Thanks to some funding from the alt-media-res project, we now have a [http://shiftspace.cc/jamie/tx_report_0.1.html draft metadata standard] [http://shiftspace.cc/jamie/re-tx_report_0.1.pdf PDF] prepared by JJ King and Jan Gerber:
This report and proposal is under consultation within Transmission-connected networks until Monday November 6th;
We are looking for substantive responses, feedback, proposals, in particular direct inputs from the other Transmission working groups - eg Translation/Subtitling, DoNoHarm, Aggregator R&D, Documentation....
I am posting [http://wiki.transmission.cc/index.php/Responses_to_draft_schema summaries of responses] in the wiki and hope that people with experience of this field of work can suggest practical ways forward to finalise and then begin to implement the standard.
After 6th November we will gather the working group together to review and improve the schema, spec out next steps etc. please add to and improve the [http://wiki.transmission.cc/index.php/Metadata_Workspace metadata to do list]
As part of this process, JJ King and I have drafted a [http://www.clearerchannel.org/transwiki/index.php?title=Proposal_for_further_funding_for_implementation_of_RDF_schema proposal for implementation] of this standard.
28th October 2006
1. Common Aggregation Site
Why? To simplify the browsing and searching of all of our content for our audiences, to cross-promote our content and get it seen by more people and enable greater collaboration
- what would this require each group to produce in order to contribute?
RSS feed with text metadata - what metadata included?
RSS2 with media enclosures - vodcast - what metadata included?
- does it require a new collective to make it?
Use Transmission site as basis of this
Develop over transmission-discuss list or make a sub-collective for developers
Look at Global Voices, are we doubling up?
- what other software functionality do we need to develop or what other software to make use of?
Search across all projects content
2. Common Data Structure
Why? To enable us to share content, opening up the data flows by letting us communicate with each other.
- what kind of projects might we want this to enable data sharing with?
- FOSS CMS friendly
- Vodcast-client friendly
- Also consider: Public Access / Community / Pirate TV
Commercial / Government TV
Small-scale DVD/VCD distribution
- how does this define what the common data structure might look like?
- Smallest amount of conversion between one system to another
- Additional metadata added for various projects but maintaining a standard core
- what do commercial and other big projects do that we might learn from, and potentially wish to integrate with?
- Google Video - what is required to be listed in Google Video?
- iTunes Store - what can be gained from knowledge of those data-sharing standards?
- what are some ideas of what the common data structure might include?
- RDF semantic web - SOMA?
- Unique identifier per program
3. Common Database of Screening Groups
Why? We all have contacts of screening groups in our local areas that would be interested in each other's content, sharing these contacts would save lots of doubling-up on research and greatly improve links in our networks
- how could we share our screening contacts?
- Use a common database
- what could we build this database with?
- Other online-community software allowing feedback through the network (Plone?)
4. Common Codecs and Formats
Why? To simplify the process of encoding video both for our projects and our audiences
- to what extent can we use common codecs?
- Recommend codec standards based on our discussions at Transmission, and adopt them if appropriate for our own projects
- what are some proposals for what these common codecs and standards might be?
- XVID SCREENING DOWNLOAD: VIDEO - 1200 kbps bitrate, xVid codec, AVI format, PAL/NTSC Resolution and Frame Rate, AUDIO 128kbps bitrate audio, lame mp3 codec
- THEORA SCREENING DOWNLOAD: VIDEO - 1200 kbps bitrate, Theora codec, OGG format, PAL/NTSC Resolution and Frame Rate, AUDIO 128 kbps bitrate audio, Vorbis codec
- FLASH WEB PREVIEW: VIDEO - 300 kbps bitrate, vp6 codec, FLV format, 320x240 Resolution, 15fps Frame Rate, AUDIO 64kbps bitrate, AAC codec...?
- what kind of applications could we design to help end users adhere to this standard?
- One-button GUI for mencoder to produce standard output video for your project, downloadable as a cross- platform binary package
- Transcoder/Uploader client to produce standard output video, attach inputted metadata including CC license as XML and in the video file, and upload to your online archive
- Simple documentation available for people to encode using popular applications
5. Ongoing Platform Design Research and Feedback
Why? If it is too much to think we could work on the same codebase, we can look for other ways to avoid doubling-up in our work, and for supporting each other
- what considerations should we have in mind for future development of our video-sharing platforms?
- Effectiveness socially and politically - working with social movements and campaigns rather than making tech that is fun for us to play with but not so useful
- Scalability - how many people can we reach with our systems?
- Targets - are we reaching our target audiences if a mass distribution is not the aim?
- Usability - major consideration for something as technical as Internet video, keep it simple
- Sharing - how content can be syndicated across other sites
- Convergence - how to make our media easy to move across mediums and devices
- Cost-efficiency - how to keep costs down
- Sustainability - how to maintain developer and editorial collectives by better software and systems design
- how can we maintain a supportive network for sharing information and constructive criticism of our approaches to distributing video online for social change?
