Content concepts.

My favourite part of attending a conference is learning all the new buzzwords. These are from last week’s DTG Summit. blog.mindrocketnow.com

Every industry has its own peculiar patois, but it seems to me that the content industry tries harder than most to come up with pithy phrases. At first glance, they elicit a reaction of ‘groan – that’s doublespeak nonsense’. But sometimes – only sometimes – a closer examination yields interesting insights. Let me try some out on you.

Content Infidelity: In 2013 Netflix released a survey that said, “51% of those in a relationship would “cheat” on their spouse/partner/significant other by streaming a TV program(s) they agreed to watch together before their partner had a chance to watch it”. (Its survey of 2017 went on to find that 14% think content infidelity is worse than having an actual affair.)  

So what? Such a high percentage tells me that watching TV is a fiercely social activity, and still a centrepiece in our everyday lives. It feels comforting work in such an impactful industry.

Work, Sleep, Stream: A survey by Rovi in 2015 found that watching TV was the third biggest daily time commitment, after working then sleeping. (Actually, it found that there was some overlap between working and watching TV, but that’s a different point entirely.) 

So what? TV is so important that some chose to prioritise it over all other activities. TV continues to take the lion share of leisure attention and therefore spend.

Show Dumping: This is a relatively new phenomenon, identified by Tivo in 2016 where viewers disengage from a show because it becomes too difficult to watch. This might be because it moves from one streaming service to another, especially if it goes from free (iPlayer) to subscription (Netflix). Or if it goes from a subscription you already have to one that you don’t. 

So what? It seems that content isn’t king, but economics is.

Path to Pay: As we’ve seen, the only growth area in Pay TV is in low-spending households. These are the ones who normally enjoy Freeview or iPlayer or shows recorded on their hard drive. All broadcasters are looking for a way to encourage these free viewers to become pay viewers. 

So what? The innovation in the UK market over the last year has been in this low ARPU end of the market, and focus will continue to be here. Existing products like Now TV, Amazon Fire TV, Netflix, Freeview Play will get better. But as a direct consequence, products like Sky Q and Virgin Media VIP will become more expensive and irrelevant. This is all good news for the viewer.

Skinny Bundles: Presumably due to the restrictive rights negotiated by Google, YouTube TV is launching with a mere 42 channels. This looks waifishly skinny when compared with the “over 260 channels” as boasted by the likes of Comcast. 

So what? These 42 channels are the ones that viewers actually want to watch. In other words, 85% of Comcast’s channel line-up is of no interest to a viewer. It should be acknowledged that there are significant issues with the rights that Google negotiated, particularly the reluctance of broadcasters to relinquish control over which ads to show to Google. Nevertheless, to misquote Andrew Neil's keynote, I hope we’ll look back on 2017 as the last time we got away with forcing viewers to buy 85% more stuff than they actually wanted.

Net Neutrality by Necessity: The growth of online video continues to be explosive. ISPs are struggling to maintain the infrastructure investment to keep up. But just as it’s ludicrous to expect to pay different electricity bills for usage by a laptop versus usage by a microwave, it’s equally illogical to have anything other than net neutrality.

So what? Viewers are moving away from delivery of TV through aerials on the roof, to delivery of TV through home Wi-Fi. The industry is having to welcome a new entrant to the TV value chain: the Internet Service Provider. The ISP has a key role in providing the broadband connectivity into the home, and usually the Wi-Fi router that flings the Internet around the home. As they did with the owners of the DTT transmitter network or satellite fleets, it’s in the broadcaster’s interest to strike a commercial relationship with ISPs. However, it’s not in the ISP’s interest to strike individual commercial relationships, not least in Europe due to legal obligations not to discriminate service based on content. ISPs will realise this once they accept that they’re nothing more or less than a fundamental utility. Which will leave broadcasters floundering to find their place in the new value chain. 

Prop-Up Programming: Even in this age of multi-channel multi-format multi-delivery fragmented viewing, some programmes still get over 10M viewers. These hugely popular programmes prop up the channel, the brand, the technology platform, and the broadcaster as a whole. However, these programmes are becoming fewer. 

