Motivation

1
Motivation

• Gaming is becoming more mainstream and more
  diverse
• Youve heard this over and over, but its true
• Gaming industry is continuing to grow at a rapid
  pace
• Gamers are no longer the stereotypical
  demographic of teenage boys
• Men, Women, Children, Teenagers, Baby Boomers, etc

2
Motivation Cont

• Gaming is no longer restricted to a few friends
  in front of an Xbox or the freshman dorms LAN
  party
• Sponsored Gaming Competitions
• Network enabled consoles
• MMO games, RPGs or otherwise
• Virtual worlds Is Second Life a game?
• The G4 Channel (I knowI know)

3
So What?

• Widening demographics, increased connectivity,
  more mainstream exposure, what does that mean?
• More for the game industry
• Many of the big companies are going to stick to
  what works, production of a blockbuster game is a
  multi-million dollar undertaking nowadays.
• Still looking for the next big thing though
• More opportunities to explore new ideas
• No longer required to pander exclusively to the
  hardcore gamer
• Different types of gamers means different
  requirements, different market niches

4
Really New?

• Microsoft obtained a patent for a system that
  allows others to act as spectators in online
  gaming matches (to be used on Xbox Live)
• The idea of watching games is certainly not new
• However, with so many interested in gaming
  nowadays, there exists a potential market
• Network connectivity is becoming standard in
  consoles
• Microsoft may be evil but theyre not stupid

5
Get to the point already

• We have opportunities to try new things because
  gaming as we know it is changing
• Virtual Cameramen for gaming sessions
• Ever been a spectator in a Quake match?
  Sometimes it gets boring, let someone else do the
  work, or at least provide you with more options
• More World of Warcraft videos on YouTube than Id
  care to consider
• Why not give the players some easy to use tools
  to support their endeavors?
• Second Life isnt just a game
• NMC Campus, Shops, Theaters, Clubs, etc
• It isnt necessarily about the gaming
  experience anymore

6
What have we seen already?

• The idea of a cinematic camera that adapts the
  rules and styles of cinema to the world of gaming
  sounds great on paper
• Unfortunately worlds are dynamic. On the set of
  a movie, the director knows exactly what will
  happen. Not so much in a game/virtual world
• Gamers will only appreciate something like this
  as long as it doesnt interfere with their
  gameplay
• Combine these two, and it turns out that
  developing a useful automated camera system is
  nearly impossible
• Also, the rules of cinema arent so much rules as
  common practices, and are difficult to capture
  quantitatively. This makes them difficult to
  translate into code.

7
Camera Systems

• Automated Camera Systems for games are HARD!
• Drucker, Bares, Halper, Cohen, He
• Primarly constraint satisfaction systems

8
Drucker

• Developed the CINEMA and CamDroid Systems
• CINEMA Attempted to define a language for a set
  of simple camera primitives that could be used by
  other applications to create high-level camera
  control
• CamDroid Defined camera modules, made up of a
  set of constraints over a period of time.
  Execution of these modules involved constraint
  satisfaction and optimization
• PhD Thesis Intelligent Camera Control
• Culmination of other works, he proposes an
  intelligent camera agent that is capable of
  performing various tasks
• Classifies different types of domains of action
  that would take place in a virtual environment
• System based on optimization of constraint
  satisfaction

9
Drucker Cont

• Ultimately Drucker showed his system to be useful
  in five different environments with five
  different action domains
• However his system still faced several of the
  problems that most automated/intelligent camera
  systems face
• How to decide on the constraints to be used
• How to make it simple for a non-programmer to use
• The language used to define camera constraints
  and modules seemed to be fairly simple, but I say
  that as someone who has been programming for 10
  years. I doubt a non-programmer could make any
  sense of it.

10
Cohen, He

• Developed CamPlan, and idiom-based automated
  camera system
• An idiom is a collection of expertise that
  defines a given situation
• Ex A conversation between two actors. Classic
  rules of cinema define that conversations happen
  according to certain rules and in a certain way.
  A conversation between two characters will likely
  use an over-the-shoulder shot and/or
  shot-counter-shot sequence
• An idiom based system is in some ways simpler
  than the constraint-based systems but is
  unfortunately limited to the idioms defined by
  the system. When a situation presents itself
  that does not fall into one of the idioms, the
  system may fail

11
Bares

• Developed several constraint based systems for
  automatic camera control
• Conceived an interesting interface in which the
  user defines the constraints for a particular
  shot through a storyboard interface
• Provides a visual interface that allows the user
  to more easily define a particular shot
• Unfortunately this system was only used to
  generate static shots

