What is HDV?

Kyle Cassidy
December 2004

In the first television revolution since the 1950s, 2006 will mark the year when America goes digital. In that year, by government mandate, digital television (DTV) and HDTV (high definition) will be a reality. Of course, many Americans will upgrade to widescreen high definition TVs over the next few years as the prices come down.

Once HD is firmly implanted in America's living rooms, there will be a demand for content. Once the djinn is out of the bottle, there's no going back: Television just won't look the same. In fact, there is a demand for HD today, from major networks and even local PBS affiliates. However, professional HD production equipment is priced at many thousands of dollars, far out of reach of most American videographers.

In September of 2003, Canon, Sharp, Sony and JVC announced they would work together on a home High Definition Video (HDV) standard. HDV allows high definition video to be stored on an ordinary DV or Mini DV tape. JVC has already come out with two cameras, Sony's announced a third and various editing solutions are already on the market, all within the price range of prosumer videographers.

HDV stores 16:9 high definition video on DV tapes, supporting both 720p and 1080i (NTSC and PAL), using MPEG-2 compression for data rates of 19Mbps and 25Mbps respectively, comparable to the bitrates of ordinary DV (25Mbps). Audio is compressed by 75% using MPEG-1 Layer 3.

Is this too good to be true? Well, sort of: There is a catch. The video streams over FireWire using an MPEG-2-TS (or "Transport Stream") format. This is a compressed format, which is incompatible with most DV editing software. MPEG compresses clusters of video frames (GOP or group of pictures) together into one stream for delivery. By taking advantage of the similarities between frames (after all, if the sky is blue in frame 1, what color is it likely to be in frame 2?) and compressing them together, MPEG is able to provide significant space saving. This also means that frame-accurate editing is difficult, to say the least. The computer must decompress each group of frames and then the whole thing re-compressed, even for simple timeline previews. MPEG, remember, is also a loss-y compression algorithm, so you can't just keep compressing and decompressing it. Multiple rendering and re-rendering of the video will produce compression legacies that get worse and worse. MPEG also bundles (i.e. muxes) the audio into the compression stream. All this means is it's difficult to edit with traditional editing tools.

The resolution of a TV image directly affects how sharp it looks. Higher resolutions provide more detail. An SD (standard definition) NTSC television frame (such as the one you watch every night) is made up of 525 horizontal lines broken up into two fields. The first field is made up of the odd lines and the second field is made up of the even lines. This is what we call interlaced video.

Of these 525 lines, only about 480 actually make it onto a normal television screen (the other lines are used for synchronization, closed captioning information and some are just plain masked). The horizontal resolution is how many dots each of those 525 lines can have going across. The resolution of a TV image is determined by the "circle rule" or the number of dots across the largest circle which can be displayed on the television. On a standard 4:3 TV, a circle will fill about 75% of the display. On a 16:9 formatted HDTV, the circle will fill about half the screen. According to the circle rule calculation, NTSC video has a resolution of about 330. A VHS tape has about 240 lines of resolution. DVDs have about 480 lines of resolution, which is the highest resolution a standard TV can display. High Definition televisions on the other hand, can display 1080 lines of vertical resolution, twice that of an ordinary TV and three times as much as broadcast television.

• Digg This!
• del.icio.us
• Technorati
• StumbleUpon
• Reddit