OTHER TECH

The considerable splashing in the Olympics racing pool also provided a tough test for the other main type of encoding artifact - "blocking," in which the image breaks up into areas of mosaic-like squares. Blocking is easy to see in pool and fountain footage because its very squarishness is extremely unwaterlike. And even though, from an encoder standpoint, the movement of splashing water is pretty much as random as the static from an empty TV channel, we know how the water is supposed to look.

The NBC high-def feed seemed to have some residual blocking in the swimming coverage. And rapid movements by gymnasts and divers often displayed both mosquito noise and blocking simultaneously. But it's not clear whether I noticed these artifacts because of poor encoding by NBC, poor reception with the indoor antenna in our offices (buried deep within a skyscraper), poor decoding by the HDTV recorder, or the peculiarities of the LCD monitor I used for most of my viewing - or simply because the characteristics of the segments made them easily visible. Quite possibly all of these factors contributed.

Gymnastics and swimming contests at the 2004 Olympics turned out to be tough tests for the HDTV system (screen grabs from NBC's HDTV coverage).

Some Sound & Vision editors re ceiving the NBC high-def feed via cable reported seeing pristine images, and the visibility of both mosquito noise and blocking decreased markedly when I watched the high-def recordings I'd made on a plasma TV. (There should be a Greek chorus calling for research on whether and how different types of screen interact with encoding artifacts, making them more or less visible.)

I also noticed some very peculiar effects, when reception was very poor, stemming from the way MPEG encoding operates - by transmitting "reference" frames and generating the frames in between by interpolation. Once a gymnast standing under the high bar dematerialized, à la Star Trek, and then rematerialized up on the apparatus! I don't think any country would protest if

I gave that a score of 11.