Hypermediated TV
10 likes755 views
The document discusses the importance of context in media interaction, advocating for tools that facilitate user-generated annotations and metadata to improve media search and remixing. It emphasizes the role of fandom and collaborative production in enhancing media experiences and creating context metadata that aids in the discovery and authorship of content. The proposed systems aim to enable incremental contributions from users to streamline the remixing process and enhance community engagement with media.
![Thanks! [email_address]](https://image.slidesharecdn.com/hypermediated-tv-15438/85/Hypermediated-TV-17-320.jpg)
![Daniel: the editor " I like to have something brand new every single video... [It's] a creative outlet for the things I want to do.”](https://image.slidesharecdn.com/hypermediated-tv-15438/85/Hypermediated-TV-23-320.jpg)
Hypermediated TV
- 1. Hypermediated TV Ryan Shaw unmediated.org UC Berkeley SIMS Yahoo! Research Berkeley
- 2. Context , not just content Early attempts at interactive TV and video focused on content But the contexts surrounding media are a better place to interact Which are we preserving when we talk about archiving TV?
- 3. Remix culture as a collective filter "Video search... will only work if we have millions of people informing our collective knowledge of what is worth our individual attention, and that can only happen if we can annotate, share, and remix video ." John Battelle, 4/27/05
- 4. Commons-based peer production Modularity Production can occur incrementally and asynchronously Granularity Contributors can make very small commitments Low-cost integration Robust systems for quality control
- 6. A socio-technical system Technological (networks, software)and social (people) components Design must be informed by an understanding of: how people interact with each other how people interact with media how people interact with technologies
- 7. Connect metadata producers and consumers
- 8. Media Streams Metadata Exchange Distributed production of metadata Fine-grained contributions, integrated incrementally Allow anyone to annotate any media quickly and easily Associate annotations with concepts to allow disambiguation
- 10. milkshake brings all the boys to the yard, and they're like, it's better Daniel “ I'm looking for an angelic female character , slowly looking up" Takuya "How about Ah! My Goddess, second season , episode 5 ?"
- 11. shot requests shot pointers music videos remix requests (ignored)
- 12. Shot Archive shot pointers & descriptions shot requests
- 13. shot pointers Shot Archive shot pointers & descriptions
- 14. music videos requests (granted!) Shot Archive Editing Rule Engine
- 15. Only store pointers to content Metadata is associated with pointers to temporal and spatial regions of multimedia content Remixers produce edit decision lists , not edited copies of content Has the Family Movie Act legalized this approach to remixing copyrighted content?
- 16. In conclusion It is still too difficult to find and author media Context metadata can solve this, if it can be produced on a large scale We can design tools to mediate a social process around media Fans' interactions with these tools will produce context metadata
- 19. Commons-based peer production Operating Systems Encyclopedias
- 20. AnimeMusicVideos.Org 323,000 members 59,000 AMVs by 20,000 creators 7,500-8,000 unique visitors per day Very active forums for discussing anime and AMVs
- 21. Studied anime fans to inform our design Catalyze creative ideas for innovative tools Use cases for prioritizing design concepts Working model of users , domains and contexts of interaction
- 22. User-centered design process In-depth interviews with remixers Online survey of 125 fans Media metadata experiment User personas Scenarios of how tools might be used
- 23. Daniel: the editor " I like to have something brand new every single video... [It's] a creative outlet for the things I want to do.”
- 24. Takuya: the expert " I'm a trivia freak, so it just completely makes me ecstatic if I know something somebody else doesn't.”