- Communicate with, inform and help each other
- what would this network entail?
- Email list
6. Democracy Player
Why? The best implementation for our purposes of a vodcast/bittorrent/player client that is intuitive and makes it simple for audiences to view our content in a tv-like way
- what do we all need for our content to be Democracy-Player-optimised?
- what can Participatory Culture do to work better with our networks?
- Package of Democracy Player bundled with pre-loaded feeds from our projects
- Feedback mechanism between our groups and Participatory Culture - email
Took me some hours, but there was nothing much to do while waiting for the PowerBook to encode videos, so, here it is: http://www.jeremiahfoo.com/blog/?p=354, pictures I took during the meetings.
I will try to compile the videos so I can start podcasting soon.
I thought I'd better at least get a couple of notes in here before I forget everything that's happened over the last two days.
It's hard to get a perspective of where things are at from where I'm sitting, I'm mostly comparing things to my own expectations. I think I was expecting things to move into being quite specific more quickly. What I am noticing instead however is that there is a lot more of getting everyone on the same page as regards each others projects and their general operations. This isn't surprising perhaps as it's (as far as I know) the first time so many activist online video distribution projects have sat in the same room. It's a bit much to expect everyone to just jump into coding and collaborating on software immediately it's much more a of a long term process. I think it will take 6 months to start seeing some real outcomes of what's going on here in Rome.
The content management system discussion today was a prime example. We didn't get too specific but had a quite a good discussion (I thought) on the basics of what people thought the key aspects of a CMS should be.
Between about 30 of us I think we covered some good ground and focussed on quite basic aspects of how we might collaborate together but there was some confusion in terms of what we should be considering a CMS and some frustration at not being easily able to see options immediately before us to work together.
My conclusion from this discussion was that basically a lot more discussions need to take place (and a lot of them need to be small conversations too i think) and it really extended my time line of how long building those solid collaborations are going to take.
I came out out feeling quite positive about it all, hopefully it's the first conversation amongst many.
oh and the food in forte rocks!!
Jaromil and Eleonora wrote a research about video streaming, commissioned by dutch media institute Montevideo.
It has not been published yet on the net, but Jaromil brought it, and it's licensed under GNU Free Documentation License.
jaromil explains that copy protection is a bad idea, and also a waste of time.
talking about streaming, there are two ways: on-demand and 'programmed' (live).
For programmed streaming, you can use Darwin Streaming Server (with MPEG4IP as encoder), or Icecast2 (using FFMPEG as encoder).
The research also talks about codecs, focussing on open-source codecs, and takes into account licensing/patenting issues. There are 2 open source codecs that are not discussed: Dirac (BBC) and Snow (ffmpeg related) because they're not ready.
The research is about Wide Area Network streaming, so streaming for the Internet. For LAN, one could also do MPEG2 streaming, but bandwidth limitations are a major concern for internet streaming.
MPEG-4 is patented proprietary technology
XviD is an open source "MPEG4-class". US patent regulations make it somehow tricky to distribute binary versions.
Player integration is problematic, plugins for players.
H.264 / AVC: very effective, can be downscaled to low bitrates, can also be used on mobile devices.
Theora: completely open source. results are comparable to MPEG-4, but it uses very different technology.
theora encoding is still slow, because there are not many optimised implementations (except MMX).
Theora is supported by various players, but there is also a java applet, 'cortado', that plays video from the web.
Only in Linux there is a problem with java applets. FreeBSD systems do have a Java environment installed.
Jaromil and Eleonora's research also has a codec benchmark table.
They found that for most artworks, a 1200kbps gives good results for PAL.
H264, using mencoder with x264, the encoder was the only one that was doing what it was supposed to do.
Theora is the best codec
For low bitrates, H264 is the only solution
- more control
- can be scripted
- can be configured using a text file
- settings can be searched, you don't have to click around
- it can be hard to find the exact command, with all the options and switches
- especially for video encoding, there are a lot of options to set.
a simple command to try transcoding, is using the programme 'convert' which is part of ImageMagick. It is installed on most linux systems.
#convert filename.xpm filename.jpg
converts the image filiename.xpm into the jpeg format.
Eleonora and Jaromil's research includes a script that was used to encode video's in various codecs using various bitrates, using mencoder and ffmpeg2theora.
This is great as a starting point to learn about buiding the correct command-lines to encode video.
mencoder is better than ffmpeg for encoding. ffmpeg is more a library with a lot of codecs.
related projects are:
FreeJ that can also do (live) streaming, with a lot of interaction control.
Dynabolic: linux distribution for multimedia by Jaromil, just released in 2.0 version.
