Google & MPEG LA Reach VP8 Agreement

[allquixotic]

If we can get IE and Safari to finally implement WebM/VP8 support after realizing that there really is no patent risk here (well, no risk greater than using ANY codec whether licensed or not), we'll finally be able to use the HTML video element and expect the latest version of all major modern browsers to support that video format. No more user agent sniffing and switching over to H264.

Problem being, people will be running IE8, IE9, IE10, older Firefox, older Safari for a decade or more from now, and those people still won't be able to view the WebM video, even if Microsoft and Apple implement it in the very next release of their browsers. I hate people who don't upgrade their software.

[archibald]

If google trully believed that WebM did not infringe then they would not have "settled" but offer indemnity to users. Google rattled the sword only to keep it sheathed and then offered a peacekeeping.

I think you're being unfair to google here, as the MPEG-LA doesn't provide any indemnity against VP8 violations. It wouldn't just be a case of google's lawyers looking through all the patents and saying "it's fine". They'd need to be so sure about it that if every phone, tablet and mp4 player manufacturer implemented VP8 in hardware, they would be able to cover the cost of every single lawsuit. Given that the MPEG-LA doesn't have a time limit on how long it can wait in order to sue (so they could wait until it's more widely used), that's a much riskier option than paying a (currently undisclosed; so it may be large or small) sum to make the problem go away.

Neither side wants a massive lawsuit; I think the most likely outcome would be that both sides would have some patents invalidated, and would be found to be infringing on some of the other's patents. Given that, I think this is the best (as in: least disruptive) outcome for both sides.

[artivision]

The author has updated it and provided a better comparision:
https://gist.github.com/Hupotronic/4645784

Here are some words from its conclusion:

What a cheating test! H264 is 15% better than VP8 in fast motion level and about 7-10% in average movies. But in slow motion VP8 start to win with 10-30% difference. In still images-photos VP8 crashes the other part (like JPEG) with 40% less data. Also its free. The real downside is that costs 2x the time to encode video, near 8x for still images with perfect size, and also has just a little more decoding time.

VP9 is just 7-10% behind H265 but Google will release it when it 20% better, as Google says. Also they try to have better encoding/decoding than H265, they try their simple profile to be fast like Xvid but many times better. Also its free. For me, if TV/monitor production companies have VP9 decoder inside them, along with H265 that will be great, and now its possible.

[liam]

VP8 isn't as good as H.264 I'm afraid. We need VP9! But I guess they're waiting for WebM to be adopted everywhere.

It's gotten a lot better. Look around for recent comparisons with x264. It definitely beats baseline, for instance.
VP9 is also pretty good, even considering these intial encoders typically aren't very good.
This is an older link, but it is nice and clear: http://downloads.webmproject.org/ngo...results.html#1
This thread has some not-very-thorough-testing but it is more recent: http://forum.doom9.org/showthread.php?t=165839

[shruti001]

i am not getting it accurately can you people make it more clear