12
Difficulties

• Developing a fully automated camera system is a
  extensive and difficult undertaking
• Many of the systems developed perform admirably
  in controlled situations, but would likely fail
  in the dynamic and demanding worlds of todays
  games
• Constraint optimization also tends to be
  computationally expensive, which is not something
  that many of todays already taxing games can
  afford
• Most of these systems require a large amount of
  data from the surrounding environment that may
  not always be available
• Gamers want a camera that will not interfere with
  what they are doing.
• An automated camera along the lines of what
  Prince of Persia implemented in order to show the
  player the room full of puzzles was well done.
  It took control from the player but only during
  non-critical moments and always in an effort to
  provide the player with useful information

13
Spectators!

• Drucker is actually one of the Microsoft
  engineers behind the spectator system that
  Microsoft patented
• Drucker et al believe that gaming can and will
  become a spectator sport
• This is a new way of looking at games
• Dont just focus on the player, but also the
  potential spectators

14
Canovaccio Background

• Black Box Theater project in Second Life
• Virtual Theater and Stage
• Virtual Actors
• Based on Commedia dellArte
• Mocap from life actors
• Masks fashioned after this style

15
Virtual Cameras

• Advantages of virtual cameras
• Can easily be placed anywhere at any time
• It is possible for multiple users to look through
  the eye of the camera simultaneously
• Virtual Cameras are cheap, they can be set up in
  multiple locations with multiple settings quite
  easily

16
Camera Tool

• To be used primarily as a tool for the actors
• Student/Teacher roles
• Allows an instructor to demonstrate certain
  theatric techniques easily
• Controls the students gaze and focus
• Can easily switch from watching the instructor to
  watching the student(s) as they each give it a
  try
• Practice
• Actors can use the tool to observe themselves
  from multiple perspectives as they practice the
  various movements that are central to commedia

17
Camera Tool Practice

• Provides two primary modes to the user
• Mirror Mode as though the user is looking in a
  mirror
• Profile Mode provides the user with a profile
  of their avatar
• Use of basic film language
• Close-up, full-shot, long shot, etc
• Allows the user to combine their selected mode
  with any of the basic shots and to easily switch
  between them

18
Camera Tool Group Mode

• The Master Control allows the user to generate
  items to track the actors/students for various
  purposes
• Establish a master shot involving 2 actors
  (currently only two)
• Identify the students and control what they see
• Specify a particular student to become the focus

19
Camera Tool and Commedia

• Working to link various prims with mocap
  animations. Different masks when equipped will
  allow different animations
• These animations can be linked with particular
  shot types as well
• A close-up would work well for a wink whereas a
  long shot would work well for a cartwheel.

20
Camera Tool Continued

• Still a lot of work to be done on the tool, but
  the foundation is in place
• The biggest hurdle at the moment seems to be
  tracking. Using the SL API, tracking the actor
  in question is very choppy.

21
Camera Tool Future

• Id like to fully develop the tool to allow
• Shots of selected avatar(s), not just the current
  user
• Selected control by the master of what each
  participating avatar sees
• Wide area security cameras showing the area of
  interest from various angles and distances

22
Camera Tool Future

• The tool (or at least the ideas behind it) could
  be valuable from a spectator perspective as well
• Audience members could be identified as
  spectators and if they so wish, they can watch
  the play/dance/game through the tool.
• The tool could then control what the audience
  sees through
• Basic rules for animation/shot combinations
• Actors currently flagged as important
• Automation is not always good. Todays sports
  broadcasts are controlled from the booth where a
  director surveys the various shots available to
  her and decides on the shots that are currently
  best for the audience
• A well-defined, easy to use interface for the
  director would require human interaction but
  would ultimately result in what would likely be a
  better experience for the spectators

23
In Closing

• Tools like this can be valuable in situations
  where gameplay is not the focus
• Second Life and other virtual worlds. There is
  no game/points to lose, so the players will
  tolerate and may be interested in alternative
  cameras
• There are spectators not playing the game but
  would rather be entertained by it, so tools to
  show them relevant events in an entertaining
  fashion can be valuable
• Some classrooms are moving into virtual
  environments and tools like these could be used
  to facilitate the teaching/learning process
• Virtual theaters, clubs, gameshows, etc. Any
  virtual environment/event that might generate
  spectators could benefit from tools that make
  that can make that environment/event more
  interesting to those watching