The MPEG machine is ready to start (again)

MPEG has developed standards that are used daily by billions of people. A non exhaustive list includes MP3, Advanced Audio Coding (AAC), MPEG-2 Systems and Video, Advanced Video Coding (AVC), MP4 File Format, DASH. Other MPEG standards were widely used in the past but their use is gradually fading out, such MPEG-1 Video, and MPEG-1 Audio Layer 2.

This because MPEG is always ready to exploit the latest technology innovations to create new standards that offer new advantages because they outperform previous generations.

At the San Diego, CA meeting JVET, a joint MPEG (ISO) and VCEG (ITU) working goup tasked with the development of a new video compression standard, received 46 proposals from 32 organisation in response to the Call for Proposals issued in October 2017.

The target of the new standard is to achieve a compression ratio leading to a bitrate reduction of at least 50% that of HEVC, including High Dynamic Range (HDR), in addition to providing native support to such emerging applications as 360° omni-directional video.

I am really so proud to say that the tests showed that several proposals in many instances already exceeded 40% bitrate reduction compared to HEVC. Considering the technology power of the MPEG machine and that there are 30 months to the expected time of approval of the new standard (October 2020), there is no doubt that the target will be reached, probably exceeded.

MPEG is always working to provide new benefits for more humans on the Earth. In this case, those still disadvantaged in delivery infrastructure will be able to access services from which they were excluded so far. Others will be able to enjoy more involving experiences.

Of course this will only happen if mindless industry elements will not blow the opportunity again.

IP counting or revenue counting?

In a now distant­ past companies used to be run by engineers. If a company had the right product – backed by good research and developed, indeed, by engineers – people would buy it. Then having a good product was not sufficient and many companies decided that the company had to be run by marketers. Then having a good marketing was not sufficient and so many companies decided that accountants should run them. Eventually many companies were run by lawyers because compliance became the priority. I do not have examples yet (we already have some from politics) , but I expect that soon companies will be run by actors and people will buy products of a company whose brand has been “sold” by a good actor.

I do not think this is the right approach. Of course we do not want as CEOs engineers who, like hammers, sees everything as a nail, or marketers who could not care less of what is inside provided they “feels” the packaging, or accountants whose sole purpose in life is the next quarterly report or lawyers who see everything in terms of compliance or actors who impersonates the company as it if were Othello or Desdemona.

I think CEOs should be the synthesis of all this. CEOs should be able to integratie the functions inside their companies overcoming the downside of sectorial MBOs.

Unfortunately reality seldom matches my beliefs.

Strictly speaking this is “someone else’s problem” (my company is small and I run all the functions I mentioned), but this situation has a strong impact on my alias, the MPEG convenor. The impact is not on the quality of the standards MPEG produces but on their viability when the standards leave the committee.

In MPEG the driving force are researchers: engineers, computer scientists, University professors, entrepreneurs and more. They love talking with their peers as if they were at conferences. Actually, I know they have better feelings because, unlike conferences, they can have memorable battles with their peers and hopefully have their ideas accepted in a standard.

Researchers typically work hand-in-hand with their companies’ IP attorneys. Both are typically rewarded on the basis of “number of patents in standards”. Although not the general rule, the IP attorney position is peculiar in that very often they are close to the CEO who is certainly attracted by the prospect of counting the dollars flowing through future royalties.

But the player in the company who is really concerned by standards is the product department. They need standards, but actually they do not care very much if they contain IP contributed by company researchers.

Unfortunately many CEOs see the product department’s role very much like the captain of a 19th century steam ship saw stokers: they take it for granted that things run. Therefore the value of products as such and its dependence on coal – I mean standards – is not properly represented to the eyes of the CEOs. Product departments are certainly happy to see that their company has contributed a lot of IP to the standards they need, but they would be much happier if those standards were also usable.

CEOs should open their ears not only to their IP and research department heads but also to their product department heads because it is the products leaving the latter that generate the revenues. Awarding “bigger bonus” for “more patents” is good, but it can become evil if the award is not connected to the actual use of the standards containing the IP.

This is not theory, it is a sad reality for one of the most important MPEG standards: High Efficiency Video Coding (HEVC). 63 months (5 years and 3 months!) after its approval, the licencing of HEVC (that is outside of MPEG purview) can be described as follows: ~1/3 of the ~45 patents holders have not published the licence of their IP and ~2/3 have joined one of the 3 existing patent pools, only two of which have published the licence of the IP they administer.

MPEG is embarking in a new video compression standard, even though the licensing of the previous HEVC standard is in the state I have described. CEOs have better not to behave like 19th century captains. The stokers are doing their work but there are more people whose actions have to be reined in.