What’s streaming: how your favorite content is delivered to you. Part 1


Pavlo Kalmykov, Product Architect at Pluto TV

February 10, 2022 Pluto TV, streaming, video streaming

Hi, I’m Pavlo Kalmykov, a Product Architect at Brightgrove and Pluto TV, a free streaming platform with 50+ million monthly active users. I’ve been working with video streaming for 4 years.

 

Online streaming accounts for 80% of all current internet traffic, making streaming a very profitable area of development expertise. As a Product Architect at Pluto TV, I’d like to share my key streaming insights with you.

 

Today I’ll cover streaming in different contexts and key modern streaming mechanisms. This article will be useful for people just getting into streaming or those who want to know more about streaming specifics.

 

If you know about the streaming mechanism, it’s easy to follow even how live footage from Dodger Stadium appears on millions of screens around the world. To understand streaming technology better, we’ll begin with its history.

History of streaming

First, let’s look at the definition of streaming in the digital field. It’s a technology for transferring data that can be received and processed in a steady stream. 

 

The first known example of streaming is smoke signals used for data transmission at a distance. Any person who sees those signals is considered a receiver, and if the person understands the meaning of the signals—an interpreter. As you can see, streaming exists even without digital technology. Now you can reach your team lead even on a deserted island vacation if he’s the only one who can fix a bug 🙂 

 

In the multimedia context, streaming is a continuous flow of content delivered from a source and consumed continuously. There’s little to no intermediate content storage or network elements in streaming.

 

The telegraph was invented in the 18th century and was the first multimedia streaming prototype. However, the first actual audio signal transmission with subsequent reception and playback occurred in 1900. It took place thanks to Reginald A. Fessenden, who’s responsible for the first two-way transatlantic radio transmission. 

 

Let’s move to the closest thing to media streaming as we know it—television. TV broadcasting began in 1920. Audio and video signals in analog format were transmitted over radio waves. Each radio station occupied a specific frequency to transmit those signals. We know those frequencies as TV channels. 

 

Frequency interference exists in analog TV causing lowered content quality: changes in color, brightness, and sound quality. And just like radio signals, analog TV transmission gets worse the further away you are from the signal source. 

 

That’s why digital television came into the picture in the late 90s. The main difference between analog and digital TV is the data transmission format. In digital television, the signal is transmitted as compressed data in packages. 

 

Digital TV content doesn’t suffer such quality loss, and users can enjoy a picture of the highest quality. Digital transmission is also more efficient than analog and has only about a quarter of the signal bandwidth difference compared with an analog TV channel

 

However, a special tuner is necessary on the receiver side to decode digital signals. Old analog TVs can only broadcast such channels with such an add-on tuner, which limits the technology’s availability. 

 

The digital signal range is way better than analog, and it doesn’t suffer such quality loss, yet it is still finite. Satellite dishes came to the rescue to deal with the distance problem. But the satellite craze slowly turned into dish oversaturation, and cable TV was the solution. It’s basically one dish that distributes the satellite TV signal to multiple users in one or several buildings via wire.

 

Let’s finally move to streaming in the context of the internet. We’ll take an in-depth look at the types of such streaming and the technology behind it.

Real-Time Messaging Protocol (RTMP) streaming

Adobe created RTMP in 2002 as a method for streaming video and audio over the internet. RTMP and Adobe’s Flash Player were the leading live streaming mechanisms until the early 2010s. These two support instant video delivery with around five seconds of latency when used together.  

 

RTMP uses a continuous TCP connection to stream audio and video fragments from a source to a single destination. There are two main use cases for this protocol. The first is to transmit video between an encoder and server, known as first-mile delivery or video contribution. The second is to transmit video between a server and a viewer’s device utilizing Flash Player, known as last-mile delivery. This one works best for streaming to a small audience from a dedicated media server.

  

First-mile vs. last-mile RTMP use

 

However, receiving and playing video with RTMP is no longer supported by many endpoints, for example, iOS devices. Furthermore, it will soon be deprecated as Adobe has announced the end of Flash Player. 

 

Broadcasters still use RTMP for live streaming to their servers and then transcoding the content for delivery to the end-user. In other words, RTMP streaming is alive and well for content contribution, but not last-mile delivery. HTML5 streaming, open standards, and adaptive bitrate delivery eventually replaced RTMP in last-mile delivery.

Progressive download or HTTP Pseudo Streaming

Progressive download is essentially downloading a video file from a server using the well-known HTTP protocol. It’s similar to HTTP Live Streaming, an adaptive bitrate streaming protocol developed by Apple. 

 

The playback is a result of moving the video’s metadata from the end to the front of the file. This action gives the media player all the information necessary for playback while the file is still downloading.

 

HTTP Pseudo Streaming overview

 

The main disadvantages of HTTP Pseudo Streaming are reduced security and longer loading times for larger files.

Adaptive streaming

Adaptive bitrate streaming is a technique to stream multimedia over computer networks. It’s a mechanism based on traditional HTTP, and it supports variable quality streaming based on the receiving device capabilities. 

 

In the past, most video or audio streaming technologies used RTMP or RTP with RTSP streaming protocols. Adaptive streaming is mostly based on HTTP and works best over large distributed networks such as the Internet. This streaming type adjusts the media stream quality in real-time based on the user’s bandwidth and CPU capacity. 

 

With adaptive streaming, the video is transcoded into a configured video size, split into fragments, and uploaded to the server. Afterward, the video is available for download, often via a caching content delivery network (CDN). The server only issues the stream description manifests to the user’s video player. In this case, we still have control over each user stream without handling the distribution of media files by the server.

 

Adaptive streaming overview

 

There are many adaptive streaming protocols, but the main ones are: 

  • MPEG-DASH 
  • Apple HTTP Live Streaming (HLS) 
  • Adobe HTTP Dynamic Streaming 
  • Microsoft Smooth Streaming 
  • QuavStreams Adaptive Streaming over HTTP 

 

MPEG-DASH is growing its market share due to its greater efficiency, but most distributors still use HLS.

Now you know how streaming originated and evolved and the main modern internet streaming types. In part 2, we’ll look at streaming latency and what causes it, digital rights management, live event broadcasting mechanism. I’ll also cover different streaming monetization tactics and how to implement them.

RELATED CONTENT

Discover more about our experiences, victories, struggles, and best practices