Originally Posted by Steve Bruzonsky
Audio video purists tend to think that the shorter the audio or video cable, or speaker cable, the better! Say it isn't always so though@
Some months ago I saw a post by Amir that only use coaxial digital cables (or for that matter component video cables if you are into analog video) of at least 6', because shorter cables are problematic for reflections impacting the quality of the signal.
Recently, I saw a post by Patrick Harkins of Lumagen, at the Video Processor forum here at AVS, dealing with the Radiance XD/XE, that you should use HDMI cables of at least 6', because shorter cables are problematic for reflections which cause HDMI issues. And that Lumagen users have found their problems often solves moving to at least 6' HDMI cables.
Here we spend big bucks for HDMI SSPs and 2D/3D projectors like Sim2,
and we can be frustrated with HDMI issues.
Comments, folks, both objective and subjective.
The suggestions by Amir and Patrick while quite accurate, are because some manufactures do not handle high frequency circuit design correctly. If the driver and receiver are properly designed, the cable impedance is properly terminated and there are no reflection problems at short lengths.
There are some exceptions outside of the consumer AV field - timing. In the days of coaxial Ethernet, before CAT5 took over, you had a minimum length of cable between two computers. This was due to the time it took for a transmission to go down the cable.Too short and it would cause issues. Other now obsolete coax cable computer networks such as ARCNET also had similar issues. Note that these topologies use the same physical wire for both transmit and receive. So you can't have two trains, one northbound, and one southbound, on the track at the same time. Timing is the only way to make this work. CAT5 uses separate pairs for transmit and receive so this is not an issue.
And of course wire/cable lengths must be matched in certain applications. In old analog NTSC television systems, you had to make sure cables were "timed" so that all sources were matched in time to a switcher. Digital video did not change this requirement. What happened is the by the time we had digital video equipment, buffer memory was cheap and easily implemented so cable timing differences can be fixed with delay memory. Hence cable length matching in digital video systems is a thing of the past, but the underlying physics did not change, only that a high tech workaround was developed. Radar is another field where precise cable length matching is important. Also look at a modern computer motherboard. See how some of the traces are "drawn" like a sinewave. That's delay to make sure all sister paths in that buss are matched. At the frequencies of today's computers, this makes a critical difference. You didn't see this on an old IBM 4mhz PC because it was not critical at that frequency.
But again this is not significant with audio systems nor is it a problem in a simple video switcher. It only matters where you have to mix two video sources. So physics hasn't changed. The shorter wire length is always the better option. One just has to take into account transmission line theory above certain frequencies. At audio frequencies,these issues are insignificant just like skin effect.