So what?  If ITV didn’t have to serve 10.4M viewers with Britain’s Got Talent simultaneously, they could abandon the expensive DTT and DSAT carriage agreements, and deliver directly over the Internet. However, if ITV didn’t have any programming that reached 10.4M viewers simultaneously, it wouldn’t command the same ad revenue, and wouldn’t be able to afford to deliver its services. A neat little circle that props up the status quo, but looks more fragile as the number of prop-up programmes decrease.

(Full disclosure – I made the last two up myself and added them to the list, hoping that they’ll gain credibility by association.)

Thoughts from the DTG Summit 2017.

The broadcast industry is catching up with the world around it. It’d better hurry up. blog.mindrocketnow.com

The DTG Summit is one of the smaller conferences, catering to the UK broadcast community. Even though it’s small, there’s a good mix of broadcasters, technology vendors, service providers, and regulators, which usually makes for a good spread of perspectives. This year, they all seemed to converge on a couple of ideas: 1) IP is a good thing; 2) Netflix and Amazon have a lot of money.

A climate of fear

However, both of theses ideas were expressed in a way that fed a pervading sense of fear. Fear of being forced to spend money on newer shinier kit, fear of other entrants taking market share, fear of losing revenue, fear of the world moving on whilst broadcasters ossify.

These are all real, legitimate fears, founded in established market trends. For example, the only revenue growth in Pay TV appears to be from the lower spending homes. For example, the most popular UK TV content is still free, so there is a large constituency of TV viewers who are happy with not paying  – even if it means missing out on the latest features – so there is a large legacy viewership to keep happy. For example, investment cycles by TV operators is still around 5-7 years, compared with software app cycles of 1-3 months, which makes it very hard to TV technology platforms to keep up with best in class software.

Disruptive effect of IP everywhere

Broadcasters have been late to implementing IP end-to-end because they’ve been trapped in their decade-long investment cycles. Other sectors like banking have come to grips transforming their infrastructure to highly available, highly secure, highly flexible, all the while reducing operational costs – through implementing IP everywhere. (The banking sector has spent a huge amount of money to get here, but the point is, they are already reaping rewards.) The broadcast sector is just starting to link its digital islands, and is finding it very hard and expensive to do so.

Whilst IP everywhere will undoubtedly lead to improved operational efficiencies, that isn’t the disruption being caused in its wake. The disruption is a consequence of the cost. The irony for me is exactly what the broadcast industry finds expensive and hard, has actually made it cheaper and easier for new entrants to come into the marketplace. The barrier to entry into the broadcast space has been lowered because: you don’t need as much scale to be economic since the incremental cost of IP delivery is negligible; new opportunities can be trialled without needing to underwrite/write off large development costs; geographic location is truly irrelevant opening new opportunities for resource arbitrage.

To put it another way, newbies can find value with IP everywhere in not only doing new things, but also doing existing things because they can do it cheaper than broadcasters. The solution for broadcasters appears simple to me: get on the IP everywhere train as soon as possible to remove the cost disparity and to find new value opportunities. It’ll cost a lot to implement, but it’ll cost a lot more to go slowly out of business.

Disruptive effect of new money

Despite a pervading air of fear, the conference never fails to self-congratulatory over UK’s talent in media and entertainment. UK broadcasters may not have the wherewithal to spend as much as Netflix on new content, but at least every villain is British. UK broadcast punches above its weight.

My problem with this contention is the aspiration to spend as much as Netflix. For a start, Netflix and Amazon together now spend more on creating content than all the Hollywood studios combined (around $12B). We don’t feel the need for BBC or ITV to be as big as Disney, so perhaps we shouldn’t feel threatened by Amazon.  My other problem is the implicit correlation between spending money and creating quality programming. Whilst it’s true that the more money you spend, the bigger the sets you can build. It’s also true that there is plenty of quality TV to be found in niche broadcasters like Channel 4.

For me, the disruption isn’t the fact that there are new entrants with deep pockets. The disruption is the fact that these new entrants have ability to move the market. For example, Netflix is pioneering the use of data and analytics in its delivery of content. For example, Amazon has implemented voice search successfully for TV (and I think this will be the interaction mechanism of choice with TV viewers within a couple of years). For example, Netflix launches entire series in one go, a practice that is now common amongst all providers.

