This may be of interest:
Amazon, Cisco, Google, Intel, Microsoft, Mozilla, and Netflix are pooling
resources to work on hammering out a standard, royalty-free, next-gen open
Cisco and Mozilla had already started working together, combining ideas
from Cisco's Thor codec and the Mozilla-sponsored Xiph's Daala codec, but
getting them together with Google's VPx team is a big happy occasion...
getting Intel and Microsoft on board means probably hardware support on
x86_64 chips and support in Windows and MS Edge, while Amazon and Netflix
pump a lot of video volume out both to browsers and devices, so should help
push adoption by ARM SoC makers.
In other words, AWESOME SAUCE!
The only missing major player looks to be Apple... so we'll see if they
eventually join on or if they double down on HEVC and we keep having to
I'm subscribing to their mailing list for updates; it may be worth looking
into if we can partner in or at least follow along and support stuff on our
Thanks to Giuseppe in ops & a bunch of other folks who have helped out,
we've got an updated server configuration for the 'video scalers' that
generate media transcodes, supporting VP9 video and Opus audio.
Initially we have one server updated and in production, and the rest should
follow by tomorrow. (Updated servers are also in place on the beta cluster.)
This should make it possible for WebM files using VP9 video and/or Opus
audio to work as fully first-class citizens in TimedMediaHandler,
generating medium- and low-resolution transcodes in Ogg and WebM VP8.
It should also fix Ogg audio files using the Opus codec to work
Support for generating VP9 transcodes as well is in the code and just needs
to be switched on -- I'll tweak that configuration once we've got the rest
of the servers upgraded and have more capacity for batch runs.
(I've got a client-side bot which I'll start once that's ready, to force
all the files to generate over time.)
(cross-posting from wikimedia-l)
Last year, the Wikimedia Foundation published our first ever video
Which covered some of the major news events of 2014 through the lens of
Wikipedia. This year, we’re opening up the idea development and
pre-production process of making a video for 2015 to everyone. This is an
opportunity for you to help shape the narrative of the events of 2015:
Last year’s video was made largely by myself and another video editor over
about 8 weeks at the end of 2014. I spent the first half of my 8 weeks
researching news, comparing that to view and edit counts of Wikipedia
pages, and searching for media to illustrate those events. After I had that
media, it was a matter of taste to place them in a video editing timeline.
When we published it, the press and the general population on the Internet
reacted positively. All things considered, I think that #Edit2014 was a
good start, and I’m happy with the final result—but I’d like to improve a
few things for #Edit2015.
Here’s the plan:
Open Collaboration: I’m opening up the whole idea-development and
pre-production process (research, scriptwriting, brainstorming, finding
media, etc.) for making #Edit2015 to on-wiki collaboration. While
experimental, we have #Edit2014 as a guide to show that a final product can
be done; it taught me that year-in-review videos cover international news
events through a brand (in this case Wikipedia) by telling each news story
in about 5 seconds and then cutting to the next one. After being multiplied
by around 20 stories, your video will be upwards of two minutes long when
the credits, logos and titles are included. If you watch other
year-in-review videos (like Google Zeitgeist Year In Search) you’ll see how
each will spend 5 seconds on a topic and then jump to the next.
First drafts of #Edit2014 were half global news and half wiki-world news. I
wanted to showcase as many Wikimedia tools, events and projects as
possible. What I found was that since this is for a wide audience, and it’s
only a few short minutes long, we only have a chance to communicate one or
two new ideas (for an ordinary person who uses the internet), so we had to
be very selective about what was showcased. In this case, it was a chance
to talk about the edit button and Wiki Loves Monuments briefly. Then we
have to get back to those global shared news events that the public may
have experienced. Aspects like ‘going down the rabbit hole,’ clicking link
after link, was something that ordinary people were familiar with, so this
is something we used to bridge stories.
The idea-development and pre-production process does not require any fancy
video equipment—just a wiki page and an internet connection. I used post-it
notes on my wall to organize my ideas. I think that we—that is, the
Wikimedia crowd—can be very good at story development and collaboration.
