Abstract The multiplayer story engine experience represents a new an exciting form of entertainment. Recording this experience to video holds the same possibilities of entertainment

Download 24.13 Kb.
Date conversion08.03.2017
Size24.13 Kb.
Video Composition for the Multiplayer Story Engine

H. Adam Lenz

Department of Digital Media, University of Central Florida
4000 Central Florida Blvd. Orlando, Floirda, 32816


Abstract - The multiplayer story engine experience represents a new an exciting form of entertainment. Recording this experience to video holds the same possibilities of entertainment. But the format requires a video composer to jump through hoops in order to craft an entertaining, or even narrativly sound story. Thinking outside the normal role of the editor or cinematographer and a bit of ingenuity will be necessary

  1. Introduction

Editing for the Multiplayer Story Engine (MPSE) is a drastically different experience then editing for a more traditional story in film or video. The multiplayer story is a linear story, which puts participants in the roll of a player. It’s like, World of Warcraft in real life. A participant is given an identity and they must play that person for the duration of the event. In a more traditional story that is edited live, an actor is given a script, or loose plan that they must follow. Protagonists and antagonists are established before the filming. This presents quite a problem for a producer who wants to create a video type story in the fashion of what is generally and popularly accepted as entertaining.

This creates quite a problem for an editor who wishes to entertain the audience. In this article I will relate the narrative elements of an MSPE to the editor. I will discus the different capturing options that are available to a video composer. I will describe our own experiments in capturing an experience. Lastly I will propose a more professional system for capturing this type of experience.

  1. Narrative in the MPSE

The overall narrative in a MSPE is very similar to any traditional video story that most people are used to. Narratologists tell us that the common framework of a story is an arch encompassing a normal ordinary world moving into a tilt, or something that changes this ordinary world. There is a resolution, which then leads into a falling action and a return to a new normalcy. It can be seen throughout history that most stories follow this structure.

Where the MSPE diverges from the normal narrative is that there is any number of possible protagonists when looking into the experience from the outside. The MSPE more closely resembles the narrative in a MMORG, where the protagonist, in regards to each player, is the avatar that they control. The experience of the user describes the story arch, and so a story really just makes up the actions one person takes during an experience. I believe that the best narrative elements taken from all of the participants will not create an entertaining story for an outside audience. An editor looking to create a story must look to each participant to create one compelling story.

  1. Capturing and Composing

Capturing video in a MSPE is a bit like rolling a roulette wheel. Because there is any number of protagonists it is almost impossible to know whom to follow to create one consistent story. In order to functionally capture a video the cinematographer must plant many cameras in many different places, which hide in the environment and can be controlled to tilt and pan, zoom in and out.

A possibility of editing an MSPE is to programmatically edit the video before and during the experience. Simple editing of a video can be broken down in to events that describe the start and stop of a video clip in the master timeline, the in and out points of a video clip itself and a description of any filters, effects, or transitions that are applied to the clip. With these edit points, one can craft and edited video simply from video time code. By setting a start time that describes exactly where an even happens, and editor can simply enter the edit points into a form that describes the clip on the master timeline. The document can then be used to playback the video, edited in relation to this meta-data.

This would most likely require a digital asset management (DAM) system to ingest all of the video for tagging, and then a playback mechanism for the output of the video. Also this requires an equal amount or editors that there are participants in order to capture and edit each participants story.

Finite definitions of edited video are used in the Apple Interchange XML format used by Apple’s Final Cut Pro. This xml file allows an editor to exchange edits between different versions of Final Cut Pro. It is also possible for a programmer to use this data to visualize an edited video, and change the data using any number of means to edit XML data.

The final composition for this type of video would looks something like the movies, Go, or, Vantage Point, where each person has a story that goes on in the same linear time. The stories are played one after another and the time is reset for each story. The benefit to this is that the editors can choose weather a participant’s story will make it to the final composition.

If looking at the story of the entire group it is possible to edit the final product collectively in somewhat the same fashion as a soap opera. Most television shows that are produced live use control surfaces to mix a number of cameras together. If the group of participants in an MSPE is playing together you could follow the group like it was the protagonists. Character rolls, would stand out and this could be the person an editor follows more closely. The biggest problem with this is that an editor will not know or be able to follow every conversation that occurs in a group, where often-smaller groups of people stand out. Also, the camera work necessary to record such a story would require a camera located in the middle of the forming circle that rotates as each person talks. Multiple over the shoulder shots could be acquired from cameras located outside of the circle, but the cameras have to be able to move so that one of the participants isn’t accidentally standing in the way.

  1. Experiments

Our experiments immediately veered us away from attempting to programmatically edit the video. We did not have a number of capable editors equal to that of the participants, nor did we have the use of small, wireless, moving capture devices, so we decided to place 5 – 10 cameras in key discussion pits within the experience space.

Because we did not want wires all over the ground we used a streaming server to send the video over a local are wireless network to a mixing booth where one editor mixed all of the video in to an extended video. Due to networking equipment failure, our last experiment did not use these wireless cameras. The possibilities of problems created by using a slew of different components in a way that they might not be designed for, is dangerous and unpredictable.

We also used robotic surveillance cameras that had yaw, pitch, and zoom controls. These at least allowed the composer to control where the cameras were pointed. The processing of the video image into information packets when using the wireless video cameras, introduces latency. This meant that the robotic cameras would have to have a delay applied to them before mixing.

The final equipment configuration (e.g. Fig. 1) used four wireless video cameras with four Apple iMac computers, for video capture. The wireless video signal was sent to a streaming server and then viewed through a web page object in the video mixing program VDMX. VDMX allows you to “build your own virtual video studio, which is equally adept at event production, post production and motion design”[2].

