VP8 Codec SDK "Aylesbury" Release | Hacker News Reader

VP8 Codec SDK "Aylesbury" Release

blog.webmproject.org

18 pointspavs14 comments

For those wondering.

  What's an Aylesbury? It's a breed of duck. We like
  ducks, so we plan to use duck-related names for
  each major libvpx release, in alphabetical order.

Aylesbury is a sleepy dormitory town in Buckinghamshire near where I grew up and is famous for very little except giving its name to a breed of duck. I was pretty surprised to see a codec named after it.

Probably more important:

  Our goal is to have one named release of libvpx
  per calendar quarter, each with a theme.

On2 (who Google bought to get the video tech/patents for vp8) were previously known as The Duck Corporation.

Nice, I was wondering if the name had any link to the town, which isn't too far from me.

The MPEG LA said that it would assemble a patent pool against VP8, did they manage to do that, or they were just FUD-ing around?

I'm not saying they will or they won't, but there is no point wasting resources assembling a patent pool against VP8 if it isn't even popular yet. When it becomes a threat, you can bet then the MPEG LA will assemble something, valid or not.

VP8 is used by skype for video and VP6 (which VP8 is based on) was used by youtube for a while on many videos. They could have made millions of dollars by suing huge companies such as Skype and Google. Why didn't they if they had patents on them? Doesn't make sense.

When did Skype switch to VP8? As far as I know they still use VP7. What does VP6 have to do with VP8?

VP8 is heavily based on VP6 and VP7. If VP8 breaches some patents, they are very high odds that so do they.

cite please.

Feel free to read the libvpx source, it's based on On2's VP7 codec and mentions it in comments in several places.

You can also compare ffvp56 and ffvp8, although there's a larger time gap between the two implementations.

What would that accomplish. The original statement was that skype uses vp8 and youtube used vp6 (which vp8 is heavily based off of). Skype doesn't use vp8, and looking at vp7 or vp8 source isn't going to tell me that vp6 infringes on any patents.

Who are they going to attack? As with Oracle's suit, you can bet Google will be chomping at the bit to invalidate anything they put in the pool, which doubtlessly would weaken their bread and butter MPEG patents.

And the thing is, hardware H.264 will almost certainly be in every phone built until the patents expire, so ragging on a free alternative is just going to jeopardize their existing revenues, for a shot at the hacker community's websites like Wikimedia running H.264 and paying royalties - which will never happen anyway. Meanwhile most likely future for WebM is that some forward thinking companies serve WebM if you're using Firefox/Opera, but most are lazy, pay their H.264 tax, and require you to use IE/Safari/Chrome.

> The MPEG LA said that it would assemble a patent pool against VP8

They said the same thing about Vorbis ten years ago, don't hold your breath.

Um, I'm pretty sure that measuring PSNR and SSIM differences in percentage makes no sense at all. PSNR is measured in decibels and SSIM is a log scale. They might as well just have said "it went up".

Reply on news.ycombinator.com