Media Semantics is your complete source for ready-to-animate Flash characters and supporting technologies. 

The Character Builder™ is an Integrated Development Environment (IDE) that lets non-animators create animated presentations involving characters and optional supporting visuals. The Character Builder can render content directly to the popular Macromedia Flash™ SWF format, as well as video.

Normally located on a web server, the optional Character Server™ can generate character animation on-the-fly for delivery to any Flash client. The Server also includes an AIML natural language engine for use in applications in which users converse with characters. The Server is a separate product from the Builder, and is intended primarily for developers.

The Character Builder comes with royalty-free, ready-to-animate Stock Characters. Additional characters and add-ons are available from the characters page, and custom characters can also be ordered from the service desk. Characters are based on an open artwork format that accommodates a wide range of bitmap and vector-based character styles.

Both the Character Builder and the Character Server support high-quality Text-to-Speech (TTS) engines as an alternative to recorded audio, however broadcast-use licensing for TTS can be difficult and costly. Optional Speech Packs are high-quality TTS engines that address this problem by specifically permitting the commercial redistribution of the resulting audio within a Media Semantics-generated media file.

Builder IDE (Integrated Development Environment) provides visual editing tools for specifying character animation, voice, and supporting media elements

Integrated media tools provide lip-syncing, TTS (Text-to-Speech), and audio processing functionality

Render-to-Flash feature lets non-animators create web greeters and eLearning content for delivery via the ubiquitous Macromedia Flash and Flash Lite players

Render-to-Video feature provides raw video output for use in traditional multimedia and mobile applications

Optional Server provides dynamic Flash SWF generation, server-side TTS, and an AIML (Artificial Intelligence Markup Language) natural language engine

Media Semantics offers some of the best technology on the market for creating applications involving animated characters.

The Character Builder uses a non-linear animation engine to create fluid animation from high-level commands such as Look, Point, and Say. Actions are independent: for example any eye or mouth action can be used with any hand action, to allow for a wide range of expression from a limited number of actions. The approach lets authors work independently from animators, and significantly lowers the cost of producing content involving character animation.

The Builder includes the ability to "lip-sync" audio files, so that a character’s lips and gestures are synchronized with its voice. Text-to-speech voice is typically used during development, but is increasingly an option for deployment, with the advent of high-quality speech engines.

Characters are created using pre-rendered 3D art or vector-based art, and can take advantage of a variety of compression, masking, and tweening techniques. A separate Character Builder Extensibility Kit includes additional tutorials and source files for creating custom characters.

The Character Builder uses a rich set of XML primitives to describe character behavior and supporting visuals. The markup language resembles HTML, but includes objects such as Stage and Slideshow, that reveal a high-level underlying application model.

Content authors normally use the Builder's Outline, Design, and Script views to specify content in a visual manner. For example, the Builder's Script View lets authors specify actions by selecting icons from pull-down menus. The resulting script can then be recorded in-place, or printed out for use by a voice artist.

Developers also appreciate the powerful underlying XML tag set, for use in creating database-generated content and end-user messaging solutions.

The Builder generates Flash files that take full advantage of streaming: they begin playing almost immediately, and then continue to play as they download.

Because character actions are interleaved with audio, the resulting streams can be very small. In many cases a presentation can begin playing within the first 10 kilobytes of download.

A simple API allows the surrounding web page or application to start, stop, and select among different presentations - the character smoothly reacts to these commands and maintains automatic actions, such as blinking or mouse-tracking, even when a presentation is stopped.

Builder-generated Flash files can be placed directly on a web page, or embedded within a traditional Flash or Director application. The files can also be embedded in a growing number of eLearning packages that support the insertion of Flash movies, including Adobe Captivate and Articulate Presenter.

A project can consist of a character and one or more presentations. An external container, such as an HTML page or a Flash application, can select among the authored presentations using a simple API.

Projects can also include supporting visuals - text, images, buttons, movies, and more. Supporting media can be controlled by a character, and synchronized with the character's voice and animation. For example a character can "point" at an image on a slide, then move it, or hide it. Similarly a character can glance at an embedded movie, watch it for a few seconds, then stop the movie, turn to the user to comment, then restart the movie.

The Builder can be used to create attractive slides and scenes, using features such as rich text, vector and raster graphics, and embedded Flash movies.

Dynamic Flash Generation

The Character Server can generate a Flash SWF stream on demand from an XML string. The XML string is typically assembled using a server-side script mechanism, such as ASP or PHP, in response to a request from a web client. The XML string can include information from the request, from a database, or from a web service. The resulting SWF stream is then returned to the web client as the web response, where it can be consumed just like any other Flash SWF file.

The Server can render anything that can be rendered by the Builder, including slides and supporting visuals. If audio files are not specified, then the Server will invoke a SAPI-compliant, server-based Text-To-Speech engine, and compress the output for inclusion directly in the Flash file.

The Server can be used by developers to create a wide range of character-based applications, including applications in which end-users can assemble their own character messages through a web-based interface.

Integrated AIML Engine

The Character Server's natural language capabilities allow users to ask questions using typed English input, and receive answers spoken in real-time by a cartoon or realistic character.

The Server includes an implementation of the popular AIML (Artificial Intelligence Markup Language) standard. By supporting AIML, the Character Server lets developers take advantage of a broad base of publicly-available "chat" rules. Conversation rules can also include character action tags, such as speech and animation tags.

The Character Server combines AIML, Dynamic Flash Generation, and Text-to-Speech into a single product, to provide a complete platform for "virtual assistant" and "learning agent" applications.