Over the last two decades there has been a lot of buzz about high-definition television, but so far, it has only slowly taken hold. The FCC mandated that U.S. TV stations are to broadcast strictly digital signals by 2006, which is often – incorrectly – interpreted to mean that they have to broadcast HDTV signals. In fact, a station’s DTV signal may include from one to four signals that can be any of a range of resolutions, from standard (current NTSC) to high-definition.
Standard definition TV (NTSC) is generally defined as an image that is 720×480 pixels in a 4×3 aspect ratio playing at 30 frames per second. It is commonly called 480i. All screen images, regardless of resolution, are based on the equivalent of 72dpi. Print resolution is considered to be 300dpi, so a standard definition video image converts to approximately 2.4 inches by 1.8 inches based on a 300dpi measurement. Of course “dpi” (dots per inch), a printer-related term, is meaningless for video and computer images, but it does provide a common yardstick. Each NTSC frame is made up of two fields of alternating scan lines. This is called interlaced scan. Each field has half the vertical information and covers 1/60th of a second in duration. One field contains all the odd lines, the other all the even lines. Despite the lower resolution, standard definition video does a pretty good job of displaying motion, like sports, since your eyes are getting a new image each 60th of a second.
The most common high-definition signal to date carries a resolution of 1920×1080 pixels in a 16×9 aspect ratio. A variety of frame rates from 24fps to 30fps may be used. Using our dpi yardstick, this would equate to a printed image of about 6.4 inches by 3.6 inches – about six times larger in size than standard definition. High-definition images can use interlaced scanning techniques, but what has generated a lot of spark is the use of progressive scanning. Progressive scanning is the way in which your computer screen displays images. Scan lines are displayed sequentially top to bottom for each frame. This means that the entire frame is made up of information from the same instance in time. Essentially this is the same method used in a film camera when exposing images onto a negative.
The 1920×1080 equipment (generally called 1080) is typified by Sony HDCAM gear, some of which can switch between progressive and interlaced modes. To make this work, Sony uses a progressive scan technique referred to as segmented-frame recording. Each frame is still made up of two fields, so in reality 24 frames per second is actually 48 fields per second. Since both fields are captured by the camera’s imaging system at the same point in time, there is no temporal displacement of motion between field one and two. Both fields, in fact, form a complete progressive frame.
The other primary hi-def format is the one typified by Panasonic’s Varicam. This uses an image size of 1280×720 pixels (also a ratio of 16×9). Although Panasonic also manufactures and markets 1920×1080 (interlaced) gear, equipment in the 1280×720 camp usually records at 60 progressive frames per second. It is generally referred to as the 720p format. Some versions, like the prosumer JVC HD camcorder record 720p at 30fps in order to keep the cost low. On the surface, 720p seems like it would have less resolution than the 1080 format, but since the fastest frame rate for 1080 is 60 fields per second, versus 720p’s 60 frames per second, the actual image information that passes in front of your eyes each second is about the same. In addition, the Varicam offers off-speed “stunt” modes that mimics film. The camcorder’s VTR always records to tape at 60fps, but the camera can capture various rates between 3fps and 60fps. This permits the camera to be used in a film-style 24fps mode, but also for undercranked and overcranked shots previously only possible with film.
The “holy grail” would be 1080 gear capable of 60fps progressive recording, but that is beyond our current technological capabilities. Since our eyes and brains are comfortable with dramatic film images shot at 24fps, the industry has grown to accept 1080/24p as a very attractive production approach. This is good for drama, but lousy for motion, and so broadcasters have not chosen to adopt transmission in this format, opting instead to use either 1080i (interlaced) or 720P (progressive) hardware. The nice thing about 24p is that we have many simple and easy techniques to get from 24p hi-def images to NTSC, PAL, 1080i HD, 720p HD or other options. As such, 1080p/24 has come to be known as the Common Image Format.
Modern life isn’t simple enough to only have these few choices, of course. There are a number of quasi-HD formats under the heading of “enhanced definition”. These include 720×480 at 60fps (progressive) as well as a few more. An interesting thing about 24fps recording in HD is that it has broken standard definition’s frame-rate barrier with the equipment makers. We now have several different 24p options with standard definition equipment, such as the Panasonic AG-DVX100, 24p mini-DV camcorder. This camera uses a similar recording approach to its HD cousin, the Varicam. The camera images video at 24 progressive frames per second, but then records that to tape using a pulldown technique common to telecine equipment. Pulldown is a way to divide 24 images into 60 fields. This permits any standard DV, DVCPRO or DVCAM deck to play any of the AG-DVX100’s tapes – yielding a video image not unlike 16mm film transferred to videotape (when the 24p mode was used).
Standard definition video is still largely the way of the world, but many people are shooting HD now, even when the current need is a standard definition master. Why? There are three main benefits in doing this. The first is future-protection. If you need HD material in the future, you have already acquired current imagery in HD and it will maintain its shelf life. Second is alternate media. HD gives you a better starting point, if you want to pull print-quality images from your footage. Finally the image quality of the HD cameras themselves is reason enough. Even when downconverted to standard definition, HD-originated video tends to maintain a better image quality than SD-originated video.
High-def offers advantages for now and into the future and no single format will win out. Modern TV receivers and display systems will be able to intelligently adjust to the incoming signals, so complete standardization will no longer be required. This makes production life somewhat more complicated but also keeps our options open.
© 2003 Oliver Peters