14 comments on “WebRTC mandatory video codec discussion: the final duel?

  1. How the 100k units per year should “be adequate for many WebRTC innovators or start-ups to try out new implementations on a large set of users before incurring any patent royalty costs” is a mystery to me. What happens after you try it out? You cannot afford it and you cast it aside. I’m part of a startup and I’m well aware of the fact we wouldn’t absolutely be able to afford the licensing costs of H.264 for the applications we’re building on top of WebRTC (conferencing, streaming, etc.).

    The choice of H.264 would only result in a full stop to all of the innovativations and fresh ideas people have brought to WebRTC so far, because the licensing burden would not only apply to us, but to tons of other implementors that are doing great stuff right now.

    Most of the H.264 proponents have contributed zero or close to zero when it comes to WebRTC: they may be doing something on the ML, but I’ve seen nothing relevant when it comes to implementations, nice scenarios and so on (and no, I don’t consider Bowser relevant, and I don’t even consider ORTC at all: as if we needed another Silverlight vs. Flash or ASP vs PHP clone). The only stuff I’ve seen done was great browser implementations by Google and Mozilla, and excellent applications by WebRTC enthusiasts that at times pushed the concept beyond what I thought possible. When you don’t write a single line of code, just stay on the side watching, and then eventually try to impose your codec, I can only see one thing: you’re scared you won’t be able to compete based on merits, and so you’re bullying to impose your coded, to make money out of other people work or stop those who can’t afford to pay you, no matter how great their efforts were. This is defending the status quo, something that unfortunayely is not new in the IETF where we all know how strong large companies are: and something I will never accept or tolerate, especially when a technology that could affect, in a positive way, the future of multimedia technology is involved.

    If it can’t be VP8, let it be something else (H.263 or heck, even H.261 would work at this point) but I will never accept H.264. Legacy applications are called like that for a reason: WebRTC is moving towards the future, and a rock tied to its foot to keep it from evolving is the last thing it needs.

  2. The quote from the Skype (Microsoft) employee is interesting, but so was the response to that message on the mailing list which pointed out that if Skype (Microsoft) wants that position to be considered they really should submit an IPR declaration to the IETF.

    I have a concern about the table showing Google as the only VP8 backer. It is correct that Google is the only big organisation backing VP8. However, I am sure there are many, many, small start-ups (who don’t have the time or money to engage with the IETF) that support VP8 simply because there is no royalty-free option for H.264 at this time (I agree with the sentiments of the first commenter here – the 100k unit thing is just a fig-leaf). I believe that the most innovate use of WebRTC will be from these smaller companies and I do worry that the need to use a royalty-bearing codec will stifle that.

    To be blunt, the big organisations that are in favour of using H.264 can afford to switch to VP8 more than smaller organisations can afford to worry about H.264 licensing. Even the administration required to count your H.264 usage (required even if the first 100k units are free) may be too onerous for one/two man organisations with downloadable software products (what do you count, downloads, license keys, API keys – how do you manage all this without cost?).

    VP8 might not be royalty free in the future, but that still makes it a far better choice than H.264 which is not royalty free now.

  3. 1) I don’t know why any non-browser vendor opinion matters (aka. Cisco, Ericsson, Orange….). So even if that table shows a lot of companies in the right side, Google and Mozilla should be 90% of the decision and Opera, Safari and MS rest 10%.

    2) The only future proof approach in my opinion is VP8 as it ensures a simple transition to VP9.

  4. “Most of the H.264 proponents have contributed zero or close to zero when it comes to WebRTC”

    +1

    “If it can’t be VP8, let it be something else (H.263 or heck, even H.261 would work at this point) but I will never accept H.264. ”

    ++++1

  5. I think you are right – it is about to get really ugly! Thank you, Victor, for your comprehensive overview and good fortunes in Vancouver. I had not quite appreciated for imbalance of those at the table for H.264 until I stared at your table, which will make it a tough fight for a royalty-free VP8/VP9 in WebRTC. Lorenzo Miniero above is right that all the small start-ups and innovators who want licensing-unencumbered video codecs are not represented except through Google. You also point to the considerable FUD floating about – “who knows is VP8/VP9 is unencumbered, we may still be able to get you!”. Google has now spent over $300M trying to “free” all the codecs and code and it is disingenuous for companies to keep turning up at IETF saying “well Google hasn’t negotiated/paid/fought with us yet” within a standards body supposedly committed to “a long tradition of preferring non-encumbered IPR whenever possible, and especially to avoid IPR where using the technology requires making agreements with and payments to third parties as part of the cost of doing business”. Fare forward.

