I think someone should say it out loud. There’s nothing magical, groundbreaking or earthshattering in H.265. and I think that is also the reason why we shouldn’t let it be our next video codec.
H.264 has been with us for a while now. Standardized in 2003 it is now a decade old technology. That’s even before the iPhone days – prehistoric by my standards. And H.265? It isn’t thrilling enough. It goes on with the usual trajectory of this family of codecs – same concepts, just more options and more computational power which enables achieving less bitrate for a desired video quality.
I am not alone in this notion – Dave Michels wrote a great piece on NoJitter, suggesting that H.265 may already be dead – even before it got adopted. The reason? Google. And VPx codecs.
With everything changes around us: social, mobile, cloud and data; I think it is time to switch gears in video coding as well. We need to change our mindset from looking at a video codec as a technology challenge to thinking how we can change the paradigm – of what can be done with a video codec.
SVC, Scalable Video Coding, was a step in the right direction. But it isn’t enough, as it still lays in the realm of technical capabilities and benefits. To that end, I think we should look at some other codecs for assistance. Namely, VP9 and even ORBX.
VP9
VP9 is Google’s successor to VP8. It is targeted to be on par with H.265 in terms of the technical spec, making them interchangeable (the same way that H.264 and VP8 are).
The difference between VP9 and H.265? Free.
H.265 continues with the pedigree of its successors – being highly patented with a large number of praying vendors waiting for their share of the pie.
There’s nothing wrong with codecs and royalties. Just that they prohibit innovation from happening and they make a great barrier of entry. Good for incumbents. Bad for customers.
Even if you want to pay for a codec, it is a complex and tedious task on your to do list. A real headache. And in the world of freemium that we all live in today? It becomes prohibitive.
I’d argue that our future codec has to be FREE. I am not alone with that desire – the IETF suggests the same as well.
ORBX
ORBX is even more far reaching. It is a video codec written in… JavaScript. Sounds weird, but why not?
The benefits here? The dynamic with which a service can decide how a video codec behaves and operates – without a need for downloads or software updates.
It also opens up the door for more developers and more creativity.
-
Our future codec needs to look elsewhere for requirements. It shouldn’t be a continuation in our current trajectory of “increase CPU to gain more quality” – it should tackle the problem from a different angle. And this is why to me, VP9 is more important as a concept – what it stands for – than what its exact spec is capable or incapable of compressing on an ARM chip.
