Page 115 - Hi-Fi_News_-_December_2020 video

JITTERBUG

USB Filter
YOUR VIEWS

Digital cable matters

QUESTIONS OVER AUDIBLE DIFFERENCES BETWEEN USB INTERCONNECTS

I’m not one to adhere to the
‘bits are bits’ dogma when it
comes to digital audio sound
quality because I can imagine
that some changes can impact Can a £39 insect make all

a signal when it is involved in your CD files sound better than
some form of conversion, such Hi-Res?
as digital-to-analogue. But I am Yes and no: Using the same
less convinced in the case of equipment and a quality DAC, a 24/96
purely digital-to-digital, as is the fi le (for example) will always sound
better than a CD 16/44.1 fi le … but,
case with USB cables. even a single JitterBug will often
While reviewing Chord’s allow a CD file to be more musical and

Epic USB cable [HFN Sep ABOVE: USB data pattern measured via a 90ohm more emotionally stimulating than

’20] editor Paul Miller describes host USB sink and comparing 1m (black) with 5m a Hi-Res file without the benefit of a
JitterBug.
audible differences when using (red) of Chord Epic USB cable
the cable in his system. Yet Noise is the problem. Real noise—
the kind you can’t hear directly. Most
the purpose of the USB cable is significant degrading of the signal, often, the word “noise” is used to

simply to convey an electric signal which does not seem evident in the describe tape hiss or a scratch on a
that pulses between 0 and +5V USB data pattern that accompanied record, but these sounds aren’t noise;
representing the serial output from editor PM’s review. Surely there must they are properly reproduced sounds
that we wish weren’t there.
a source component’s parallel-to- be some other contributing factor Problem noise is essentially random,
serial conversion. This signal is responsible for the change in sound resonant or parasitic energy, which
then presented to the destination quality that PM describes? has no meaning. It can’t be turned
component which in turn takes the Paul Radford, via email into discrete sounds, but it does
signal and, via serial-to-parallel compromise signal integrity and the
conversion, reconstitutes the original Paul Miller replies: No argument from me, performance of everything it touches.
digital word in a shift register that’s Paul, but you offer a clue to the answer JitterBug’s dual-function line-
conditioning circuitry greatly reduces
then sent for further processing. in your opening paragraph – the key to the noise and ringing that plague both

I find it hard to understand how the understanding how digital interconnects the data and power lines of USB ports,
USB cable could degrade the signal (whether carrying USB, S/PDIF, AES, whether on a computer, streamer,
so much that it causes errors to be I S, etc) may influence sound quality lies home stereo or car audio front-panel
2

USB input.
introduced into this chain of events. in the analogue, not the digital domain. A single JitterBug is used in between
The serial-to-parallel circuitry uses Indeed, our group tests [HFN Jul ’13/ Jul devices (i.e., in series) as shown
either the rising or falling edge of ’14] were the first to demonstrate that USB below. For an additional “wow”

the signal voltage combined with a – a synchronised, packet-based digital experience, try a second JitterBug
clocking mechanism to determine interconnect standard – was also cable- into another USB port on the same
device (such as a computer). Whether
which value to load into the shift sensitive. Clear differences in the ‘shape’ the second port is vacant, or is
register. For this to fail in some way, of the data waveform arriving at the USB feeding a printer or charging a phone,
and thus lead ultimately to changes ‘sink’ were illustrated, just as the Graph JitterBug’s noise-reduction ability is
in the sound quality, would require [reprinted above] accompanying my likely to surprise you. No, the printer
won’t be affected—only the audio!

review of Chord’s Epic USB also showed
While a JitterBug helps MP3s sound a
differences with the length of the cable. lot more like music, high-sample-rate
The shape of the waveform files have the most noise vulnerability.

is indicative of response and Try a JitterBug or two on all your
characteristic impedance equipment, but never more than two
per USB bus. There is such a thing as
which, in turn, tells us something
about the bandwidth of circulating too much of a good thing.
interference that might occur. This
is just one factor among many
(unknowns) that have no impact
on the description of the data in the
digital domain but, along with differences
in jitter, do influence the types of noise

and distortion that arise during D/A
conversion. Computers and the DSP in
every outboard USB DAC operate in the
ABOVE Chord’s Epic USB cable sports digital realm, but we audiophiles process
bespoke silver-plated A- and B-type plugs our ‘data’ in the analogue world...

DECEMBER 2020 | www.hifi news.co.uk | 115