WO2001035245A1

WO2001035245A1 - Method and apparatus for optimizing computer training files for delivery over a computer network

Info

Publication number: WO2001035245A1
Application number: PCT/US2000/031027
Authority: WO
Inventors: Guy Robert Hermann; Clive Gerald Hermann
Original assignee: Guy Robert Hermann; Clive Gerald Hermann
Priority date: 1999-11-12
Filing date: 2000-11-09
Publication date: 2001-05-17
Also published as: AU1762401A

Abstract

A method for optimizing training files (10) for delivery over a computer network (400). The method comprises capturing the full-motion video (12) in a vector graphics format; capturing audio (13) associated with the full-motion video (12); compressing the audio (140); and combining the compressed audio with the vectorized full-motion video (110). The method may additionally include further compression of the full-motion video (12), particularly in instances where such compression is necessary to deliver the combined video and audio at Internet speeds. A system for optimizing computer training files for delivery over a computer network is similarly disclosed.

Description

METHOD AND APPARATUS FOR OPTIMIZING

COMPUTER TRAINING FILES

FOR DELIVERY OVER A COMPUTER NETWORK

Background of the Invention

1. Field of the Invention

The present invention relates in general to optimization of computer training files for delivery over a computer network and, in particular, to a method and apparatus for optimizing full motion video, animation, and audio files for delivery over a computer network. The present invention has particular application in association with low bandwidth computer networks such as the Internet via POTS connections.

2. Background Art

Every industry has been impacted by the dramatic rise and incredible success of the Internet, and probably none more so than the education industry. Education and corporate training companies have been trying to find a way to deliver cost-effective and compelling software training across the Internet for at least the last 10 years. Many of the earliest attempts at on-line training failed because the content was too static and presented in much the same format as a book. Students were compelled to read descriptions of processes, and look at pictures, with everything presented in static page-after-page format. There were several attempts to add animations and audio to the content to try and liven it up a bit, but the costs of developing the animations and audio was prohibitive, and only the most well-funded training groups could afford to devote the resources necessary to produce engaging animations. Also, the animation and audio files tended to be very large, and thus took a long time to download to a user's computer across the Internet. While bandwidth is constantly increasing on the Internet, most users in the United

States (and more so Internationally) are still using dial-up modems to connect to an Internet Service Provider (ISP) over the plain old telephone system (POTS). Under such a system ~ even if the user takes advantage of the latest 56Kbps (bits per second) modem technology - an ISP (in view of bandwidth pooling, resource sharing, and cost considerations) is likely to provide less than 28 Kbps in actual data throughput. A 28.8 Kbps modem can accept up to 29,491 bits per second (or approximately 3.5 Kbytes per second), however approximately 10% of the bytes (8 bits per bytes) transmitted over a network are devoted to overhead or other activities not related directly to data being transmitted through the network. Thus, a 28.8 Kbps modem connected through an ISP is typically able to deliver something in the order of 3.0 K Bytes per second of sustainable data throughput to the user.

Some training companies have developed systems for training in large corporations. These systems are thought to deliver training at "Intranet" speeds (i.e. bandwidths of approximately 10 Mbps or greater) rather than "Internet" speeds. Most of these "full-screen" training solutions were only viable at speeds of at least 800 Kbps, and higher. While these solutions were acceptable for some corporate customers, many of such customers desire the ability to facilitate employee training from home, which again often requires delivery at standard modem speeds (i.e. less than 19.2 Kbps). In addition, most corporate Intranets span several cities and locations, and it is not reasonable to assume that every branch office or other type of location is connected by "high-speed" data lines. It is quite common for a branch office to be connected through dial-up modems, ISDN, or other economical connection types. Thus, making it highly desirable, even in a corporate training system, to produce low bandwidth computer training files that can be hosted on a wide variety of standard computers.

Of course, another factor that is often overlooked when discussing the importance of bandwidth to the online training environment is the impact of training to an existing network. While many corporate training clients have 100 Kbps network connections available to most users, much of this bandwidth is likely already being used. Thus, the addition of training video traffic could severely impact the overall throughput of even the broadest broadband network. Consequently, even in the corporate training environment, optimization of a computer training file would be beneficial; making online training more attractive to a large corporation.