It’s also a fallacy that these new entrants have a money-no-object approach to new technology. They have deep pockets because they’re successful, not because they’re profligate. It’s this success that we should emulate, not the spending.

A future of opportunity

These musings may sound gloomy, but I’m actually very upbeat. I came away from the conference with a renewed sense of confidence. There’s a key advantage to being a second entrant into a market – others have created that market space for you. There’s already a clarity of vision, and someone else has shown all the steps (and missteps) to getting there.

Broadcasters can learn from the many lessons from other industries, and from first movers. Learning these lessons should reduce the inevitable cost and effort required to re-platform an entire industry onto IP everywhere. Learning these lessons will help to create industry-leading content without industry-leading spend. The opportunity is to learn from the pioneers and to do it better.

I’m not ready for the 4k future.

Ultra HD is a confusing confection of standards, and is getting in the way of watching amazing 4k TV. blog.mindrocketnow.com

I want a new TV. It’s not that my current TV isn’t any good – the pictures are crisp, and the screen is large enough – it’s just that the new sets are even shinier. And the newest of the new TVs have the next super-duper feature – Ultra High Definition. Not just HD, but ULTRA. Sweet.

The problem is that the UHD future is still very uncertain. There are competing and non-interoperable technology decisions that will impact upon which TV I should choose – and getting the decision wrong could mean that I won’t be able to watch the programmes that I want to.

It’s content that drives behaviour that drives technology. This lesson has been applied with variable conviction through 3D, HD, widescreen, and NICAM stereo. In the UK today, Netflix, YouTube and Amazon Instant Video are already streaming content in 4k. BT has announced BT Sport UHD. The new UHD-capable SkyQ box is strongly rumoured to release before Christmas. Sony provides a 4k media hub – essentially a NAS pre-loaded with Sony 4k movies (and the Playstation original Powers) to provide content for your new Sony 4k TV. Samsung uses the same approach, but without its own movies. So the content is starting to emerge, to push the technology.

The first issue for the early-adopting consumer, is that UHD content will be limited and expensive. Netflix 4k needs a £8.99 per month subscription and the catalogue is limited to some flagship content from Netflix (House of Cards), AMC (Breaking Bad) and Sony back catalogue movies (Jerry Maguire). YouTube has a 4k quality setting and some 4k videos uploaded (beautiful pictures, chillout music, but very jerky and makes my laptop crash). Amazon Instant Video has some Amazon originals 4k content included in the prime subscription, and other back catalogue from Sony available to rent at £6.99 or buy at £18.99. You can bet the subscription hike from BT and Sky for 4k will cause a sharp intake of breath.

Once you’ve committed to spending on 4k, the second issue is that it may not play. To understand fully why, we need to unpick what UHD actually is.

UHD bundles together a whole set of advances in TV technologies. This bundling of advances makes UHD such a significant step forward. But because these are all individual steps, with no mandated correlation, there’s no guarantee that UHD means the same thing to all broadcasters and manufacturers.

First is the resolution: UHD is 3840 x 2160 pixels, which is four times the number of pixels as full HD. (Full 4k cinema is 4096 x 2160, which is slightly wider, as is currently the case.) UHD also allows for faster refresh rates, starting at 24 frames per second (current cinema rate), 25 fps (current TV rate), 30, 50, 60 and 120 fps. This promises a level of detail where the eye would have difficulty in seeing individual pixels, with no motion blur – at a cost of 20x current bandwidth.

Next up are the advances in colour reproduction. All current TVs use a mathematical representation of colour based on red-green-blue described by the standard Rec.709. Though this standard brought worldwide agreement on how colours (and white) should be represented, the range of possible colours is less than the range experienced by the human eye. To get close to human experience, UHD mandates the newer Rec.2020 standard.

To be able to describe colour more accurately, UHD allows for spending more bits to describe colour – 10 or 12 bits per sample rather than the current 8 = 1.5x current bandwidth. These are either sampled twice as frequently for brightness (luminance) than for red, and half again for blue (4:2:0), or twice as frequently for brightness than each of red and blue (4:2:2), or same frequency of samples for all three (4:4:4). 4:4:4 requires 2x the bandwidth as 4:2:0.

