Home » Content Delivery » Understanding OTT: The Basics

Understanding OTT: The Basics


Introduction – The future is software-defined media supply chains

Broadcasting and the technology that underpins it is in the midst of profound change.

Purpose-built facilities are reaching end of life. There is no longer a need to build expensive rooms full of racks of equipment that serve just one purpose. Systems built with software are transforming the economics of the industry.

To quote John Honeycutt, Chair IBC Council, who pioneered the migration of linear television to the cloud as the CTO of Discovery, “Infrastructure used to be an asset – now it’s a liability.”

Legacy distribution systems are infrastructure-centric and hardware-based. Trying to cantilever them to new requirements isn’t really progress – it’s trying to extend the life of something that should be evolved.

We need technology that looks far beyond today. That technology is software-defined media supply chain and is at the heart of the OTT revolution. OTT, short for Over The Top, is the modern way of delivering content to consumers over the internet, bypassing traditional broadcast, cable, and satellite to home delivery. OTT is built on scalable and adaptable web-centric platforms that can only be achieved using software-based systems.

This ebook explains what that means and how OTT works.

Why change at all?

What is the imperative to change decades of highly reliable infrastructure and working methods?

Essentially, because consumer expectations have moved on and at a pace that current systems cannot match. Businesses will fail if they do not adapt.

Since television began it has been a one-way delivery. Broadcast means the transmission of content from a single point to all possible endpoints within reach on the network whether that is by satellite or cable.

Sure enough, cable TV revolutionized content choice – cable stations could be aggregated in a cable headend or bundled into a channel bouquet distributed through a coax tree ultimately to a TV via a set-top box. Some home recording (for playback) became possible and the number of channels exploded but the bulk of programming remained scheduled, pushed to the viewer, with next to no chance for the consumer to interact.

The hardware-centric nature of the infrastructure meant production workflows remained static. This has defined generations of TV production, transmission, and consumption.

Consumers now expect a multiple choice of content and to choose the time and method of its consumption. They want to watch what they want, where, when and on what device. There’s also the expectation of upgrading the experience to new formats including HD to 4K to 8K, High Dynamic Range, and superior audio. Innovations like these will not stop being developed and consumers expect the latest tech to be in their hands.

The problem is that legacy systems are increasingly insufficient to meet these requirements. The broadcast model uses specialized technology, designed for a purpose and unique (therefore expensive) to our industry. It is too rigid to react to consumer expectations at anything like the speed required for business success.

Especially when there is an alternative.

Overview of OTT as a distribution method

The internet offers resources for distributing TV and movie content without having to reinvent a lot of wheels. Its core infrastructure is agnostic to any computer system – that is, it is not dependent on a very specific kind of infrastructure other than a generic one.

The internet is software-based from end to end and using software we can build new applications quickly to match the pace of market demand. It is futureproof.

The internet gives consumers choice – including the ability to swap and change their content providers.

Programming can traverse the internet over the top of existing and legacy distribution – hence the acronym OTT – and provide consumers access to a vast array of live and on-demand content sources. It can be monetized in several ways (free ad-supported television – FAST, advertising-supported video on demand – AVOD, subscription video on demand – SVOD, and pay as you go transactional video on demand – TVOD are the prime models). OTT distribution is geographically diverse and regionally aware allowing new business opportunities by targeting content and ads with granularity as fine as the individual viewer. Importantly, OTT is a pull model. Consumers request the content they want; it is not broadcast to them.

These are things that legacy infrastructure cannot do.

Adapting legacy to OTT

The move to OTT from legacy is an evolution. In part, this is because there are expectations around the delivery of content built on decades of traditional broadcasting which OTT has to match.

For example, broadcast distribution is by protected private content highways which are tightly managed and highly secure. There are standards for media, and transmission formats as well as over-the-air regulations which guarantee services to consumers.

Since broadcasting is deterministic, we need new technologies that create delivery determinism on the internet. Pushing content does not work on the internet; we need pull-based mechanisms for transport.  We need to take the best of breed from today – everything we need to keep business going – and move forward with it.

Luckily, the internet has some resources on which to build these services so we don’t need to reinvent wheels.

What standards can we re-use?

We can reuse a number of existing standards to maintain business continuity. These include:

MPEG2-TS – A robust carrier for media, MPEG2 transport streams are self-describing and self-synchronizing. Originally developed to packetize and carry MPEG2-encoded video, MPEG TS today is used to carry almost any compressed video streams regardless of the underlying compression codec.

Video image formats – Resolutions, framerates, and color volumes are already well established and supported by consumer devices. High Dynamic Range (HDR) formats vary too (DolbyVision, HDR10, PQ are among its flavors).

Video compression – MPEG-2, MPEG-4, HEVC are well established with more compression schemes emerging capable of higher efficiency and better performance (AV1, VVC)

Audio formats – similarly there are established schemes for compressing audio signals which can be carried into OTT. These include AAC, MP3, MP-4A, (E)AC-3, DD+ /Atmos

Unlike broadcasting, the Internet isn’t driven by hard ISO-compliant standards but request for comment (RFC) documents. There are several rigid requirements, established protocols, and evolving technologies we can lean on.

First among these is HTTP, the foundation of data communication for the World Wide Web.

How does OTT transport work: Overview

OTT distribution is based on the concept of packetization – taking a data stream and chopping it up into pieces or “chunks” for ease of transport using ubiquitous web browser protocol HTTP.

However, if we try and push packets across the internet just using HTTP the result will be latency (delay), packet loss, jitter, re-ordering.

Consequently, OTT uses HTTP over TCP/IP (Transport Control Protocol) which uses acknowledgments to guarantee packet delivery. This means that if a packet is lost it is retransmitted. In normal everyday operation packets are getting lost all of the time but TCP takes care of the problem.

Unlike broadcast transport streams where content is pushed to users, this web-centric model means content is pulled by end devices. When user’s watch content OTT they are making a request and the device browser starts to pull that content to the device.

It’s worth reiterating the key difference between broadcast/TS and OTT. With broadcast TS the sender controls the transmission while in OTT the receiver controls the transmission process. In broadcast, there is an engine creating an MPEG TS with very precise timing, a process that consumes a lot of CPU cycles just to create a perfect cadence. In OTT, we’re merely hosting media chunks as small files on an FTP server waiting for them to be pulled by the viewer’s OTT player; there is no ‘sender’, just a file transfer. This enables immense scale and significant cost reduction to originate the feed for each recipient, as the OTT player itself stitches these chunks together to create the properly timed cadence of pixels and video frames.

Browsers can run application code. That means they can run things like Video Players. SVODs are likely to run their own Player so that they can have a dialogue with their customer (the customer can browse the SVOD library, and the content service provider can deliver targeted ads to them).

OTT transport is based on file transfer of chunks of media. OTT also allows multiple versions of content to be made available simultaneously to each player. The player then decides which version is the most appropriate one and can switch between versions dynamically to adjust to changing network performance.

Ads can be inserted at any point in the path from the headend to the Player. In broadcast, supplying a different ad to each recipient is uneconomical as it requires creating a separate MPEG TS output for each recipient. In OTT, the ads are simply a different set of file chunks that hold the ad content that the player requests and the player seamlessly switches between the program chunks and the personalized ad content.

To learn about OTT Reception, wrappers and protocols, packaging, adaptive bit-rates, and receive the full OTT dictionary- download the full guide here.

Do NOT follow this link or you will be banned from the site!