As can be seen, there is a need for a realistic on-line learning solution, including the ability to deliver computer training files (including any combination of screen capture information, full motion video, animation and audio) at Internet speeds (i.e less than 19.2 Kbps in overall bandwidth requirements). Creating computer based training programs has tended to be time consuming, resource intensive and slowly deployed. In many instances, hours are spent in "post- production" for every hour of training developed, thus delaying the deployment of such training for weeks perhaps months. Furthermore, some training companies distribute computer training via physical distribution of removable computer media such as CD

ROMs. However, even with the advent of writable CD ROMs, creation and distribution of CD ROMs has an undesirable fixed-media cost and requires at least one additional days' time for delivery to the end users (assuming delivery via an overnight express company (another undesirable expense)). Still further, the use and distribution of such physical media makes revisions and "training material" updates difficult, which could lead to inadvertent use of outdated materials. Still further, distribution of such physical media may facilitate unauthorized redistribution of training content resulting in lost future revenues.

Thus, there is a need for a substantially "real time," low overhead, Internet deliverable training solution. For instance, if a training provider could record a seminar topic in real time and make it available to customers via the Internet that would improve distribution speed, avoid shipping overhead, minimize the use of potentially outdated content, and minimize unauthorized use of training materials, thus increasing a training provider' s market success.

It would also be desirable to format such training materials such that they can be hosted on "standard" web servers and received by "standard" browsers, such as Netscape Navigator, Microsoft Internet Explorer and the like. Several alternatives to standard web servers require expensive "streaming server software" (such as RealSystems' G2 Enterprise Edition or Microsoft® Windows Media™ Services), which have typically required training providers to pay by the number of simultaneous streams, or the number of Megabytes served to the end-user. It is desirable, of course, to avoid the investment in a proprietary streaming server solution or pay hefty bandwidth related fees to deliver training. It is likewise desirable to create a solution that does not require specialized server software and, thus, would be readily implementable on our end-user's own file servers, as well as the cheaper web hosting servers that are commercially available. Current examples of streaming server software, in order to communicate efficiently with a client, may require the opening of additional ports on a corporate firewall. Specifically, the initial exchange may take place on the standard HTTP port (well recognized as port 80), but the client will request additional UDP or TCP ports be opened up for subsequent communications with the server to try and ensure synchronization. Opening additional ports through a corporate firewall has the potential of decreasing security and increasing vulnerability.

It would also be desirable to configure the system for optimizing computer training files that does not require Java, ActiveX, or any other proprietary programming language or scripting language that could be perceived as a potential threat to computer networks of corporate clients. In particular, it would be ideal to find a solution that could get through all of the standard corporate firewalls and policies, thus, indicating the desirability of an "open-source," standard Internet protocol.

Summary of the Disclosure

The present disclosure is directed to, in part, a method for optimizing computer training files for delivery over a computer network. The method comprises capturing the full-motion video in a vector graphics format; capturing audio associated with the full- motion video; compressing the audio; and combining the compressed audio with the vectorized full-motion video. The method may additionally include further compression of the full-motion video, particularly in instances where such compression is necessary to deliver the combined video and audio at Internet speeds.

The disclosure further includes a system for optimizing computer training files for delivery over a computer network.

Brief Description of the Drawings

Fig. 1 of the drawings is a flow diagram of a method for optimizing computer training files for delivery over a computer network;

Fig. 2 of the drawings is a block diagram of the system components utilized in the present system; Fig. 3 A of the drawings is a graphical representation of the bandwidth utilized in downloading the computer training file of Fig. 3B; and

Fig. 3B of the drawings is a graphical representation of a preferred temporal playback of a computer training file for creation and delivery under the present invention.

Best Modes of Practicing the Invention

While the present invention may be embodied in many different forms, there is shown in the drawings and discussed herein a few specific embodiments with the understanding that the present disclosure is to be considered only as an exemplification of the principles of the invention and is not intended to limit the invention to the embodiments illustrated.