  6. UPDATE: Cisco to open source H.264 implementation, distribute binary module and absorb MPEG-LA license fees http://blogs.cisco.com/collaboration/open-source-h-264-removes-barriers-webrtc/

    See below the announcement from Cisco

    ———- Forwarded message ———-
    From: Jonathan Rosenberg (jdrosen)
    Date: Wed, Oct 30, 2013 at 1:28 PM
    Subject: [rtcweb] Cisco to open source its H.264 implementation and absorb MPEG-LA licensing fees
    To: “rtcweb@ietf.org”

    I’d like to make an announcement material to the conversations around MTI video codecs in rtcweb.

    Cisco is announcing today that we will take our H.264 implementation, and open source it under BSD license terms. Development and maintenance will be overseen by a board from industry and the open source community. Furthermore, we will provide a binary form suitable for inclusion in applications across a number of different operating systems (Windows, MacOS, Linux x86, Linux ARM and Android ARM), and make this binary module available for download from the Internet. We will not pass on our MPEG-LA licensing costs for this module, and based on the current licensing environment, this will effectively make H.264 free for use on supported platforms.

    We believe that this contribution to the community can help address the concerns many have raised around selection of H.264 as MTI. I firmly believe that with H.264 we can achieve maximal interoperability and now, do it with open source and for free (well, at least for others – its not free for Cisco J)

    More information on the open source project can be found at http://www.openh264.org, which is sparse now but more coming soon.

    Thx,

    Jonathan R.

  7. All the people working on getting the openh264.org and announcements out over the past few weeks were reading this article but sort of afraid to post anything on given what was coming. Great article – In my mind the big thing that has changed today with respect to this article is that Firefox is going to have both VP8 and H.264.

    We don’t think choosing an older codec really works – if you look at the video, it is so awful, no one wants it. Thus we think we have to break the log jam at IETF. H.264 as MTI does that and we hope that browsers support VP8, H264, VP9, H265, Daala and any good codec. If the handful of major browsers have wide codec support, and a common fallback of H.264, many applications can be build that use the codec they want and things will work well.

    Give users the choice, but make sure there is a common fallback. And my view is microsoft and apple were never going to do VP8 as the MTI.

    • And my view is microsoft and apple were never going to do VP8 as the MTI.

      “were never going”? Cullen, you’re talking as if the call for consensus had already been taken 😉

    • I think Cullen makes a key point here – if Microsoft and Apple are ever going to eventually arrive on board the WebRTC train the path of them having to adopt Google-driven VP8/VP9 as the MTI in the standard is probably not the most plausible path. So this is a very interesting response by Cisco to the video codec dilemma. Plus, in some sense, Cisco is putting a challenge on the table here to get Apple to do the same thing – make H.264 easily available in iOS to other apps without further fees.

  8. As Shakespeare would have said, from my point of view it’s “Much Ado About Nothing”.

    I don’t see any real value in this Cisco opening. First of all, as per the FAQ you cannot ship the module when you’re installing your application that depends on it: you have to download it on the fly, possibly getting the free license somehow in the process. Far from ideal, especially in case the servers are not reachable for any reason. Besides, as far as I’ve understood the “free” only applies to Cisco patents: the FAQ again says that this does not cover other potent holders tha may assert claims, which is EXACTLY the same situation VP8 is allegedly in now.

    Make the codec REALLY free and then we’ll get somewhere.

  9. I find it unbelievably rude to credit Wikipedia for the painting illustrating your article, where in fact the work of Ilya Efimovich Efimovich Repin, courtesy of the Pushkin Museum. As if I were to credit Saint Gobain for the Chateau Margaux that I drank last night :o).
    Other than that, great article !

Leave a Reply

Your email address will not be published. Required fields are marked *