie).

To deal with these scenarios and variations in video complexity levels, a better approach to video encoding could be considered replacing the static encoding ladder model. Therefore, the solution: Per-Title Encoding.

‍

‍

Benefits of per-title encoding

1. Improved quality: Per-title encoding maintains a consistent delivery of high-quality video. By adjusting encoding parameters to match the content's characteristics, per-title encoding minimizes noise or visual errors (artifacts) and while maximising compression, resulting in a better viewing experience.
2. Bandwidth efficiency: Per-title encoding maximizes the efficiency of bandwidth usage. Videos are encoded at the lowest bitrate that still maintains the desired quality, which is especially important for viewers with limited internet connections.
3. Reduced storage costs: Video streaming services can significantly reduce their storage costs by implementing per-title encoding. With smaller video files, less storage space is required, making it more cost-effective to store and deliver content.
4. Device compatibility: Per-title encoding ensures that videos are optimized for a wide range of devices, from smartphones and tablets to smart TVs and desktop computers. This means that viewers can enjoy high-quality video content on any device without the need for manual quality adjustments.
5. Viewer retention: Delivering high-quality video with minimal buffering, startup time and interruptions enhances viewer satisfaction, ultimately leading to increased viewer retention and engagement.

‍

FastPix helps you achieve better results with video experiences to summarize

Per-title encoding helps us enhance the video quality and deliver high-quality content efficiently and effectively. As streaming services continue to compete for viewers' attention, per-title encoding has become a critical tool in ensuring a superior streaming experience. It not only benefits viewers by providing a seamless and exceptional viewing experience but also allows content providers to save on storage costs and reach a broader audience across various devices and network conditions.

Click here, to discover how FastPix Video API lets you to effortlessly create and deliver high-quality video experiences to a global audience in a matter of minutes.