Simply, in large part, the present invention boils down to the ability to capture a full-screen movie, along with an audio narrative, and perhaps animation and compress these files down to an aggregate bandwidth suitable for use on dial-up modems across the Internet, preferably in the range of 19.2 Kbps to allow acceptable delivery to a 28.8 Kbps Modem, however, it should be understood that use of this technique could be limited for use with minimally a 56.6 Kbps modem across the Internet thus requiring an aggregate bandwidth in the range of only 40 Kbps. As the allowable bandwidth increases, the quality and size of the image can increase proportionally. Using the present system and method, it is entirely possible to compress an 800x600 size movie into the realm of 28 Kbps, and this plays well across the Internet when using a 33.6Kbps or faster modem.

Fig. 1 of the drawings is a flow chart of a method for optimizing computer training files for delivery over a computer network. As shown in Fig. 1, computer training file 10 may include one or more of screen capture data 11, full-motion video 12, audio data 13 and animation files 14. The screen capture data 11 may be created by Lotus ScreenCam, Microsoft

CamCorder , Hyperionics HyperCam, Microsoft PowerPoint Presentation Broadcast (for capturing MS PowerPoint only), Motion Works International CameraMan - PC Version, and Motion Works International CameraMan - Macintosh Version, MatchWare ScreenCorder, TechSmith Snaglt, CyberCam, Evolvable EvoCapture, RBCap, Video Capturix 2000.

In addition, there are a series of program that capture single screen images which may selectively be animated using animation tools such as Macromedia Director, Asymetrix Toolbook, Macromedia DreamWeaver, Macromedia Flash, Animated GIFs, etc.

Presently, the preferred solution utilizes Motion Works International CameraMan Macintosh Version. The Macintosh version is entirely different however, as one of its native output formats is Apple's QuickTime. Capturing the PC screen movies on a Macintosh in QuickTime format using CameraMan, we are able to get a file that can then be processed and compressed for delivery at Internet speeds.

Although Motion Works International CameraMan - Macintosh Version is preferred, with some minor additional steps, however, each of the foregoing screen capture programs will work well. For instance, Hyperionics HyperCam allows the user to adjust captured screen area, the number of Frames/Second captured, frame compression, and audio sample rates. However, HyperCam' s only available output is still an .AVI file, which cannot be processed and compressed as effectively for Internet delivery.

Full-motion video 12 may be created using any digital format but is preferably saved as an MPEG file. For instance, Adobe Premiere or equivalent type programs could be used to accomplish the creation of the full motion video potion of computer training file 10. Audio file 13 may be created using any number of techniques so long as the resulting file is digital audio, preferably in a .WAV format, although other formats can alternatively be utilized. For instance, Lotus ScreenCam can be used to capture audio. Similarly, Microsoft Sound Recorder, Cakewalk Pro Audio and Sonic Foundry Sound Forge can be used to record audio, but Motion Works International CameraMan - Macintosh Version is preferred.

Animation track 14 is preferably created using Macromedia Flash animation software. Flash files offer vector flexibility, compactness, and quality for high- performance image delivery on the Web. As binary files, Flash graphics are extremely compact and stream from any Web server. Hundreds of thousands of developers currently develop Flash-based content, and over 88% of browsers have the Flash Player already installed. The Flash Player is freely available for Macintosh, Windows, Solaris, Java, Linux, and IRIX platforms. Additionally, Flash already ships with most leading operating systems, media players, and Web browsers, including Windows 98, Mac OS 8.6, America Online 4.0, WebTV, Apple QuickTime, RealNetworks' RealPlayer, and current versions of Netscape Navigator and Microsoft Internet Explorer. Other animation programs such as Macromedia Director, Macromedia DreamWeaver, Asymetrix ToolBook, Microsoft PowerPoint, and many more may selectively be used, however none is currently integrated within the common media players such as RealSystems Real Player and Apple's QuickTime Player.

The interaction of video vectorization means 110 and each of the data capture/creation files is highly significant in obtaining the optimization levels desired for Internet transmission. It is appreciated by those skilled in the art, that vectorization as used in this application is the technique of representing video data as images based on lines drawn between specific coordinates. For instance, the standard video representation technique supported by WIN 9x API is pixel-based. Clearly pixel-based video representations are inadequate for producing the desired optimization levels. Of course, it is contemplated that those of ordinary skill in the art, having the present specification before them, will appreciate that other graphical representation techniques, such as fractal equations, may also facilitate optimization to Internet transmission bandwidth.

