Standard media formats and licensing: JPEG versus MPEG

Free as in beer or Free as in 'This is a holdup!'

By Rick Jelliffe
July 31, 2009 | Comments: 7

It is interesting to see the differences in approach between the ISO JPEG group and the ISO MPEG group on IP

The JPEG FAQ says

the JPEG committee have always tried to ensure in their standardisation work that the 'baseline' part of their standards should be implementable without payment of either royalty fees (volume related) or license fees (non-volume related).

The MPEG site says

Various MPEG standards have benefited from the existence of patent pools where organisations wishing to license technology included in such standards can negotiate a single license covering all the essential patents identified by those organizations which administer the patent pools.

Licenses, even peppercorn payments, are a real stumbling block for artisan developers, who were the bedrock of FOSS until the coporates co-opted it. You cannot attach payment to something you give away for free: the difficulties with H.264 are a current example of this. The solution seems to be that the codecs must be provided as part of the graphics card (which is paid for): NVidia is associated with this approach. Developers have to reach outside the self-constructed FOSS gilded cage to something posing as hardware.

It may be that this single issue of media standards licensing is what has prevented (and will continue to prevent) SOHO adoption of Linux. The question needs to be asked whether the MPEG group really has the necessary balance of interests (though I am not saying it necessarily does not: just that it is an economically, socially and technically important issue.)

While I don't think that all voluntary IT standards automatically should be RAND-z, I cannot think of any examples where I think it (RAND) has been a win for everyone, off-hand. Patents shape the direction of the R&D that corporations do, but the corporations need to do R&D to compete and live regardless of patents: innovation is in no way predicated on ownership. And certainly I think that strong market success obliterates the value and social usefulness of IP protection.

I have recently blogged about the need for standards bodies to have a balance of interest so I was intrigued to see Microsoft's Sr Director for Interoperability Jason Matusow's take:

The idea that "open" = no IP in a standard is an overbalance in favor of implementers just as the idea that a single party has unequal say in a working group can overbalance in favor of a contributor.

But the practical issue is not ownership but RAND versus RAND-z licensing. I don't get Jason's point about imbalance. The paying users get a good technology and the IP-owning vendors get a good profit would be quite an eccentric definition of balance.

Spurred on by these kinds of considerations, I have sent an email to the ISO/IEC JTC1 SC34 secretariat today, to draw attention to the JPEG group's resolutions on IP (as linked above). I think it reflects where SC34 has long been at, but I would like it to be more explicit, and helpful for all concerned, and for the public who have heard talk of submarine patents and so on. Of course, SC34 are no more bound to pay any attention to me than to Charles Manson: but I hope NB delegations and other standards committees will follow the JPEG lead. Anyway, here is one of the resolutions:

Again, at every JPEG meeting the following resolution is also passed unanimously concerning the work of the JPEG committee's parent body within ISO, JTC1/SC29. "SC 29 affirms and supports ISO policy that requires disclosure of the existence of Intellectual Property (IP) rights or pending rights (such as patents or pending patent applications), hereafter referred to as "IP rights", associated with any technology submitted to SC 29/WGs for consideration for inclusion in any ISO/IEC standard. Specifically, SC 29 affirms the ISO policy of only considering technology that is free of "IP rights" or which is available on a royalty and license fee free basis or which is available under reasonable terms and conditions on a non-discriminatory basis. "

You might also be interested in:


You may also be interested, if you're not already aware, in the current push for royalty free "internet codecs" in IETF:

Dave: Thanks for the reference. My readers might see this as an interesting project to track or participate in. There needs to be reliable unencumbered codecs available, certainly at the base level.

I am afraid that the shortsighted academic policies of the 1990s contributed to the rot: researchers were encouraged to patent, sometimes even measured by their patents, and a lot of research on public money that should have been for the public good ended up in no good for anyone, with the research institution forced to see research as an investment for its future survival. Fraunhofer is not in an unusual position in all this.

It is almost hypocritical: governments adopt open standards-encouragement policies, but only for standards in areas that they use; for standards that they are not big users of, the encouragement dries up.

But all the research and patenting has made it a very difficult area to work in, for groups wanting to make FOSS codecs, I imagine.

AVC baseline profile is also royalty free. The patent issue is not an MPEG issue, it is an ISO issue. Note that JPEG (WG1) and MPEG (WG11) are part of SC29 and thus, your last quote also not only applies to JPEG.


Christian: All good points. It is an all-ISO issue.

But it is also an MPEG group issue.

For readers: Wikipedia's entry on H.264 AVC has:
"Baseline Profile (BP): Primarily for low-cost applications that requires additional error robustness, this profile is used rarely in videoconferencing and mobile applications, it does add additional error resilience tools to the Constrained Baseline Profile. The importance of this profile is fading after the Constrained Baseline Profile has been defined."

Rick - I really like the thoughtful posts. I have just posted a longer piece to clarify what I was saying about balance. I completely agree with you that greater participation (beyond vendors) is important. I was using "balance" in a different way and hope to have clarified that a bit.


Here's my response to Jason (which is presumably awaiting moderation):


I spend the vast majority of my time working on cloud standards these days which are typically creative commons licensed and associated with royalty free patent pledges (just how I like them).

First let me refer you to Microsoft's standing definition of open standards courtesy Wikipedia[1]:

"Let's look at what an open standard means: 'open' refers to it being royalty-free, while 'standard' means a technology approved by formalised committees that are open to participation by all interested parties and operate on a consensus basis. An open standard is publicly available, and developed, approved and maintained via a collaborative and consensus driven process."

See that? "'open' refers to it being royalty-free". What is "open" if *not* royalty-free? Merely published? That's just a standard I'm afraid, per my "degrees of standards freedom" diagram[2].

It looks to me like what we've got here is a brain fart... which is fine - we're all human. The question now is, will you admit to it or deny having done it?

The thing is that Microsoft is just starting to win back some points with cynics like myself (I'm now more supportive of MS than I ever have been), and then you go sprouting stuff like this which makes me realise that a leopard can't change its spots.

Open is trending towards more open, not less. Resist at your own peril.



Sam: Just to note, what you call "Microsoft's standing definition" seems to come in the course of a newspaper interview on a particular topic by a particular MS national spokesman, so I wouldn't get too carried away with thinking it necessarily represents some official corporate-wide official definition. It is not something you can really hang a gotcha on. (I think you have to take all the different comments and try to get a sense of the range of opinions floating around, over the years.)

For readers interested in Vijay Kapur's thoughts on standards, I draw their attention to his 2007 article and in particular to following sentence, which I think encapsulates Microsoft's approach at that time A good strategy to achieve interoperability is to apply widely accepted de facto standards and open standards based on pervasively used technologies.

I don't think that is very objectionable in itself, providing you don't conclude it is the only strategy. But I think the debate however is on the desirability of the kind of interoperability that following-the-leader results in.

The micro-optimization of a single dominant technology may need to be weighed against the need for the macro-optimization of a competitive market/bazaar.

News Topics

Recommended for You

Got a Question?