DVx-HD uses "striping", not cubing or tiling. I don't think that Striping Microsystems, or Slicing Microsystems would be a good name:-)..................................
www2.digitalbroadcasting.com{8FCA30BF-62DC-11D3-9A60-00A0C9C83AFB}&Bucket=HomeFeaturedArticles
C-Cube Announces New HD Codec Targeting Broadcast and Production
9/7/99
By Tom Butts
C-Cube Microsystems (Milpitas, CA) has unveiled its next generation HD codec, which it says is the industry?s first and only high-definition platform architecture for broadcast and video production.
The DVx-HD platform uses C-Cube?s proprietary HDScan process and is compatible with all major HD compression formats as well as international broadcast standards, including ATSC and DVB. The company will be demonstrating the architecture at IBC in Amsterdam this week.
?C-Cube?s announcement marks another significant leap in technology for film and television production,? says Rose O?Donnell, chief technology officer, Avid Technology. ?The DVx-HD platform offers exceptionally high-quality, compressed images at a fraction of uncompressed bandwidth and storage costs. At the same time, it addresses some of the critical market issues needed to establish an infrastructure for the creation and transmission of HD content.?
HDScan?s Striping Approach
What distinguishes the DVx-HD architecture from other platforms on the market is its use of ?striping? to provide communication among processors. Other HD processors on the market use the tiling approach, separating the picture into blocks, which Bob Saffari, director of marketing for broadcast and professional products for C-Cube, compares to multi-monitor videowalls.
?There is no communication between the processors,? Saffari says. ?You end up with a system that has difficulty with bit allocation across the image in a uniform fashion.?
This ?tiling and stitching? approach can result in poor compression and artifacts, according to Saffari.
HDScan divides the image into horizontal slices?up to nine slices for the 1080I HD format. The data between adjacent slices such as motion vectors and information on the complexity of the image is shared among the slices, allowing more efficient allocation of bits.
?The end result,? Saffari says, ?is uniform quality at the highest efficiency.?
Saffari adds that the DVx-HD architecture also is optimum for the video production environment for its ability to encode and decode for editing.
|