- 25. Kira: the newbie "hi, i'm new here and i've downloaded alot of AMV's in my time here. unfortunately, i am unable to make any videos as of yet; but, i have an idea for one... i really really really want to see this one done! please, please!! "
- 26. "Cal"
- 27. My contributions Reconceptualization of media search and media authoring Detailed data on how a community describes, looks for, and remixes media content Framework for understanding interactions with media in a particular domain
- 28. Next steps Begin prototyping interfaces for MSMDX tools Analyze queries for anime content from non-fans to see if metadata is useful outside community Identify other communities that might use such tools (video bloggers, indie journalists)
- 29. An explosion of on-demand multimedia
- 30. But we lack tools to manage media Multimedia content is still: Hard to search for Difficult to author with
- 31. People perceive characters, action, events Arthur Dent in his bathrobe , shortly before the Vogons destroy Earth
- 32. Computers perceive files, pixels, bits hhg-tlr3_m480.mov Quicktime 00:00:37:2
- 34. Analyzed shot requests "I need at least one anime where the character is eating pancakes , preferably out at a diner ." Object Setting Action
- 35. series "Cowboy Bebop" cast "Miaka" role "villain" body "muscular" object "pancakes" number "three" setting "enchanted forest" action "eating" genre "sci-fi" mood "scary" plot/theme "about a breakup" production "loopable"
- 36. Asked fans to describe shots Select temporal or spatial regions and add free-text annotations
- 37. Good match between requests and descriptions
- 38. Community-editable domain ontology Links episodes to characters , character names to physical appearances , etc. If someone labels a scene with "Vincent" associate that with "male, tall, long dark hair, tattoo" Match up requests specifying body types with descriptions specifying characters
- 39. Closing points anime fans collaborate to produce new works out of existing media we can design tools to mediate and improve this collaboration fans' interactions with these tools will produce semantic metadata semantic metadata can make it easier to find and author media
- 40. Fans are active cultural producers Already finding, describing, and remixing and sharing multimedia Dedicated/obsessed enough to spend time contributing Fandom is "serious leisure"
- 41. milkshake brings all the boys to the yard, and they're like, it's better milkshake brings all the boys to the yard, and they're like, it's better
- 42. milkshake brings all the boys to the yard, and they're like, it's better Music and lyrics provide the structure for the remix
- 43. Planning on paper Lyrics and timings are written out with notes on what shots to use
- 44. Metadata for credibility and reputation All data is linked to identity of its creator System for implicitly and explicitly rating contributions Allow building of social and cultural capital
- 45. MSMDX media player "View source" for multimedia A media player which allows jumping from remixed content to the original source from which it was taken Viewers can bookmark and tag audio and video of interest: tags are saved to central database
- 46. MSMDX editor plugin Use song lyrics or keywords to directly query for relevant scenes Selected clips can be drawn from personal collection or downloaded on demand for editing Labels that remixer adds to source content while working are uploaded to central database
- 47. MSMDX template builder Load a digital music file to create a template with timed lyrics Add high-level descriptions of scenes to be used Resulting template can be filled with specific content by a remixer, or automatically
- 48. Proposed data model ( data model diagram goes here)
- 51. Several steps to create an AMV Storyboarding or other planning Ripping source anime to a hard drive Finding and isolating desired clips Synchronizing with music Lip-syncing Creating visual effects
- 52. Some steps are tedious Storyboarding or other planning Ripping source anime to a hard drive Finding and isolating desired clips Synchronizing with music Lip-syncing Creating visual effects
- 53. Other steps are enjoyable Storyboarding or other planning Ripping source anime to a hard drive Finding and isolating desired clips Synchronizing with music Lip-syncing Creating visual effects
- 54. Interviewed 3 creators One pair interview, one solo interview, about one hour each All male, early 20s Interviewees were surprisingly reflective Expressed great interest in the project and staying involved
- 55. Ran a web-based survey
- 56. Analyzed scene requests "I need at least one anime where the character is eating pancakes , preferably out at a diner ." “ I'm looking for characters who have been betrayed/burned by another character romantically ." “ A clip of Tamahome with money and Miaka hitting him ."
- 57. Analyzed scene requests “ I'm looking for characters who have been betrayed/burned by another character romantically ." Theme/Plot
- 58. Analyzed scene requests “ A clip of Tamahome with money and Miaka hitting him ." Object Action Character
- 59. Using tags to find photos
- 60. Remix culture as a collective filter "Video search... will only work if we have millions of people informing our collective knowledge of what is worth our individual attention, and that can only happen if we can annotate, share, and remix video ." John Battelle, 4/27/05
Editor's Notes
- #2: I spent a lot of time this semester watching cartoons and music videos, because I think they hold the key to the future of the Internet. I'll explain why in a moment, but first let me tell you about a problem.