In a preferred embodiment, video vectorization means 110 is provided by performing screen capture on a Macintosh type Computer from Apple Computer Corporation of Cupertino, California (or an Apple clone type computer) running at least System 6 because the Macintosh operating system displays video in a vector format. Although the presently preferred approach uses a Macintosh computer to achieve vectorization, it should be understood by those of ordinary skill in the art that any computer (multi-purpose or perhaps even a dedicated graphics computer) can be used so long as the graphics are generated in a vector format, or any other efficient screen description language which will meet the requirements. In the Macintosh embodiment, to create training for WINTEL PC systems and software for use therein, a PC screen is displayed in on a Macintosh type Computer. There are various approaches to facilitate this functionality such as: Timbuktu Pro by Netopia of Alameda, CALIFORNIA(a cross-platform remote control/file transfer application); Virtual PC by Connectix of San Mateo, CALIFORNIA^ software based system capable of running WIN 9x, WIN NT, WIN 3.11, IBM OS/2 and Linux, among other PC based operating systems and applications written therefor);

SoftWindows/RealPC by Insignia Solutions of Fremont, CALIFORNIA(a software based PC emulator); Blue Label PowerEmulator by Lismore Software Systems of Dublin, Ireland (a software based PC emulator that facilitates use of Intel x86 Linux); Chameleon Host Link by NetManage of Cupertino, CALIFORNIA(a software based UNIX, xWindows, AS/400 (and various other mainframe systems) emulator); Exodus by White Pine Software of Nashua, NH (a software based emulator for Linux in an xWindow environment); and OrangePC 660 by Orange Micro of Anaheim, CAL_FORNIA(a hardware based PC emulator). Of course, other software based emulators and hardware- based emulators exist and still more are likely to be developed in the future that will facilitate capture of good quality video data from another platform on the Macintosh system. The currently preferred solution utilizes Timbuktu Pro because of its cross- platform capability and its ability to transmit images without pauses or losses.

There are various full-motion video players available for use on the Macintosh, such as Apple QuickTime Player. Again, the key is for the full-motion video to be played in a vector graphics format, such as on a Macintosh type computer.

As shown in Fig. 1, following video vectorization of the computer training file, the resulting file may be run through none, one or both of video codec 120 and video codec 130. As would be understood by those of ordinary skill in the art, following vectorization of the video portion of the computer training file, the resulting file may already have a sufficiently small bandwidth. The target bandwidth for optimized computer training file 200 (combined video, audio and animation) is something less than 19.2 Kbps. Thus, the target for the video portion is likely to be in the vicinity of 10 Kbps. So, between processing steps it may be desirable to preview information regarding the video file to determine the current bandwidth of the file. For instance, in Apple QuickTime Player, the "Movie Info" function will show file size and bandwidth. Other programs have similar functionality, such as Media Cleaner Pro and Adobe Premiere. Assuming that vectorization did not produce a sufficiently optimized bandwidth in and of itself, then the video file would be need to be additionally optimized by video codec 120 and/or video codec 130. It is appreciated that a codec (compressor/decompressor) is any technology for compressing and decompressing data. Codecs can be implemented in software, hardware, or a combination of both. Some popular codecs for computer video include MPEG, Indeo and Cinepak.). There are various methods known for compressing video and audio files. Some techniques are iterative others are Single-pass, and some of them are stand-alone (i.e. not usable with additional codecs). Some compression techniques are temporal (i.e. compress the file in time by replacing some full data set representations with a representation of the differences between frames in time), others are spatial (i.e. compress within a frame by removing redundant information and replacing bitmaps with either vector, shape, or other algorithmic representations of spaces and colors). Every streaming (and non-streaming) media delivery technique relies on the ability to compress the video and the audio streams down to an acceptable level. Each compression method has strengths and weaknesses, which may even be dependent on the data set, data type, etc. Some compression methods introduce more distortion than others. Some methods compress slowly and decompress quickly, and vice versa.

