CS7497 Spring 2013

DART: A Toolkit for Rapid Design Exploration of Augmented Reality Experiences

Designers have very strong skills on how to use popular environment for displaying 2D content or creating animation using script. But when they have to create an Augmented Reality environment (which try ti augment a person’s perception of the world wuth computer generated information), they encounter an issue. Even if they are familiar with scripting language, AR tools are developed by non-designer engineer and then use complex language such as C++. Hence, they cannot use those tools, even if those ones have been design for them. AR systems make an intensive used of tracking. Unfortunately, libraries that offer computer vision tracking are complex and there are a lot of issue when a developer tries to use several of them into a single program. Moreover, there is actually no current way, or very inefficient way, to manage information about the real world. AR is all about enhancing the Real-World. Then, an AR system has to manage carefully those information.
The DART team has worked closely with designers. Hence, they have been able to create a product that fit the designers’ needs. Darts is based on Macromedia Editor and offer simple scripts to create 3D content from 2D storyboard. DART also offers a complete access to physical sensors. Hence, the tracking can be done easily, even in the case where several sensors are implied.

DART has been design to be as close as possible as DIRECTOR. The idea is if you are familiar with director, you are already familiar with Dart. Director uses a stage, which is basically a scene), casts, objects where contents are stored, a score (timeline), spirits (members of cast) and scripts. Darts reuses that design by using stage and spirits on where scripts are bound.

The Effect of Latency and Network Limitations on MMORPGs

MMORPGs are popular multiplayer games. generally, they offer a giant open world and interactions between hundreds of player. Because the workload can be handle by only one server, that world has to be splitted into several servers. Currently, there are two different methods to do that. The first is the really basic. The world is divided into several areas. A user can go from one area to another freely, but there is a loading time. The transition between areas is not transparent with such design. A more elegant way is create a border area that create a link between two areas. Hence, the transition is transparent for the player but there are several drawbacks.  Indeed, the workload is more important because both adjacent areas have to be used. Another drawback is the difference between the player and PNG. A PNG is restricted to its area when a player can freely go to one area to another. Hence, it can use those limitations to make easiest some fights or moves.

The study also tries to understand the effect of latency. For that, three situations have been created, based on the MMORPG Everquest2. First, movement and combat are studied. Authors estimated that a latency greater than 500ms would have been the limit. It appears in reality that the breaking point is around 1250ms. This can been explain by an internal mechanism of Everquest. Indeed, even if the character is currently doing an action, the player can planned the next action (the last key pushed is remember in a waiting queue). Because each action takes about 1s, this mechanism allows the player to still have a great experience even with 1s latency. The second situation studied is the effect of latency on teamplay. Here again, the authors estimates the latency breaking point around 500ms. The test shows that the limit is also about 1s. The last situation evaluates the display accuracy of a player in the environment. Latency can have a huge impact. Indeed, first the player’s computer request the game server for the action. This server computes the movement and if it is legal, sent an acknowledgement. On the player computer, the movement is displayed only after the acknowledgement has been received.

Alice: Lessons Learned from Building a 3D System For Novices

Alice is a new tool for describing the time-based and interaction behavior of 3D objects. It has been designed according two assumptions. First, a new audition is rising. This audition is populated by people interested in 3D displays but without necessarily having a strong mathematical and programming background on that area. Second, logic is still needed to perform non trivial interaction. Then a kind of if-else statement has to be created, but with the requirement to be accessible.

A typical Alice workflow has two phases : creating a scene and adding scripts. To test the software, the authors pair subjects and ask them to follow tutorials. The good thing with pairing is that it makes the subject to talk to each other. Then, it is easy to take notes about conversations and hence, understand their feeling about the software. According to their work, for non-specialized people, it is more convenient to use forward/backward/left/right command than just have access to X,Y,Z. Generally, the idea is to hide things under basic and understandable concept or object. Optional parameters can be used to offer both easy and complex interactions. When the user feel the need or the motivation, it can use optional parameters to enable the full capacity of the software.