I was able to find a test done a few years ago by PC World. I'll ignore the HDMI section, they confirmed what we've all already agreed upon, HDMI cables at typical distances don't matter. Unless it's so cheap that the connector falls off (which may happen) you get perfect video. This is what they wrote in their test of component video cables:
"We started by measuring characteristic impedance--the extent to which a cable hinders the flow of a signal. The standard impedance for each wire in a component-video cable is 75 ohms. If the impedance in any one wire is far off the mark, it produces an impedance mismatch with the devices it connects to; as a result, some of the signal may be lost in transmission, or it may bounce back along the cable to the source, producing smeared colors or blurriness in the picture's fine details.
Monster's M500CV was the winner here, as all three wires inside the cable varied within a negligible 1 ohm of 75 ohms. Translation: This cable imposes as little distortion as possible. Other cables didn't do as well. The three wires included in the CableWholesale.com cables hovered between 63 and 64 ohms, while the Kimber Kable's wires measured between 85 and 86 ohms. The AudioQuest's wires varied from about 71 to 75 ohms. And the StarTech.com's varied from about 67 to 69 ohms.
But here's the rub: Virtually every consumer component cable uses RCA-style jacks. Originally used for analog audio connections, RCA plugs have an impedance of about 50 ohms, creating unavoidable impedance mismatches at both ends of a cable. How well a cable manages the impedance at every point of the cable, not just at the connectors, affects its performance. But the impedance mismatch between a cable's wires and its RCA connectors has far more impact on performance than any other attribute.
We next calculated return loss, a measure of how much of the signal bounces back down the cable. According to the experts at Tektronix we consulted, 14 decibels is ideal. CableWholesale.com's product exhibited the least amount of signal bounce in our tests, at 13 decibels. The other four brands did worse (all at about 8 decibels). The Tektronix experts explained to us that practically all signal bounce is a result of the impedance mismatch between the wires and their RCA connectors.
We conducted one final test: Insertion loss, measured in decibels, gauges how much of the video signal gets lost as it runs through the cable. Four of the cables managed roughly equivalent performance. The worst performer in the group, the Kimber Kable V21, lost less than 2 decibels--an insignificant amount.
Working with an AccuPel HDG-3000 HD/SD/DVI Component Video Calibration Generator, we sent 720p test patterns through our cables to the Epson PowerLite 500 projector. None of the cables transmitted a perfect signal, but the imperfections were minor. In crosshatch patterns (a grid of fine horizontal and vertical lines), some lines displayed slightly smeared edges or shadows, rather than sharp pixel-for-pixel transitions from white to black. But we had to get within a foot of the screen to see any of this, and we saw the same problems regardless of which cable we used.
Another set of test screens displayed multiburst patterns, featuring several swaths of parallel vertical lines that get progressively finer from left to right. On every screen, the finest swath--where the lines were just a single pixel wide--looked blurry for each cable, indicating that even a good display might smear small details. The other swaths were sharp, with well-defined transitions, regardless of the cable.
Bottom Line: Though the analog cables varied slightly in our instrument tests, they did not produce distinguishable differences in transmitting real video content."
The conclusions I take from this are not really any different than what I've though previously about audio cables. Different digital coaxial cables will measure differently, just as these CV cables did. You're after 75 Ohms including the connectors, and you often don't get it. Just throwing money at any brand is no guarantee of results. The problem I have with their conclusions is that they judged generally by eyeballing it, and sorry but our eyes are just not very accurate. Now before anybody starts I KNOW how that sounds. This is aiming to be a scientific test though, and I would've preferred to see more instrumented results. There is extremely accurate equipment to measure things like light and color from a display, and unless I'm missing something those measurements weren't done.
