→ April 2006 Contents → Column
Now that small-camera high-definition video is a reality, there's an awful lot of confusion about the relative quality of the myriad of variations. Dirck asked me to attempt to explain the differences between HD formats like Sony's HDCAM or Panasonic's DVCPROHD, and the new HDV. That's a tall order, but I'll try.
There are several factors we need to consider when comparing digital HD formats. They include:
Before we take a look at each let me throw some numbers at you.
Sony's HDCAM format records a 1440x1080 signal using 4.4:1 compression at a bit rate of 144mps and 3:1:1 subsampling.
Panasonic's DVCPROHD format records a 1280x720 signal using 6.7:1 compression at a 100mps bit rate and 4:2:2 subsampling.
HDV records a 1440x1080 signal using MPEG-2 level H-14 compression at a bit rate of 25mps and 4:2:0 subsampling.
Okay, thoroughly confused? Let's try to sort this out.
Resolution: The accepted resolution for any 1080i or p HD image is 1440x1080, or 1.5 million pixels. 720 is a progressive-scan-only format, recording 1280x720, or 900,000 pixels. Some super-high end HD systems will record at 1920x1080.
Compression: This is the amount of data squeezing a format employs to reduce the overall size of the video "files." Back in the days of analog recording there was, of course, no compression. When digital came along the amount of data increased substantially, so manufacturers looked to compression as a means of fitting the signal into the required space. The ratios used by Sony and Panasonic's high-end formats are, today, considered "mild." And if you've ever viewed the direct output of their cameras you'd agree that, despite compression, there appears to be no sacrifice in quality.
Bitrate: This is the amount of bandwidth a format uses to record the data stream. A comparison of the three formats listed above shows that HDCAM and DVCPROHD have several times the bitrate of HDV. The higher the bitrate, the higher the quality. Sony's new HDCAM SR format records data at 440mps, or at an astonishing 880mps in HQ mode.
Chroma Subsampling: This describes the amount of color data recorded. The subsampling rates make use of human physiology -- our eyes are less sensitive to color than they are to brightness. So, engineers figured out that they could reduce the resolution of color data compared to luminance. While the Holy Grail of subsampling is 4:4:4, none of the major formats employ it (except HDCAM SR). The first number refers to the relative resolution of the sampled luminance data ("Y"). The second and third numbers refer to the sampled color data ("U" and "V"). Thus, 4:2:2 means that the color resolution is sampled at half the rate of the luminance. And 4:2:0 means that the color information is alternated between video lines (thus, the ratio in practice is actually 4:2:0/4:0:2/4:2:0 and so on.).
I know that some of the folks who rent our Sony HDV Z1U at Videosmith notice digital artifacts and aren't too happy about it. That demonstrates that a discerning eye will pick up qualitative differences between HDV and HDCAM/DVCPROHD. But there are so many other variables involved in shooting, editing and displaying HD signals, that HDV will hold up very well indeed against its more sophisticated brethren. One place where common folk might see a difference is when HD is transferred to film and projected on a large screen. The limitations of HDV will be more apparent then. But I don't think that's a pressing issue for readers of The Digital Journalist.
When the time comes to choose your HDV camcorder you don't have to worry about all the geeky numbers - the manufacturers in the HDV Group have done that for you. So, just concentrate on the features, and pick the camera that best suits your needs. It's going to do you proud.
© Steven Trent Smith
Back to April 2006 Contents