Avid’s approach to compressed HD
Before Tichenor’s keynote, Chas Smith, general manager of Avid Video, gave a brief presentation on his company’s take on the evolution toward HD. (HD is present in about 12 percent of American homes, he reported.) His points were quite simple: Traditional I-frame-only 1080i video contains about 6X the pixels as NTSC. That means 6X the required storage for 1080i compared to 480i. HDV, which is compressed to allay that challenge (among others), presents problems for postproduction.
The long-GOP structure of HDV means that in the editing processing, a lot of cuts will begin on an incomplete non-I-frame. This requires re-encoding around the edit to create a new legalized GOP around that cut. This essentially introduces a second generation of the media. Also, rendering for output after, say, color correction requires another re-encode. Finally, horizontal sub-sampling (1440 pixels across vs. 1920) that’s common to HDV camcorders requires compression and decompression cycles during monitoring and compositing.
Avid’s approach is to offer an “open timeline” in its NLE. The timeline accepts native media of several types (DV, HDV, DVCPRO, uncompressed) simultaneously. (An alternative solution that many NLEs offer is an intermediate codec used during editing.) When the cut is done, choose to render out to a media type that wouldn’t subsample the raster.
For more on Avid and HD workfow, read “Beta Sight” from the April issue of Millimeter.
Related Topics: NAB 2006







