Video streaming looks simple from the outside: you press play and it works. Under the hood, it’s a sophisticated orchestration of encoding, packaging, protocols, CDNs, and real-time decision engines. This post explains modern streaming protocols and then goes deep into each protocols.

1. Why Streaming Protocols Exist

The internet wasn’t built for video. It was built for file downloads.
If video were delivered as a normal download:

• You would wait for the full movie to download.
• Network fluctuations would break playback.
• Mobile users would suffer huge buffering.

Streaming protocols solve this by enabling adaptive delivery — video arrives in small chunks, at different quality levels, adjusted in real time.

Core goals:

• Start playback instantly
• Avoid buffering
• Adapt to bandwidth changes
• Scale to millions of viewers

2. Traditional Streaming vs Modern Streaming

Old approach: Progressive Download

The browser downloads a full video file and starts playing once enough is buffered.

Problems:

• No adaptive bitrate
• Large buffering time
• Poor mobile experience

Modern streaming uses chunked adaptive streaming.

Adaptive Bitrate Streaming (ABR)

Adaptive Bitrate Streaming is the foundation of modern video streaming.
Instead of delivering one large video file, the content is prepared in multiple quality versions and delivered in small chunks.

What “Adaptive” Means

Traditional video downloads required choosing a single quality before playback.
If the network became slow, the video would buffer or fail.

Adaptive streaming solves this by allowing the player to continuously adjust video quality in real time during playback.

The player constantly measures:

• Available network bandwidth
• Device capabilities (mobile, TV, laptop)
• Current buffer health

Based on these signals, the player automatically switches between different video qualities without interrupting playback.

This switching is seamless and invisible to the user.

How the Video Is Prepared

Instead of one file, each video is:

1. Encoded into multiple resolutions and bitrates

Example ladder:

• 240p → 300 kbps
• 480p → 800 kbps
• 720p → 2 Mbps
• 1080p → 5 Mbps
• 4K → 15 Mbps

This set of qualities is called a bitrate ladder.

2. Split into small segments

Each quality version is broken into small chunks (typically 2–10 seconds long).

So instead of downloading a full movie, the player downloads small pieces continuously.

How Playback Works

1. Player downloads a manifest file first
   (.m3u8 for HLS or .mpd for DASH)
2. The manifest lists all available quality levels and chunk URLs.
3. Player begins playback with a safe quality (usually medium/low).
4. While the video plays, the player continuously monitors bandwidth:
5. • If bandwidth improves → switch to higher quality chunks
   • If bandwidth drops → switch to lower quality chunks

Because switching happens at segment boundaries, the viewer never notices the change.

Who Introduced Adaptive Streaming

Adaptive streaming was pioneered by Move Networks (2006), which later influenced the development of modern standards.

Today the most widely used standards are:

• HLS (HTTP Live Streaming) — introduced by Apple
• MPEG-DASH — open industry standard

These technologies power Netflix, YouTube, Prime Video, Disney+, and most modern streaming platforms.

3. Why Multiple Streaming Protocols Exist

There is no single “best” streaming protocol because streaming has very different use-cases:

Each protocol optimizes for latency, scale, reliability, or interactivity.

4. Protocols

4.1 HTTP Live Streaming (HLS)

Developed by Apple

This is the most widely used streaming protocol on the planet.

Used by:

• Netflix
• Apple TV+
• Disney+
• Prime Video
• Most OTT platforms

Supported by:
iOS, Android, browsers, Smart TVs, consoles — almost every device.

How HLS Works

Instead of sending one big video file, HLS:

1. Splits video into tiny chunks (2–6 seconds)
2. Creates a playlist file (.m3u8 manifest)
3. Player downloads chunks via HTTP/HTTPS

The player continuously requests small video pieces just like loading webpages.

Why HLS Became the Industry Standard

Key advantages:

• Works over normal HTTP (CDN friendly)
• Firewall friendly
• Scales to millions of users
• Supports Adaptive Bitrate Streaming (ABR)

This is why Netflix relies heavily on HLS.

Simplified HLS Workflow

HLS is optimized for scale and reliability, not ultra-low latency.

Typical latency: 10–30 seconds

Perfect for movies and TV shows.

4.2 DASH (Dynamic Adaptive Streaming over HTTP)

The open standard alternative to HLS

If HLS is Apple’s version, DASH is the industry’s neutral version.

Used by:

• YouTube
• Netflix (alongside HLS)
• Prime Video
• Many Android and Smart TV platforms

How DASH Works

Technically very similar to HLS.

Main difference:

DASH is codec-agnostic, meaning it works easily with:

• H.264
• HEVC
• AV1
• Future codecs

This makes DASH popular for high-quality streaming.

Why Netflix Uses Both HLS and DASH

Different devices prefer different formats:

• Safari → HLS
• Chrome / Android → DASH
• Smart TVs → mixed support

So Netflix packages videos in both formats and serves the one your device prefers.

4.3 RTMP (Real-Time Messaging Protocol)

The veteran of streaming

Developed by Adobe in the Flash era, RTMP is no longer used for video playback — but it is still extremely important.

RTMP is mainly used for stream ingestion.

Meaning:
Sending video into a streaming platform.

Where RTMP Is Used Today

When a streamer goes live on:

• YouTube Live
• Twitch
• Facebook Live

Their software (OBS, Streamlabs, etc.) sends video via RTMP.

RTMP is perfect for:

• Sending video from camera → platform
• Stable, continuous upload
• Low latency ingest

But it doesn’t scale well to millions of viewers.

So platforms convert RTMP streams into HLS/DASH for delivery.

Modern Live Streaming Pipeline

RTMP is the input door of streaming platforms.

4. 4 WebRTC (Real-Time Communication)

WebRTC powers real-time communication apps like:

• Google Meet
• Discord
• Zoom
• Microsoft Teams

This protocol is designed for ultra-low latency.

Typical latency: < 500 milliseconds

Why WebRTC Is Different

Unlike HLS/DASH:

• Uses UDP instead of HTTP
• Supports peer-to-peer streaming
• Optimized for real-time interaction

This makes it perfect for:

• Video calls
• Gaming streams
• Interactive streaming
• Virtual classrooms

But WebRTC does NOT scale well to millions of viewers.

So it’s used for small group communication, not Netflix-style streaming.

4.5 SRT (Secure Reliable Transport)

The rising star of live streaming

SRT was designed to solve a real production problem:

Sending high-quality video over unreliable internet connections.

Used by:

• Broadcasters
• Remote production teams
• Live sports contribution feeds

Why SRT Was Created

Traditional live broadcast used expensive satellite links.

SRT allows broadcasters to use the public internet while maintaining:

• Low latency
• Packet loss recovery
• Encryption
• High video quality

It’s perfect for sending video from:

• Stadium → Production studio
• Field reporters → TV station
• Remote events → Cloud

SRT is replacing satellite and expensive fiber links.

Putting It All Together

Modern streaming platforms use multiple protocols together.

Example: Live YouTube stream

1. Camera sends video via RTMP or SRT
2. Platform transcodes video
3. Viewers watch via HLS or DASH
4. Video calls use WebRTC

Each protocol solves a specific piece of the streaming puzzle.

The Big Picture

There is no single “streaming protocol.”
There is an entire ecosystem of protocols working together.

And that ecosystem powers everything from Netflix nights to Zoom meetings to live sports broadcasts.

References :
https://dev.to/varungujarathi9/introduction-to-video-streaming-h92