21 Jul 2014

Media Devices for India

… and other developing countries.

(Disclosure: I work for Google, but not anything related to media.)

There’s a lot of focus on media streaming devices like the Roku, Apple TV, Amazon Fire TV, Chromecast, etc. This may work in developed countries, with reasonable Internet speeds, but not in developing countries with slower speeds. For example, the average speed in India is 1.3Mbps, compared to a global average of 3.8Mbps, with 10Mbps in the US.

Not only is the bandwidth in India low, but the latency high. A 1mbps connection in the US is some ways better than a 10mbps connection in India. Finally, Indian internet is also flaky, with the power going out all the time, low-quality routers that need to be rebooted multiple times a day, low-quality ISPs, etc.

In such a situation, you can’t stream, not even SD in some situations, to say nothing of HD.

Further, downloading movies is slow and hassle-prone (you have to keep and eye on it and retry if the connection breaks). It prevents anyone in the home from having usable Internet access during the hours it’s in progress, and it consumes a good fraction of your monthly quota. Mine, for example, is 40GB, of which a single high-quality 1080p video can eat up 9GB or so. Given these realities, it makes sense to keep a local copy forever once you download something. Storage is cheap, and bandwidth is expensive. Tons of people in India have a 1080p-capable TV or other display, and enough storage, but hardly anyone has the bandwidth to stream high-quality 1080p content.

This requires a model where you download ahead of time [1], and keep it forever. Like BitTorrent.

I’m not averse to paying a reasonable price for content. For example, I wanted to buy the Bluray version of Star Trek: The Next Generation, and I was prepared to pay a fair price of $100. This is ₹6K, which is way more than the average Indian would be willing to pay. For context, the average monthly income in India is ₹ 7 - 8K. But Paramount wanted me to pay around $700. And the discs are region-coded to the US and won’t play on an Indian Blu-ray player (even if I had one). Is it any surprise that when the studios charge exorbitant prices and impose unreasonable conditions, people turn to piracy?

Given that the content owners have unreasonable expectations, you can’t design a player to support only licensed content. It should support pirated, local content as a first-class citizen, in addition to streaming. Or, to put it differently, to decouple content from the device. This is, in fact, the norm in the computer industry — my phone plays music purchased from legal sources like Apple and Flipkart as well as music that’s pirated. The phone doesn’t know or care how you got your music. Similarly, a media player must support playing content from local sources like USB hard discs and network shares [2], in addition to streaming, if it is to be usable in developing countries.

Another advantage of downloading content ahead of time is that it can use the network when it’s less in demand, like overnight. This also benefits the other users on the network, who’ll get higher speeds, and the ISP, who’ll find that their network can deliver more data per day without a greater investment from their side.

In fact, the idea of downloading content ahead of time is also applicable to developed countries, who can watch 4K if their connection supports only 1080p. Rather than complain that there’s no distribution medium that can handle 4K, let’s use the infrastructure we do have to support it.

It’s also sad that no one has figured out apps on a media player. There are many opportunities, like a weather app that geo-locates you and automatically shows you weather. Or a sports channel that let you choose cameras, and to replay something in slow motion from a particular camera angle. Or a talk show that lets you pose questions to an expert over video chat. Or even talking to your family member via Skype or Viber. This is a goldmine of opportunities, but no one has figured it out.

A media device can also double up as a PC. Many devices already support the requisite hardware, such as a video out port, USB to connect keyboards and mice, a network connection, etc. There’s also plenty of CPU power available in embedded devices. My router has a gigahertz processor, and I’m seeing quad-core TVs. There’s no reason a media device can’t run apps, to double up as a PC [3]. Or even just a browser, to function as a Chromebox. In price-sensitive markets, a single device that does away with the need to purchase and keep upgrading two separate devices will be tempting. It’s also tempting in developed countries (who wants yet another box in their home?), but all the more so in developing countries.

In summary, there are many ways to design a media player so that it’s also usable by people in developing countries, which is a huge market. And the best part is that these ideas don’t make the device any less appealing for people in developed countries. Nor does it add to the cost of the device. This is a good opportunity, if only someone wants to take it.

[1] Yes, you do miss out on the instant gratification aspect. But then, no one forces you to download ahead of time. It’s just an option. If your connection can handle SD video, you can watch whatever catches your fancy, on the spot. Or you can download a better quality of video, like HD, for later viewing. It’s just another option, and doesn’t take anything away.

[2] In all formats: H265, H264, FLV, AVI, WMV, M2TS, VOB, MPG, MOV, ISO and so on. Don’t make me do stupid things like convert media from one format to another. I have better things to do. Just throw CPU at it, and solve the problem. On a mains-powered device, this is not a problem, as opposed to a battery-powered device.

[3] If you’re worried that the device may not be powerful enough to handle both media playback and PC use at once, one can have a device that can perform only one of these roles at once. That’s still way better than the status quo.

No comments:

Post a Comment