I’ve been collecting imagery and ideas online, and I’d like to allow anyone
to use this space as a place to collaborate on this project:
An idea I had for this year is to somehow showcase the talk pages about
Wikipedia articles, to show how we arrive at consensus and a
neutral-point-of-view. Finding the right article(s) and talk page quotes to
use to illustrate that would be key. Last year, we showcased the edit
button using the 2014 Israel-Gaza conflict. For that, we see closeups of
‘citation needed’ and ‘disputed-discuss’ then we cut to the different
languages of that article. Imagine if you saw a tiny fraction of the
behind-the scenes talk about an article like that and how it aims for
Rules: These are some basic criteria I made to guide what content got into
Has the event made it to the international press or wide regional press?
Does the event have corresponding view or edit counts?
Do we have freely licensed imagery for the event?
Was there a special circumstance about this event per Wikimedia projects?
Does this illustrate some aspect of Wikimedia that the public should know?
Is the media beautiful?
Does the Wikimedia Foundation legal team approve of the media?
Do we have some media and news from every major region of the world?
As for production and post-production – Continuity, music, audio mixing, et
cetera are all things that should ideally be online and in a collaborative
manner but currently there is no system in place to collaborate on those
things using Wikimedia projects. I’d love to develop that system, but I
don’t think that it is practical for this year. I’d also like to aim to
make the video as close to 2 minutes in length as possible.
Schedule: So the logical publication date for #Edit2015 is December 15th
because that’s when the press, who would republish and spread the video on
social media, are still at work and this is an easy story for them to
publish before they go on vacation at the end of December. Getting this in
the press gets more eyeballs on the video. That means that actual video
editing should be well on its way in October and November. This is my
current schedule (for now):
Brainstorm and pre-production: now – October 1st
Production (assemble the footage): October 1st – November 15th
Post-Production (lock all the details): November 15th – December 1st
Distribution (captions and translations, thumbnails, text copy, uploading,
and any last-minute edits): December 1st – December 15th
Internationalism: My biggest problem with #Edit2014 was that so much of it
was in English. While we tried to cover as many regions and languages as
possible, a non-English speaker probably had to watch it with the captions
on—and then your eyes are stuck reading text on the bottom of the screen
rather than viewing the interplay of images and text. I think that opening
up the development and pre-production phase would flatten out the
perspective quite a bit, or at least help to point out flaws and suggest
other ideas. We shouldn’t have to rely on captions to make it universally
understandable, and since we’ll be jumping from one story to the next in 5
seconds, we can express a story in any local language. There may also be
‘universal’ communication media like video, imagery or numbers that are
associated with the text that they can understand.
Media Content: There are a few sources for freely-licensed imagery that we
can use for #Edit 2015: still imagery, video imagery, Wikimedia project
pages, audio, and imagery we make ourselves. I’d love if we could somehow
have more audio/video content for #Edit2015. I looked for freely-licensed
video and .gifs on Wikimedia Commons, Vimeo, Internet Archive and YouTube,
and I know there are many more I could have used. The first few versions of
#Edit2014 incorporated more video than the final cut did, but much was cut
out because it was too busy or complicated to communicate an idea quickly.
Sometimes a still frame of a Wikipedia article or a still photo might
communicate the idea more neutrally or succinctly than portions of
freely-licensed videos could.
For every still image you see in #Edit2014, there’s probably 10 more that
didn’t make the final cut. It took a lot of research to find appropriate
and compelling imagery.
I’m very optimistic that this will be a fruitful initiative, and I can’t
wait to see all the usernames of everyone who contributes. Please share
this link with your friends:
Let’s collaborate and tell the story of Wikipedia and 2015 together.
Storyteller <https://wikimediafoundation.org/wiki/WPZeroPetition> and Video
FYI, this is the video-related submission for Wikiconference USA, October
9-11, 2015 in Washington DC. We plan on having part of the Sunday
unconference day to work on next steps related to possibly working with
Mozilla and Internet Archive to further a full editing system.
Hope to see some interested folks from this list there in DC.