Introduction to Video Streaming

Like billions around the world, I regularly consume endless hours of on-demand and live video through platforms like YouTube, Netflix, and Twitch. However, as an engineer, I'm curious about how these services reliably deliver high-quality video to my device. Streaming video content seamlessly is a marvel of modern technology, but what goes on behind the scenes to make this possible?

Recently, while contributing to an open-source project that enables users to take control of their live streams, I delved into the intricacies of media streaming. Various protocols have been developed to support different use cases such as low-latency streaming, TCP-based or HTTP-based streaming, adaptive bitrate streaming, and many more. Platforms like Zoom, Netflix, YouTube, Twitch, Spotify, and others utilize different protocols based on their unique architectural needs.

In 2020, I also developed my own WebRTC-based video file streaming platform to overcome the shortcomings of existing screen-sharing features in video conferencing apps. This hands-on experience deepened my understanding of the complexities involved in delivering smooth and efficient video streaming.

In this series, I aim to share my knowledge about the whole video streaming space and cover as much ground as possible. Let's start by exploring the different types of streaming and their respective complexities.

Different Types of Streaming

On a high level, streaming can be categorized into three main domains:

Live Streaming

• Definition: Streaming from a client to a server, which then serves the content to viewers in real-time.
• Examples: Twitch, YouTube Live, Instagram Live.
• Key Characteristics: Low latency, high interaction, real-time engagement.

Video-On-Demand (VOD)

• Definition: Streaming videos that are pre-recorded and stored on servers, available for playback at any time.
• Examples: Netflix, Amazon Prime Video, YouTube.
• Key Characteristics: High quality, adaptive bitrate streaming, extensive libraries of content.

Video Calling/Conference

• Definition: Streaming video in real-time between a group of users simultaneously, typically using a peer-to-peer architecture.
• Examples: Zoom, Microsoft Teams, Google Meet.
• Key Characteristics: Real-time communication, low latency, often includes additional features like screen sharing and collaborative tools.

Although these categories might seem simple and similar on the surface, there are numerous complexities involved depending on the use case. Here’s a brief overview of some of these complexities:

• Protocol Selection: Choosing the right protocol (e.g., WebRTC for low latency, HLS for scalability) based on the streaming requirements.
• Bandwidth Management: Ensuring that streaming quality adapts to varying network conditions without significant buffering or loss of quality.
• Latency Reduction: Particularly crucial for live streaming and video conferencing, where delays can disrupt the user experience.
• Scalability: Handling a large number of concurrent streams, especially in popular live events or platforms with massive content libraries.
• Content Delivery Networks (CDNs): Distributing content efficiently across the globe to minimize latency and improve streaming quality.

In the following articles, we will dive into each of these aspects in greater detail, exploring the technologies and techniques that make modern video streaming possible.