Industry Leading Patent Pending Compression Technology

Broadcast International
CodecSys Codec Switching Proof of Concept

February 24 , 2004

INTRODUCTION


CodecSys™ is a multi-codec system that allows companies to create high quality video and audio transmissions at a fraction of the bandwidth traditionally required by other streaming technologies. CodecSys™ manages libraries of standard and specialized codec’s and dynamically changes codec’s and codec settings to produce the best possible full-motion video using the least amount of network bandwidth. This codec switch can happen on a scene-by-scene or even frame-by-frame basis. CodecSys delivers full-screen, full-motion video at the lowest possible bandwidths for video-on-demand, live broadcast, videoconferencing, and other applications.
 
In January 2004 Broadcast International contracted with KeyLabs™ to perform an independent assessment and evaluation of its new multi-codec operating system, CodecSys.
 
The CodecSys technology represents a paradigm shift from the traditional single generalist codec compression approach to a codec team approach, which uses many different codecs and algorithms in the compression process. The codec system selects on-the-fly the most suitable codecs available from the various codecs stored in its library to compress a single video stream. As a video frame, or a number of similar frames (a scene), is compressed, CodecSys applies the codec from the library that best compresses that content. CodecSys repeats the process throughout the video encoding process, resulting in the use of numerous codecs on a best performance basis. The resulting file is typically substantially smaller than situations where a single codec compression method is used.
 
The first applications to employ the CodecSys technology are Video Conferencing and Video-on-Demand (VOD), which were used to produce and validate the results in this test report. The CodecSys technology is anticipated to have many other product applications including video broadcast, security monitoring, information/education, telemedicine, entertainment, gaming and archival storage systems.
 

Test Objectives

The specific objectives of this project were to:
  • Verify the utilization of multiple codecs on the fly in the compression of a single video stream.
  • Verify the ability of the player software to receive and decode the multi-codec stream.
  • Verify the use of dynamic setting switching and its ability to function in a pre-determined bandwidth range at full screen size with 24 – 30 frames per second frame rate.
  • Compare the CodecSys compression file size to the same content in a MPEG2 video format.
  • Verify that the video on demand transmission stream complies with predetermined bandwidth rates.
  • Verify the CodecSys Interact video conferencing system complies with pre-determined bandwidth ranges of 384 and 64 kbps at full frame and full screen rates.
  • Verify the use of dynamic setting switching and its ability to produce full screen, full motion video output.

Test Results Summary

Method and process of determining utilization of multiple codecs

KeyLabs tested CodecSys by streaming various Video On Demand (VOD) clips and viewing real-time video conferencing. VOD clips were verified for file size and compared with native DVD file formats. Network bandwidth utilization was measured for both VOD and video conferencing applications. Visual inspection of the video streams on the client computer verified that videos played without skip or pauses and that appropriate quality was maintained throughout the entire clip.
 

Multi-codec Verification

KeyLabs verified CodecSys’s multiple codec switching technology by viewing two multi-scened video clips taken from commercial DVDs. Each clip was encoded using four different codec’s. Different players were used with one of the codec’s missing. The entire clip was first verified using a player that was equipped with all required codec’s. Then subsequent players with one codec missing were used to view the same clip. As expected, the CodecSys-optimized video clip would play until it reached the section of the clip that was encoded with the codec that was missing from the player. At this point the player did not terminate when it encountered the missing codec; rather the video display was halted (but the stream continued) until it reached a codec that was recognized by the player. It then continued displaying the clip.

Video Stream Throughput Validation

Network throughput of the different video streams was measured using Ethereal, the multi-platform, open source network protocol analyzer. Ethereal was configured to filter all packets except those created by the server and client connection. Once the packet capture had begun, the video stream was started. The resulting packet trace was then analyzed to determine throughput characteristics of each of the video clips.

Pre- and Post-Compression File Size Verification

Two video clips were used to measure file compression capabilities of the CodecSys engine. The clips were scenes taken from the commercially available DVDs Executive Decision and Gladiator. The uncompressed (native DVD VOB) file sizes were 238 Mbytes and 362 Mbytes respectively. Both clips were encoded at a target bit rate of 1024 Kbits per second. At this rate, we saw no perceptible difference between the native DVD format and the CodecSys-optimized video stream. Resultant compressed file sizes averaged about 1/6th of their original DVD file sizes.
 

 

Overall Observations

Throughout the testing KeyLabs test engineers saw no visible difference between the native DVD video stream and the CodecSys-optimized video stream that was encoded at 1024 Kbits per second. Both streams were displayed in full-screen mode. Video and audio were both checked for skips, pauses, and distortion throughout the duration of the clip.
 
The 512 Kbit per second encoded video streams were very good as well. Minimal degradation of the video display manifest itself in slight pixilation, that could be easily overlooked, and there was no degradation in the sound. At the 256 Kbit per second encode rate we saw obvious degradation in the form of pixelation, but were still impressed with the clarity of the full-screen picture. We did notice an occasional skip in the audio portion of the streams.
 
Overall, Broadcast International’s CodeSys creates a compelling platform for increasing the quality and reducing the bandwidth of video streaming across the network. This technology will particularly appeal to those organizations with streaming video needs and bandwidth constraints between facilities.
 

Conclusions

After reviewing the products and the data presented in this paper, there are several conclusions that can be drawn:
 
  • The CodecSys technology is codec agnostic, it does use multiple codec's in the compression of a single video file or video stream.
  • The process is accomplished on-the-fly and produces a substantially reduced video file size.
  • The technology is “Evergreen” in the sense that as new codecs are developed, they can be easily integrated into the library system. Client devices are kept current with updates embedded in the streaming of newly produced content.
  • The CodecSys approach is “bandwidth sensitive”; it stays within or below the pre-determined bandwidth.
  • The CodecSys enhanced products tested (VOD and Video Conferencing) are full screen, full motion at 24 - 30 fps. rates with high quality viewing.
  • Although CodecSys represents a major paradigm shift in the way video compression is accomplished, it is an inclusive technology that incorporates the “best of breed” compression algorithms of multiple Codec's.
  • CodecSys offers a higher quality presentation at the same bandwidth or same quality presentation at lower bandwidth level options
  • CodecSys technology demonstrates the ability to enhance numerous product and service market sectors.

The CodecSys enhanced products tested provide both real-time and on-demand IP network output.
 

Configurations / Settings

The CodecSys system was tested at KeyLabs facility in Lindon, Utah. The network test bed consisted of a video conferencing PC connected to our TestLink bandwidth shaper which we used to throttle back the available bandwidth. We then connected to a local server or to Broadcast International’s test server, both of which served up the video streams.
 
The specific configurations are as follows:
 
Video Conferencing PC:
  • Single Intel P4 2.6 GHz
  • 40 GB Hard Drive
  • 512 MB RAM
  • Intel Corp 82845G/GL 32 MB Video Card
  • Intel Pro 10/100 NIC
  • Intel AC97 Audio Card

Monitors:
  • Dell M780
  • 1280x1024 @ 75Hz maximum resolution
  • 17-inch CRT with 16-inch viewable image
The video conferencing PC was connected to the following big-screen TV for the video conferencing portion of the test:
  • Mitsubishi 37” Flat Screen TV
Speakers:
  • Standard PC speakers
Additional Hardware:
  • Two (2) Cisco 2500 routers acting as a T1 WAN simulator
  • TestLink Corporation Data Link simulator (used for restricting network bandwidth)

Operating System
  • Red Hat 8.0

Playback Software
  • CodecSys Player

 

 
Screen size and frames per second were all measured from within the CodecSys player.