In a preferred embodiment, the Apple QuickTime Player (MAC version) in graphics codec mode is used to implement both video codec 120 and 130. Again between compression iterations it is desirable to check bandwidth of the resulting file before continuing on to perform another compression. Simply put, the greater compression, the greater decompression required at the client computer. Consequently, it is desirable to compress only so far as to facilitate Internet speeds, so that the resulting file can be run from a slower computer (i.e. i486, WIN 3.11, 28.8 Kbps Modem). It has been determined using the Apple QuickTime Player that a third compression iteration and beyond produces only a marginal improvement.

As is understood, codec software such as Apple QuickTime Player automatically selects the video compression techniques to be used in association with a file based on an initial analysis of the file contents. Thus, it is possible, even likely that on subsequent passes through QuickTime Player different iterative compression techniques are used to compress the file. As would be understood to those of ordinary skill in the art, it is equally possible to implement a different codec in hardware, software or some combination thereof. No matter the type of codec used in the present invention, the key is the use of iterative compression techniques such that on subsequent passes, the file is further compressed while maintaining acceptable loss-levels.

If compression alone cannot provide sufficient bandwidth shrinkage, it is possible to change the parameters of the various video capture programs to produce further compression. Additionally, the training instructor could modify the demonstration techniques by slowing down mouse movements, change screen background, minimize length and/or movement within inserted full-motion video clips or other various techniques for minimizing the initial graphics file bandwidth and size.

Returning to computer training file 10, we see that audio data 13 should be preferably compressed along its own processing path, however, as would be understood by those of skill in the art, it is possible to compress the audio and video using the same software processes perhaps even at the same time.

As an example, the audio data 13 may be optionally processed along a separate processing path including an audio processor 140 and audio codec 150. The audio processor 140 may selectively be software that processes the audio file 13 to optimize it for transmission over the computer network. An example of audio processor 140 is the software program Media Cleaner Pro. The audio codec 150 is preferably a compressor/decompressor that is either implemented in hardware or software.

As an example of Audio Codec 150, the Apple QuickTime Player (MAC) may be employed for Audio compression because it has two native codecs that may selectively be used for producing audio streams. The first is the Qdesign Music 2 which had been designed primarily for music (as opposed to voice compression). Q design allowsenough control of the compression to allow selection of a preformed bandwidth. It is appreciated that some of the choices include 10 Kbps, 16 Kbps, 20 Kbps, 24 Kbps, etc.

The second audio compressor included with Apple's QuickTime Player is the Qualcomm PureVoice codec which has been designed for voice compression. As an example, the Pure Voice codec is able to compress voice audio down to 8Kbps while still having acceptable performance.

First Example Procedure

To create a new screen-movie for use at Internet speeds, we conducted the following example process:

• Use one of the many techniques described above for getting the image, desktop or application that you want to capture onto a Macintosh computer screen (Mac OS, Virtual PC, Timbuktu, Blue Label PowerEmulator, etc.)

• Launch CameraMan on the Macintosh and set it up to capture only the window in which the action will take place. o Set the Format to "QuickTime", set the Frames/Sec to 1, and set the Capture Mode to "Capture Area". o Now adjust the Capture Area so that it captures exactly the screen area of the window you want to record. o If you want to record the audio track at the same time, make sure that

CameraMan is setup to "Record Live Soundtrack". o Verify the hot keys that CameraMan will recognize to Start, Stop or Pause the capture, we use F9, F10 and FI 1 respectively as they do not interfere with any of our software emulation programs on the Macintosh, but it is easy to select different keys if this is necessary. o Take note of where the file will be created when you capture with

CameraMan, you may want to change the directory where the output will be saved. o Hide the CameraMan application so you will not be capturing any part of it when you record your screen actions.

• Switch back to the window you want to record, and press F9 to begin the capture process. • When you are finished with the segment you want to record, press F10 to stop the recording process

• CameraMan has now created a file that has been saved as a QuickTime movie. Lets call this file the "Raw QuickTime Capture.mov" file. • Launch Apple QuickTime Player . o Open up the "Raw QuickTime Capture.mov" file. o Bring up the Export dialog box. o Select the "Movie to QuickTime Movie" option as your main starting point. o Click on the Options button to verify your settings. o Video settings should be "Graphics codec", "Color", "Best", "1 Frame per Second", and Key Frame Rate of 15. Verify that there are no filters in use, and that the size is set to "Use Current Size". This ensures that there will be no lost data in the compression process. o Audio settings are dependent on whether you recorded voice or music along with your narration. If you only recorded a voice narration, use the "Qualcomm PureVoice" codec at 11.025 Khz, 16 bit, mono, half-rate for streaming and optimize compression for streaming. If you music has been recorded, select "QDesign Music 2 at 44.100 Khz, 16 bit, mono, 16 kbps" as your audio compression setting. o Export the screen-movie to another file named "Compressed QuickTime Capture.mov".

