After more than 40 years of operation, DTVE is closing its doors and our website will no longer be updated daily. Thank you for all of your support.
2024: the year in view
For our final Week-in-View article of 2023, well-known industry technologist Thierry Fautier makes some predictions about the key trends we are likely to see in video distribution technology in 2024.
2024 is set to be a year where many transitions observed in previous years will be reinforced. Let’s take a tour of the different initiatives the industry should work through and try to quantify the impact.
AI
It’s impossible to talk about the future without mentioning AI. OpenAI’s release in November 2022 has opened a Pandora’s box and made AI a reality for hundreds of millions of people.
Up until now we have seen a major focus of AI on production and post-production, but we have not seen it reach its full potential on the distribution side, especially for real-time applications.
In 2024 we should see more AI-based applications in the real-time domain, including live subtitling generation from any language to any language with the same language transcription being compliant with local regulations.
On the LLM side, we should start to see deployments of new services where a user chats with their remote to find the content they are searching for without browsing, and only uses a remote interface for key sports events (goals, penalties, cards). Using AI will remove the complexity of legacy systems, accelerate innovation – and will wow customers.
Public Cloud will be a mandatory tool to support all AI training.
Quantifying the impact is complex as AI will introduce new concepts or replace some un-deployed legacy ones.
Scaling delivery
We have seen in past years the acceleration, especially in the US, of massive live sports events, with exclusive rights owned by hyperscalers (Amazon with NFL and Thursday Night Football, Google with NFL and Sunday Ticket, Apple with MLS, etc.).
Traditional public CDN has been used for all those events. On the broadcast side, with the likes of the BBC trailing the switch-off of DTT in 2030, a scalable delivery scheme needs to be developed. In 2024, we will see the emergence of more complementary ways to scale delivery.
On the caching itself, we will see Edge Caching being more deployed where CDNs will collaborate with ISPs that will host the Edge CDN inside their network, with a private cloud-based approach to offer on-demand capacity aligned with the elasticity required by sporting events. Open caching is one flavour for participating content providers. For others, a traditional CDN caching approach, but with a private cloud flavour, can also be used.
On the ISP side, we have seen already several tier-1 ISPs build their own CDNs – Verizon with Qwilt, Orange with Gcore, Telefónica, Altice, Comcast and LG Uplus on their own – so this trend should expand among other tier-1 ISPs, providing another alternative way to scale delivery.
Multicast ABR, initially targeted for ISPs’ own services, will be more likely adopted when an ISP is able to strike a deal with a well-identified content provider: DAZN and TIM in Italy or Orange in Spain, or BT and the BBC in the UK.
Private Cloud, with a Multi access Edge Compute (MEC) flavour, will be key to this transition.
Quantifying the impact of this trend can only be made when those new workflows are deployed at scale, well beyond 2024.
Streaming cost optimisation
With the explosion of traffic using OTT apps with more and more subscribers (Omdia predicts 1.7 billion subscribers in ’24, meaning more than three billion users) the cost of streaming is becoming more and more a source of financial stress.
New codecs take years to be deployed at scale, and therefore the low-hanging fruit is the use of legacy codec optimisation techniques that can bring about huge savings. We know Context-Aware Encoding (CAE) is the solution to the problem, but we also know there are multiple flavours of this. A solid 100% AI-based solution will be deployed in 2024 on existing codecs (AVC for HD and HEVC for UHD), in addition, other bitrate reduction techniques where resolution and frame-rate can also be changed dynamically are likely be trialled initially, as they might impact device interoperability. Public Cloud is a great resource to optimise VOD encoding, while for live streaming private and public cloud could both be used.
Given that the CDN media market for video delivery is worth over US$1 billion, if only 10% of the traffic could be optimised, this would represent a saving of US$100 million. Don’t expect this to come about overnight though.
FAST
We have seen FAST reach near saturation in the US, especially in relation to the number of channels a human can cope with, the depth of the AVOD library and the ad-available inventory based around sometimes poor-quality content.
In 2024, we will first see the quality of content improve (the ‘churn’ on FAST platforms can be as high as 30% a year). This will attract more users and more advertisers.
Second, new business models will emerge, especially outside the US where broadcasters will be proposing some of their content for FAST. Pay TV operators that have already made the switch to full unicast will also bring FAST platforms onboard.
At the point where the business model and technology meet, beyond existing targeted advertising, we will see content recommendation systems target groups of users who have similar interests, as well as targeted advertising not only based on the user type but also based on the content.
Of course, in the case of pay TV operators, cross-referencing with data collected by the operators will improve targeting, a win-win with supply-side platform companies. Public Cloud is the perfect tool to support a flexible playout/recommendation/targeted advertising scheme.
We hear of leading FAST companies’ revenues already topping US$1 billion a year, so with a 40% CAGR growth, based on an Omdia estimate of FAST ad revenue in 2023 of US$6 billion, we can expect an additional US$2 billions in ad revenue in 2024.
Pay TV 2.0: disintermediation continues
Pay TV will continue its super aggregation journey, taking it to the next level in 2024.
We will see major channels not only offer their live feed, but also their own app that will encompass live, catch up, cloud DVR, xVOD, e-commerce and betting. We have seen deployments and announcements already happening – in Portugal with SIC channel on three different ISP services: in France with TF1 announcing that TF1+ would be on three different ISPs; and in the UK with ITVX deployed on the Sky Q service. Another route that some telcos will take in mature OTT markets like the US is to directly receive ABR feeds from the content providers and get out of the encoding and distribution business. Of course, in the case of pay TV ISPs, given that the traffic is on their network, they will cache the content themselves on their own CDN and therefore improve the user experience and cost of delivery. Content providers will have to decide the best workflow to use to publish (playout/origin) to pay TV operators.
This will be a systemic transformation that will hugely change the physical location of video processing and will have an impact on the encoding market.
Sustainability
I could not close my predictions without touching on sustainability. In 2023, we have seen organisations like Greening Of Streaming set the stage. In 2024, we should see real-world measurements become available to indicate precisely where power is consumed. This could influence the way content is encoded and transported, reopen net neutrality discussions, bring the consumer into the loop by having them choose, for instance, an ‘eco-mode’ at peak time on wired or wireless networks, and, last but not least, force the industry to realise that HD transmission to a UHD screen will use the same power on the decoder and display as UHD itself. This is the biggest element of the cost of end-to-end streaming delivery workflow. Operators need to focus more on how to optimise HDR rendering on UHD TV sets to save power on hundreds of millions of TVs.
Being able to quantify how much can be saved on the device side will come once we know more about overall power consumption within the media and entertainment market. The industry has not seen previous attempts to calculate this as reliable.
Thierry Fautier is managing director, Your Media Transformation.