[M4IF Discuss] The Discussions So Far...

[Rob Koenen]

People,
First details about Visual licensing were announced 5 weeks ago, and 
we have seen many responses, many of which with a concerned tone of
voice. 
It is good to see where we stand.
MPEG standards are developed with contributors agreeing in writing to
license any essential patents on Reasonable and Non-Discriminatory
terms and conditions (RAND). Most of the discussions on licensing on
this list have concentrated on whether the tentatively announced terms
are indeed reasonable. 
There seems to be general consensus that, in order for MPEG-4 to succeed,
the licensing scheme: 
1) Must be reasonable, for *all* markets, applications and services to 
   which MPEG-4 provides an solution;
2) Must allow solutions to be built that are competitive to 
   proprietary solutions in the same market.
There are 2 elements to the license:
1) The encoder and decoder fee of 25 cents. Most people think this is 
   acceptable, although some people favor free decoders and more fees
   on the encoder.
2) The use fee of USD 0.02 per hour (actually calculated per minute).  
   The use fee is controversial, and there is much debate over how 
   this works out in various circumstances. 
   If we accept that 2 cts is reasonable for MPEG-4 Visual Simple and 
   Core Profiles, then we must accept that a similar use fee is 
   reasonable for
   * The other MPEG-4 natural video profiles (?)
   * Synthetic MPEG-4 visual tools
   * MPEG-4 Natural Audio
   * MPEG-4 Speech coding
   * MPEG-4 Systems
   * MPEG-4 DMIF
   (* and perhaps further non-MPEG-4 technologies encumbered by patents)
   In other words, in assessing how reasonable 2 cents per hours are 
   under various circumstances, we need to assume that the use fee for 
   a complete MPEG-4 System is a multiple of 2 cents per hour.
Below are the main application areas of MPEG-4, with an assessment of 
how the fees may work out, derived from the discussions so far.
(please add if you see more...)
A. Person-to-person communication. No use fee; just the encoder/decoder
   fee. Most parties seem to think this is reasonable, certainly in 
   hardware implementations.
B. Personal content for which there is no remuneration. There is no
   use-fee, and again the encoder/decoder fees are within the realm of
   what pro versions of players/rippers are being sold for.
C. Individual on-line access to content for which there is remuneration.
   The use fee applies, and applies every time the content is accessed.
   If the content is of high value, and the price too, the use fee is
   low relative to the remuneration. There are concerns about 
   accounting.
D. Media play-back where remuneration is indirect, e.g. ad-supported,
   sales-generating, for courseware, etc. Depending on the exact 
   business model, the use fee seems to be able to become a real burden,
   and many concerns have been raised. Note well that this is the 
   dominant usage on the Internet. Whatever technology is to become
   a standard for on-line usage, must support this business model.
E. Subscription content. This one sits between C and D. 
F. On-media distribution. There is a one-time decoder fee per play 
   medium and a one-time content use-fee, no matter how often the 
   content is played. It is very similar to the current DVD MPEG-2 
   royalty. There doesn't seem to be contention, although in highly 
   interactive multimedia content it will be impossible to constitute 
   what the play time is.
G. One-to-many distribution, such as in webcasts or classical 
   broadcasts. Not much has been announced, but there is some fear 
   that the use-fee will be calculated as follows: 
     actual viewers x hours x 2 cts
   according to some statistic on how many viewers there are.
   There are concerns about the competitiveness of this scheme, and
   about accounting, see below.
In general, there are also concerns about the viability of keeping 
track for all potential uses. Not all broadcasts come with statistics,
certainly not the smaller ones. 
Also, it is unclear what will happen if service providers (e.g. 
Content Distribution Networks) facilitate file distribution, without 
them being aware or even wanting to be aware of what is being 
distributed and in which format. They could well be participating 
in the distribution of MP4 files, perhaps even zipped (although that
make fairly little sense, but that is irrelevant), and if I understand 
correctly, be 'eligible' to pay use fees if there is remuneration. 
This situation, in which a party that does *not* employ any MPEG-4 
technology is required to pay, may create problems that people will 
want to avoid.
Unfortunately, we do not have a lot of time to resolve the issues.
Technology choices are being made today, and company plans for adding
or dropping MPEG-4 support are being modified right now. I want to
dispel the myth that there is plenty of time to resolve the issues. 
There may be in some markets, but surely there isn't in most. There
especially isn't in the market that should lead MPEG-4's adoption,
the interoperable streaming media arena.
I look forward to continuing a focused discussion; one that hopefully
bears a real impact on what is finally agreed between the licensors as 
organized in the patent pool and the community of potential licensees.
I would also like to extend an invitation to the licensors and their
representatives to keep the world informed about their thinking. 
this is crucial if we want to stay with the facts, to counter the FUD 
and to allow informed decisions to be made by potential users.
Kind Regards,
Rob Koenen
President, M4IF