• Open the "Compressed QuickTime Capture.mov" and check its Movie Info to verify that has been compressed to an acceptably small size for Internet delivery. • The "Compressed QuickTime Capture.mov" can now be loaded onto a Web

Server and embedded into any HTML page using the standard Apple QuickTime HTML embed command. Of course the same basic technique can be used with multiple source files to create the same result.

A Second Procedure Example

In this example, pre-existing full motion videos and Lotus ScreenCam files (continuous screen capture program) were converted for use as low as at Internet speed, as follows:

• On the PC: o Use Lotus ScreenCam to export a .WAV file o Process the .WAV file through Media Cleaner Pro o Use QDesign2 Music2 at 2.5 Kbps or the Qualcomm PureVoice codec to compress the audio track o This creates a .MOV file with the audio that can then be used to merge with the video file

• On the Macintosh: o Run Virtual PC on the Macintosh in 800x600 mode o Use Lotus ScreenCam to open the .SCM file o Running CameraMan on the Macintosh, capture the screen show in Virtual PC using: o 16,44,816,644 as settings for the Capture Area o Do not record sounds o This creates a QuickTime Movie .MOV file

• Another way - apparently more reliable on the iMac is to use Timbuktu o Set it up so the PC screen is at 1024x768 in thousands of colors o The Macintosh is also set to 1024x768 in thousands of colors o Position the PC window on the Mac in such a way that Cameraman records just the ScreenCam portion o We used 40, 840, 48, 648 for an 800x600 ScreenCam file o And we used 120, 760, 108, 588 for a 640x480 ScreenCam file o Capture the ScreenCam' s to the Macintosh using the CameraMan technique described above • Open the .MOV file in QuickTime for Macintosh o Export the file using the Apple Graphics Codec at 1 frame per second with a keyframe every 15 frames

• On the PC: o Merge the Audio and Video files together using QuickTime o You must hold down ctrl-alt and then select ADD from the Edit Menu when you paste the tracks together to get the desired merge result - otherwise QuickTime Player appends the new track to the end of the existing track.

A Third Example Procedure

We also developed a technique to take our existing videotapes and convert them into a format suitable for the Internet. To convert a videotape (or any other live-action video sequence) for use on the Internet:

• Capture the movie from video into a digital format compatible with the PC. o Capture the movie into a digital format on the PC and save it as an MPEG file. This could be done with any number of off-the-shelf PC products designed for capturing video. We primarily use a Sony VAIO with a built-in Fire Wire port (IEEE 1394 - this is sometimes given proprietary names by various manufacturers such as Sony i.Link) to capture directly from our camera format to the PC. • On the PC: o Use Xing Mpeg Encoder to convert the audio portion of the .MPG files to MP3 o Use Electronic Cosmo's MPEG Suite to convert the .MP3's to .WAV files o Process the .WAV files using Media Cleaner Pro o Use the same method used to process the ScreenCam .WAVs o This leaves you with a well-compressed QuickTime format .MOV audio- only file.

• On the Macintosh: o Open the movies in Macintosh QuickTime o Export the movie using:

^■ Sorenson ^■ 30 fps

^■ No audio

• On the PC: o Process the .MOV files using Media Cleaner Pro with the following settings: ^■ Sorenson