The research PDF and the Dynabolic image will be uploaded to the local FTP server.
Links only valid during transmission, at Forte Prenestino:
Engage! Tactical Media http://www.engagetacticalmedia.org
I briefly spoke about three related projects than I am involved in. All are related to mobile phones.
1. Audio presents at radio documentary Grenzeloos Geluid
I worked with Nathalie Faber and Carolien Euser of artist collective cut-n-paste on "Callcentre 211" that ran during radio documentary festival Grenzeloos Geluid. Participants registered their phone number at a desk, and were called 15-20 times during the event. After answering their phone (or dialing their voicemail) they heard poems, narrations, soundscapes, music, etc.
I developed this system mostly using Asterisk, a fabulous open-source voice over IP system.
2. Video art presents during Playmobiel exhibition on art on the mobile phone
Every day of the exhibition (except mondays when the museum is closed),
subscribers get an art video, made especially for the mobile phone, on their handset. Subscribing works by sending a text message. The video's can also be seen in the upperleft corner of the playmobiel website that integrates the mobile site.
Technologies used: Kannel sms gateway, php, mysql, drupal, 3gp.
3. Drupal 2-way SMS module
A module to send sms text messages to a group of drupal users.
But the text message is also used for querying the site: for example, by sending "events" via sms to the site, you will receive a message with a list of the first upcoming events.
Tech: Kannel, Clickatell sms-api, Drupal
See attachment for a flowchart of how it could work.
Collaborators wanted please!
Upcoming documentation is this:
Zoe, Rama and me (=maxigas) are compiling notes of the procession.
i am compiling a directory of people with images.
i would also like to post some pix i made.
Today i about presentation of people and projects and a roundtable on videodistros...
using gstreamer-0.10 to stream DV video with audio via alsa, could look like this:
gst-launch \ oggmux name=mux ! shout2send id=icecast.v2v.cc port=8000 mount=/theora.ogg password=hackme \ dv1394src ! dvdemux ! dvdec ! ffmpegcolorspace ! theoraenc ! queue ! mux. \ alsasrc ! audioconvert ! vorbisenc ! queue ! mux.
I just thought I'd add a bit of silliness as the name of an album by a space-disco band I used to know just came to mind..
Perhaps it's the cleaning products and dust that are going to my head! Or perhaps I am floating around on a space-station listening to Grace Jones after all.. that would explain why Andrew, Andy and Agnese are dancing around me, doing the hustle in silver spacesuits.. I definitely have dust on the brain.
Transmission. I'm looking forward mostly to meeting people from other projects and learning about our different approaches to similar challenges. I think we have much to learn from each other, which will help in critiquing our current methods and help us all plan better for the future.
If in addition to this, ongoing collaborations can come out of the gathering, or a network of support can be maintained after it has finished, then I'll be very happy.
The idea of moving towards common platforms may or may not make sense to various projects attending, but I would like to engage in those discussions as there are definitely major advantages to sharing code when we are often developing similar functionality with limited resources for development.
Linking development of multimedia modules to a large open-source content management system has been a way for EngageMedia to both contribute to a large community of users and developers, while benefiting from the huge amount of work that has already been done, and from regular updates on areas of code we would not have the time to develop ourselves. It would be good to get other people's opinions on this as a way forward.
Specifically, I'm also interested in talking about creating an aggregate site of all of our content - this could evolve out of the RSS feeds on the front page of this website. Being able to search for content across all of our databases would be very useful for people looking for progressive content worldwide.
Another outcome I am interested in is documentation - how to work together on common help files and tutorials for web video.
I'm also looking forward to hoovering up a few gigabytes of video from the media server to take with me to screen on my further travels and when I get home! We'll be putting video from the Oceania Indymedia Newsreals for others to come and grab.
Also I'm interested in discussing with our peers all sorts of similar issues we face such as..
The perennial codec question - what formats and codecs should our projects be supporting that currently strike the best balance between ease-of-use/ubiquity and free/open-source software considering that many video producers may be highly skilled in production but may not be particularly net-savvy or familiar with video encoding.
How to improve the usability of our websites in general, and how to facilitate community building by implementing elements of social software. Making sure that we look at the networks and communities that underpin our projects to see how we can reflect the way they organise offline.
Learning how to bring together audiences and screeners with independently produced content without the technology getting in the way.. while helping to facilitate the breakdown of the divide between producers and consumers of media, as is happening everywhere on the web.
Looking at ways of working with individuals, campaign groups and social movements to make the step from information to action more possible.. and many more.
See you all in a few days!
Oh, here are some pictures of some of the posters stuck up around where we have been sleeping at Forte Prenestino. There is a comics festival happening here at the same time as Transmission, so even more of an opportunity to catch some dynamic Italian graphic art! Here is a taste of what you will find at Forte..