Digital (especially HDMI) cable quality

RichTeer -- Thu, 07/17/2008 - 12:20

As a long-term audiophile I know too well that cables (interconnects and speaker) are an important part of an audio system. But I am at a loss to understand why digital cables--and I'm thinking especially about HDMI cables here--can sound different to each other.

Could someone please explain why digital/HDMI cables make a difference? In another forum I frequent, the prevailing "wisdom" when it comes to HDMI cables is "bits are bits", and I'd like to be able to refute that.

TIA!

Al Sekela -- Thu, 07/17/2008 - 14:57

All cables in an audio system participate in the system noise environment. Even though the bits that come out are the same as the bits that go in (for a non-defective digital cable), the electrical connection also conveys noise between the connected equipments. The noise level is usually too low to invert the bit identity, but the noise gets into the equipment through parasitic pathways, where it can affect the bit timing and may affect the audio signal.

Cables can also modify the noise, by acting as antennas and resonators. The best analogy here is an organ pipe: it generates a loud tone with a well-defined fundamental frequency and overtone spectrum from the soft "white noise" at the input generated by turbulent air flow. Similarly, an audio or digital or power cable can be stimulated by conducted or induced noise to resonate at particular radio frequencies. The levels of the resonant tones depend on the details of cable construction.

RF noise is much too high in frequency to be directly audible, but it can generate spurious audio tones by intermodulation with the audio signal. These tones mimic natural overtones and affect our perceptions of instrument and voice tonal balance, richness, sound-stage depth and illumination, and microdynamics.

Robert Harley -- Thu, 07/17/2008 - 18:17

I haven't researched how signals are transmitted in the HDMI interface, but I have some experience with digital signals in the SPDIF interface. It would seem that a simple piece of coaxial cable carrying a digital signal from, for example, a CD transport to a digital-to-analog converter would be transparent. But digital cables exhibit an analog-like variability, even though the bits are identical. The mechanism by which this occurs was discovered by Malcom Hawksford and Chris Dunn in their Audio Engineering Society paper "Is the AES/EBU SPDIF Interface Flawed." A synopsis is beyond the scope of this post, but I describe the phenomenon in detail in TAS Issue 180. If you're interested, you can also buy paper preprints from www.aes.org.

RichTeer -- Fri, 07/18/2008 - 11:34

Thanks for the pointer, Robert. I'll look at my copy of TAS180 to look for that article. BTW, your recent sidebar article on jitter was *very* informative, so many thanks for writing it.

Robert Harley -- Fri, 07/18/2008 - 13:53

Thanks, RichTeer.

Incidentally, I was once measuring varying levels of clock jitter in a DAC (at the DAC's 352.8kHz word clock, the point where jitter matters) with different transports, and when I repeated the measurements just as a check, got different results. I eventually tracked it down to the directionality of the SPDIF cable. The way the RCA jack is soldered to the cable can cause some of the signal to be reflected back to the input, increasing the amount of jitter in the cable (and thus in the DAC). I'd look at the test instrument reading, reverse the cable, and watch the jitter drop in half (in some DACs).

I was looking for how transports affect jitter in the DAC's word clock and stumbled on the cable-directionality phenomenon.

All content, design, and layout are Copyright © 1999 - 2011 NextScreen. All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.