The Streaming server we used is the Wowza Media Server [3]. This server has a built in function to record incoming streams as flash movies to disk. This is important because it later gives us the option to remix the video, although the person remixing would be editing flash video files, which are notoriously hard to convert to an editable file type. In order to tap into the server from VDMX I created a PHP page that allows me to add a stream name to the URL in the fashion: “http://localhost/stream.php?streamName=fuu”. Then I used VDMX’s ability to capture web content, browsing to the local page for each of the streams that were to be mixed in.

We also use two robotic surveillance cameras plugged into a mixing interface and then connected to a laptop computer to control the positioning. The mixing interface sent the final mixed video from the robotic cameras to the laptop running VDMX, where a 5 second delay was added to synchronize the video from the robotic cameras with the wireless cameras. A DVI to s-video cable was run to the computer that was acting as the video server for final capture.

After the experiment was complete the editor would cut the extended edit down into a short highlight reel. This did not create a compelling or entertaining story. The edit was required to be produced immediately after the experience, not allowing for the editor to review the content. It could be described as a bunch of people hanging out in a room, with 2 – 3 minutes of exciting, compelling story line. The best and most exciting part of our experiments was the previously tapped and edited news broadcasts, which severed as the tilts in our story.

Fig 1. The Video Flow of our Final experiment

In the last experiment we added a quartz composition that placed a small stopwatch at the bottom corner. This gives the audience a sense of time in the final edit.

  1. Room For Improvements

While out experiments were meant for testing the feasibility of recording a MPSE experience, we did not have any budget. With this in mind much of the equipment used was personally owned or borrowed. The equipment shown in figure 1 is valued at about $10,000[4]. Improvements on this system that I will proposed will cost considerably more and will encompass considerable more integration tine allowed in a typical college semester. This should be considered when attempting to capture a video of this type of experience.

A possible (yet somewhat improbable) camera setup would be a super-array of cameras like that seen in the movie, The Truman Show. In this movie the audience looks down at the protagonist, Truman (Jim Carry), while he lives his every day life. Smaller robotic camera systems like this are currently used for surveillance. In the case of the MSPE, an editor must watch every participant looking for a compelling story. It could be said that the editor watching the protagonist in the story is acting just like the user of a MMORG. The editor is omnipotent, but with out any control of what their protagonist will do. Given this story structure it most fitting to present the story in a manner the chronicles each of the protagonists separately, giving their viewpoints in a monologue after the event. In any case, an editor will only be able to monitor a small number of protagonists (1 – 3), if they wish to successfully produce a convincing story.

Fig. 2 A possible nametag bug with video camera, microphone, wireless antenna and location tracker

Another possible way to capture the MPSE experience is through bugging each participant with a camera that looks outward from the participant’s view, a microphone that picks up the local audio, a sensor that gives the editor a special position of the participant, and a wireless antenna to transmit the audio/video signal to the editor. As seen in figure 2, this use of technology would require very small, wireless components. It would almost certainly be necessary for the bug to be hidden in a prop given to the participants, in a nametag or set of glasses. A participant knowing about the bug would mean that they might act differently in the experience, but it would also mean that the participants would play into the camera, which helps to make a connection between the in-story characters, and the out-of-story audience.

Being able to robotically control the camera is very important to the composition of shot. Robotic cameras that are able to tilt, pan, and zoom will allow this type of composition, but this still does not allow the camera to move through the space. Recently, there have been some projects that attempt to place a live actor into a three-dimensional space. The Mona Lisa Project [5] attempts to fix the issues that make you weather man not look like he is part of the alpha channel background. These issues cause the actor to seem outside of the virtual setting thus not realistic or convincing. The Newtek Tricaster [6] also can place people inside of a virtual space in the same manner as a weatherman is placed in front of the weather map. This still does not solve our problem of being able to move the camera around in this world

Real-time three-dimensional worlds that allow an editor to place a filmed actor in the environment have recently been developed [7]. If you could create a 3d space by using many different video cameras or laser beams and processing the information into a virtual world, a virtual camera could be used to move around the space and get any angle necessary.Precise directional microphones would be needed in order to process any dialogue to the editor.

  1. Conclusions

Capturing and composing a compelling story from a multi-player story engine experience is difficult because of the possibilities created in the interaction of participants. The for editor, who is comfortable with editing a planned narrative, and the cinematographer, who is comfortable with creating the storyboarded shot need to step outside the comfort-zone in order to capture anything close to a convincing, entertaining product. Only through the invention, culmination and integration of rather advanced monitoring systems will the ability for a true narrative to emerge in a recorded format.


    [1] E. Branigan, Ed., Narrative Comprehension and Film, New York, NY, USA: Routledge, 1998

    [2] Vidvox Homepage. [Online]. http://www.vidvox.net/

    [3] Wowza Media Server Homepage. [Online]. Available:http://www.wowzamedia.com/

    [4] Prices from B&H Photo – Video – Pro Audio. [Online]. Available:http://bhphotovideo.com/

    [5] L. Blonde, M. Buck, R. Galli, W. Niem, Y. Paker, W. Schmidt,. G. Thomas, A virtual studio for live broadcasting: the Mona Lisa project , Multimedia, IEEE, vol 3, pp.18-29, Summer 1996.

    [6] Newtek Product Pag on the Tricaster. [Online]. Available:http://www.newtek.com/tricaster/

    [7] M. Petrov, A. Talapov, T. Robertson, A. Lebedev, A. Zhilyaev, and L. Polonskiy, Optical 3D Digitizers: Bringing Life to the Virtual World , IEEE Computer Graphics and Applications, vol 18, pp. 28-37, May/Jun 1998

The database is protected by copyright ©hestories.info 2017
send message

    Main page