THE FRONT LINE: APRIL 5, 2005
Left Behind: Analog Video Processing Goes ‘Down the Tubes’
Many consumers who buy their first EDTV or HDTV set are shocked to find out just how bad analog video looks on these sets, and are returning them to stores. Why?
For years, the only way we could view moving electronic images in the USA was using the NTSC (analog/interlaced/525-line) video standard. There was no argument, no contention, no disputing. Although video content constantly changed, the ‘wrapper’ it came in was always the same – a luminance signal with color subcarriers and a sync burst, all traveling down a 4.25 MHz pipe (RF modulated or baseband).
As a result, when the first multi-sync large-format monitors came to market, there was a flurry of activity to clean up the intrinsic by-products of this composite video system such as cross-luminance and cross-color artifacts, including loss of fine picture detail (the result of using notch filters) and color moiré (cross-color artifacts).
Some manufacturers took the approach of converting the interlaced signal to a progressive-scan signal to take advantage of the higher horizontal sync rates (31.5 kHz) of these new multi-sync CRT monitors. Of course, that created more problems such as motion and scan line artifacts.
Other manufacturers began playing with ways to preserve image detail by using different filters to extract the color burst information at 3.58 MHz while not removing precious image detail in the 300 to 400-line area. Still others were experimenting with edge enhancement to make pictures appear sharper than they actually were.
With the explosive growth of large-screen displays and front projectors in the 1990s, this activity reached a fever pitch. Companies vied with each other to bring line doublers and even line quadruplers to market. Terms like “3D comb filter” were prominently featured in ads for LCD projectors and early plasma monitors.
While the first line doublers offered by Faroudja Laboratories retailed for close to $20,000 in the early 1990s, prices had absolutely plummeted by the end of the decade when DVDO brought out its revolutionary iScan line doubler with motion compensation for an astounding $900.
The reason, of course, was the inevitable increase in density and power of application-specific integrated circuits (ASICs). Today, that same processing power and then some can be found in many garden-variety DVD players for under $100. The term “line doubler” has essentially disappeared from the industry lexicon, replaced by “progressive scan” and “480p”. (We don’t even refer to the full 525 scan lines anymore!)
With the evolution of HDTV and image processors that can create and show video with all kinds of pixel resolutions and aspect ratios, there is simply no ‘buzz’ anymore in composite video.
Based on tests I’ve run in the past year and product reviews I have conducted, it’s apparent that manufacturers of video scalers (that’s what line doublers and quadruplers have become) don’t even care about composite video. Nor do many projector, plasma, or LCD monitor manufacturers.
Surprisingly, several video scalers I have tested recently, including DVDO’s iScan HD, have mediocre composite video decoding and image processing. How ironic, given that just five years ago, it was DVDO who turned the interfacing world on its head!
The test for composite video decoding is easy to perform. Simply obtain a copy of Digital Video Essentials on DVD (available on line for about $25) and load up the Snell and Wilcox Zone Plate test pattern. Connect the composite output (yep, DVD players still have ‘em) from your DVD player to the composite input of your display, and let ‘er rip.
Chances are; you will see something that looks very much like Figure 1, with missing detail at 300 and 400 lines and color moiré all over the place. In the event that your projector or monitor does have a top-notch composite processor, you may get lucky and see something more like Figure 2, although the moving target won’t be as free of color artifacts.
Figure 1
Figure 2
So why are manufacturers giving up on composite video? Well, let’s see how many signal sources still deliver it. HDTV set-top receivers? Nope, unless they’re set to 480i output, and even that signal is often available as S-video or component YPbPr.
DVD players? Nah, most of those have gone progressive scan. Cable and satellite set-tops? Unless they are processing analog video, the digital signals they demodulate all come through as component video (unless they started as analog video).
I have tested several front projectors this year that are specifically intended for the home theater market, but have inferior composite video decoding and motion compensation performance when compared to my Faroudja-equipped Panasonic RP56 DVD player. (That player cost about $225 when I bought it a few years back, but equivalent models are less than half that price now.)
The real reason could simply be that so many video players and receivers now include some sort of composite video processing, and to have it in a projector, monitor, or TV adds an unnecessary expense if most people aren’t going to use it.
Many electronic displays or signal sources also include Firewire ports so you can plug in the family MiniDV camcorder or still camera. Neither of those are analog composite formats.
Certainly a person putting together a nice home theater to watch DVD movies and HDTV programs isn’t going to care much about composite video decoding. Why else would DVDO and other video scaler companies step down from higher levels of composite decoding found in older products?
Only one company – Silicon Optix – has announced any kind of advanced processor for analog video. Their new Realta HQV series of chips do an amazing job when cleaning up analog signal artifacts, including scan lines, noise, and inter-field motion errors. But very few TVs and DVD players use this new processor so far.
I recently tested a new TiVo player with built-in DVD recorder. Even though this product only time-shifts analog TV signals, it also has a progressive-scan mode for both the TiVo PVR and the DVD player, complete with component video outputs.
That means I can now replace both my old (vintage 1998) Philips TiVo receiver and Sony DVD player with one box that also records DVDs. And the composite video signal processing on this combo is pretty good – not great, but certainly better than that found in some of the projectors and monitors I have tested lately.
Even the time-honored S-video connection isn’t getting as much of a workout as it once did. It was a big deal back in the early days of DVD players to have that S-video connection, but once consumers glommed onto the value of a “component” video output (which S-video technically is) and then embraced progressive scan, S-video connections became an afterthought.
True, most consumer TVs sold these days have a full rack of composite and S-video inputs (although most of them are never used). But professional-grade displays are cutting back on the number of composite and S-video input jacks as more and more customers demand analog component and digital interfaces.
A typical home theater projector might have one composite input (again, with minimal processing), one S-video input, and three component inputs – one YPbPr, one RGB, and one digital HDMI or DVI. And the DVI input is often configured as DVI-I, which means with an adapter cable it, too, can accept RGB and YPbPr inputs. A typical plasma or LCD monitor will offer pretty much the same line-up.
Three things continue to drive the analog composite marketplace – off-air analog TV, which will go by the wayside in the next ten years; analog cable TV, which may start to disappear in as little as two years, and VHS tapes, sales of which are already plummeting as DVD formats increase their market share.
With issues of digital rights management and ‘smart’ displays coming to the fore, the entire world of analog video signaling is undergoing a remarkable changeover to digital. That means there won’t be any room left for a 60+-year-old analog video system; one that was once designed to be an efficient, bandwidth-conserving scheme for transporting high-resolution interlaced monochrome video images.
How the mighty have fallen………………