So where is h.265 anyway?
Firstly, what is h.265?
Well, it’s a compression algorithm. It’s the successor to the immensely popular h.264 codec, which in turn was the successor to h.263 and so on and so on. h.264’s current popularity is immense, and it’s come to drive probably 90% of video made, shared, broadcast and downloaded. It, sadly, has one terrible problem.
It isn’t very efficient.
h.264, as well as all the ‘h’ codecs are based on M-JPEG technology…sort of. If you’re not familiar with the way JPEG compression works, I suggest some google fu. It’s incredibly complicated, but does a good job of reducing file sizes without destroying too much image quality. It might be hated by professional photographers, but JPEG drives images, including this site. I’ve talked about the problems of compression before, so I’m not going to go into that.
h.265 also has a massive problem. It might be a lot better at cramming more into a smaller space, but that extra efficiency requires a lot of extra power. With a lack of hardware en and decoding, playing back h.265 is hard enough, let alone (god forbid) trying to encode to it. At first, it seemed that only “top of the line” CPUs could play back the new codec, and that has been an issue in its introduction. However, the popularity of h.265 is gaining traction. It powers 4K on Netflix, can be run on *most* Android phones running Android L (it definitely works on a Nexus 5), and software optimisation has made playing back h.265 in VLC easily doable, but it does drain your battery faster.
h.265 has been around officially since 2013. It’s not a long time, but for a technology that promises such rapid changes to the industry, the lack of progress seems rather backward. In ‘pro’ video circles, h.265 encoding is still looked down upon. It’s not accepted by most (if not all) editing software, so has to be transcoded beforehand, and whilst it may save a lot of space on a hard drive, it requires a phenomenal amount of processing power to run.
One of the best analogies of compression I ever heard was about apples. Your challenge is to fit 100 apples into a crate. You could just pour all 100 of them in, but some will become damaged and inevitably a few will spill out of the crate. Alternatively, you could slowly (and carefully) place each apple into the crate, to fit as many as possible, using all available space. Your third option is to turn them all into juice, and then pour that into the crate. As such? Less compression = apples. More compression = apple juice.
More accurately perhaps, your challenge is to fit 100 apples composed of 10 different varieties into a crate. Some of the apples will be very similar, and as such many of them can be replaced with a piece of paper that reads “#apple 12 looked like apple 6”. The important thing about h.265 is that it can do this faster, and better. It’s a mixture of apples and apple juice that can more accurately be reconstructed.
h.265 is also essential for the future of broadcast television. In the UK, a standard terrestrial channel gets around 19Mb/s of bandwidth, but this can be subdivided into several smaller channels. Satellite providers have more frequency to play with, and as such each ‘channel’ (which can include subdivisions) is around 25Mb/s. The critical issue here is 25MB/s (more like 11…) is not enough for h.264 MPEG-2 broadcasts like what are currently used. Whilst h.264 4K video can be compressed down to 11Mb/s, it requires too much processing power to be done reasonably well in real time.
The lack of perceived roll out h.265 feeds a content issue. Very little 4K content is being made, because there are very few ways to show it. Consumers are then less interested in buying 4K TVs, because there is no 4K content.
The rumours here are that Sky will launch a 4K set top box and a few channels at some point this year. Whether this will actually happen or not, remains unknown.