^■ 7.5 fps

^■ 3.5 kbps

• On the PC: o Merge the Audio and Video files together using QuickTime

The resulting files will be perfectly acceptable for use across any network medium, including of course the Internet. The final bandwidth requirements will be entirely dependent on the size of the original video capture, its overall length, and the amount of screen changes that take place in the movie, but average expected bandwidth requirements for 640x480 and 800x600 ScreenCams should be approximately 19 Kbps and 25 Kbps respectively. We have seen results as low as 11 Kbps and 17 Kbps (respectively) but these should not be anticipated to be the norm. Most of the alternatives that have been discussed herein will work with essentially anything that can be rendered on the screen and can be captured at real-time speeds, and played back in movie form at bandwidths that are acceptable to the vast majority of Internet users. Ability to develop or modify training modules rapidly - essentially a "real-time" production pipeline wherein screen movies can be recorded "live" and be processed and placed online within minutes following the conclusion of the recording session. Universal playback - using QuickTime 4 Player as our format (only temporary as we are developing our own codec so we will be able to use competitive formats as well). No streaming server requirements - delivery performance can be enhanced or optimized by using one of the many streaming technologies on the market, but it is not a requirement for our system. Our files are so compressed that they can be delivered using the standard HTTP protocol with no ill effect.

Fig. 2 shows a block diagram of the components utilized in association with the present system and method.. As shown, the present system and method are operated in conjunction with a general-purpose computer system such as an IBM compatible, Apple, or other equivalent workstation type computer. This general-purpose computer system must include a vector graphics engine, which may be implemented in hardware, software or some hybrid thereof. The general-purpose computer may further include a digital signal processor, such as those available from Motorola to facilitate processing of the digital signals. The general-purpose computer may also include a network interface to facilitate transmission of the optimized computer training file to other computers. The general-purpose computer may further include a removable storage media system, such as a rewritable CD ROM drive as an alternative distribution method. Ultimately, this general-purpose computer or in some instances another general purpose computer will act as a file server to host the optimized computer training file for use by various clients.. Computer network 200 is any computer network that allows multiple computer systems to communicate with each other such as a Local Area Network (LAN), Ethernet or the Internet. The client is preferably a general-purpose computer system such as an IBM compatible, Apple, Unix type workstation, or equivalent. Both the client and server computer contain a network interface that allows for communication with computer network 200. This network interface may selectively be any Internet capable software program such as Netscape Navigator, Microsoft Internet Explorer, Mosaic, or the equivalent in combination with various hardware. In the client computer for instance, it is presumed that most clients will be using a 28.8 Kbps Modem as the hardware portion of the network interface. However, as discussed above, the present method and system provides advantages even where the client computer connects to computer network 200 at Intranet speeds.

Fig. 3 A depicts the download of a training file from the server computer to the client computer, while Fig. 3B shows the contents of the display. These figures have been matched in time to show one possible approach to programming using the present system that further facilitates the provision of an optimized multimedia experience. As shown in Fig. 3a, the training file begins downloading at tO. From tO to tl, an animation file is downloaded to the client. As depicted this animation file has low bandwidth and so downloads quickly regardless of the network parameters. In conjunction with the quick download, however, the animation is programmed to play beyond tl through until t2. Between times tl and t2, the first multimedia portion(s) are being downloaded from the server to the client, which requires significant bandwidth and processing power. Because the low bandwidth, low processing power animation is being shown during this resource intensive download, the visible and audible jitter generally associated with the commencement of such intensive video files is unnoticed by the user. By the time the prolonged animation sequence has completed the resource-intensive download has completed. In this way the user's experience is further enhanced.

Claims

WHAT IS CLAIMED IS:

1. A method for method optimizing computer training files for delivery over a computer network, the computer training file including at least full-motion video, the method comprising: - capturing the full-motion video in a vector graphics format; capturing audio associated with the full-motion video; compressing the audio; and combining the compressed audio with the vectorized full-motion video.

2. The invention according to Claim 1 wherein the method further includes: - determining whether bandwidth of the vectorized full-motion video is sufficiently small for transmission at Internet speeds; and compressing the vectorized full-motion video if its bandwidth is not sufficiently small for Internet speed transmission.

3. The invention according to Claim 2 wherein the method further includes: - determining whether bandwidth of the compressed vectorized full-motion video is sufficiently small for transmission at Internet speeds; and compressing the compressed vectorized full-motion video if its bandwidth is not sufficiently small for Internet speed transmission.

4. A system for optimizing computer training files for delivery over a computer network, the system comprising: means for capturing the full-motion video in a vector graphics format; means for capturing audio associated with the full-motion video; means for compressing the audio; and means for combining the compressed audio with the vectorized full-motion video.

- 1!