Brightness is also described more accurately using High Dynamic Resolution (HDR). HDR doesn’t necessarily increase the number of bits used to describe brightness, but how. Instead of the current mathematical Electro-Optical Transfer Function, a Perceptual Quantisation is preferred. This enables describing a wider dynamic range.

TVs currently go from 0.05 cd/m2 at black to 120 cd/m2 for full brightness.

The best-performing OLED TV goes from 0 cd/m2 to 800 cd/m2. HDR enables video to be described to take advantage of this available dynamic range – and more. The leading contender specification for HDR, Dolby Vision, claims an extra 20% bandwidth would be needed for metadata.

So we see that UHD = resolution + frame rate + colour gamut + sample rate + sub-sampling schema + HDR. Total bandwidth uplift would be 72x current bandwidth! Thankfully, there are new codecs that significantly reduce the bandwidth requirements. Both H.265 (HEVC) and Google VP9 promise 40% over current HD codec H.264, which was 40% improvement over MPEG2 as used in SD DTV broadcast. So use of HEVC or VP9 could reduce bandwidth uplift to 4.5x, but that still means bit rates of over 30Mbit/s.

However, the main reason why your content might not play is DRM. Studios are following the same tired old path of maximising the money they wring out of broadcasters and the public by the implementation restrictive DRM. However, there are a few contenders out there. Netflix and BT mandate Microsoft PlayReady. There are also solutions from Verimatrix and Nagra. Each DRM schema must be implemented in hardware (the current market leader is Broadcom with their Sage system-on-a-chip), so each DRM schema requires a different STB or TV.

The Samsung TV that is capable of playing Netflix 4k may not be capable of playing the UHD content from AcmeFutureTV if the DRM is different. And the studios focus heavily on DRM; there’s no guarantee that each studio will choose the same DRM vendors to approve. So different content may need different DRM and therefore a different STB. (I hope I have enough inputs on my TV.) The equation is now: UHD = DRM + codec + resolution + frame rate + colour gamut + sample rate + sub-sampling schema + HDR.

So before I buy my next TV, I should ask myself: is the TV full 4k or UHD1 resolution? Does it have a Broadcom Sage chip? Is it using the right DRM variant of the chip? Will I be connecting to the TV using HDMI 2.0? Does the output of the for the content STB support HDCP 2.2? Does the TV support Rec.2020 colour space? 10-bit/ 12-bit colour depth? 4:2:2 or 4:4:4 chroma sub-sampling schema? Does it have sufficient dynamic range to reproduce HDR? Does it have a frame refresh rate of 60 fps or even 120 fps? Does the player within your STB or smart TV support the HEVC or Google VP9 codec? And now that I’ve answered these questions for the TV, was the movie that I want to watch encoded with the right combination of PlayReady/ Rec.2020/ 12-bit/ 4:4:4/ 120fps/ HEVC?

Why should I, as a consumer care, when all I want to do is to watch a movie?

With all this variability in what makes up UHD, it’s understandable that some broadcasters are being cautious with their investment. They need to ensure that their network infrastructure can cope with additional bandwidth demand, invest in head ends to process content provider feeds into coherence, and invest in STB because Smart TVs are unlikely to be smart enough to cater for their particular combination that makes up UHD.

This caution allows OTT providers to lead the way with 4k. With no network investment to worry about, Netflix, YouTube, Amazon, Sony et al can focus investment on content and hardware. The approach is different in each case: Netflix has bet its farm on its tie-up with Broadcom, hoping that everyone will put the chip in their TV or STB to get Netflix, and it looks to have been a wise bet. Amazon develops its own STB, Sony its Smart TV and Playstation. YouTube does not have studio content licensed for 4k (though GoPro's Adventure of Life in 4k is arguably commercial content but with a free license) and is agnostic on hardware. 

Future-proofing my next TV purchase will probably be impossible. The panels to support Rec.2020 and Dolby Vision just aren’t made yet. However, 4k is gathering commercial momentum, so realising the full UHD1 spec in a TV will be irrelevant if I actually want to watch 4k content. I think it’ll be enough to have a 4k panel, and have lots of HDMI 2.0/ HDCP 2.2 connectors. However, I think I’ll have a cluster of STB around the base of my TV. It seems there’s no getting away from a rat’s nest of cables behind the TV.