Why do digital cables sound different?


I have been talking to a few e-mail buddies and have a question that isn't being satisfactorily answered this far. So...I'm asking the experts on the forum to pitch in. This has probably been asked before but I can't find any references for it. Can someone explain why one DIGITAL cable (coaxial, BNC, etc.) can sound different than another? There are also similar claims for Toslink. In my mind, we're just trying to move bits from one place to another. Doesn't the digital stream get reconstituted and re-clocked on the receiving end anyway? Please enlighten me and maybe send along some URLs for my edification. Thanks, Dan
danielho
09-05-08: Rja
Palerider,
Why do you assume no one here has used AT&T digital cables?
And why do you consider it better than other digital interconnects?
You keep making statements with no supporting information."

- as I sayd..
EASY....
digital cables sound diferent becouse are diferent, diferent materials, diferent quality, diferent lenghs, diferent manufacturing process, etc..
without the technical stuff, i will be looked like a dumb.
but thats the real answare, plain and simple.

to understand deeply why,
must understand the diferences between analog audio, and digital audio, and the diferences between good and bad cable.
in the end your ears will be the judge.

for example:
usually analog sound has 24khz bandwith, and two cables, one for L & R. "red & white RCA cables"
line level has arround 1 to 5v p-p aprox. and arround 50ohms to 600ohms for the emmiter and 10k to 20k for the receiving unit. those are electrical parameters.
some vintage analog gear its 600ohms all the way, and some rack gear has switches, like aphex dominator 2 720. for 600ohm or 10/20kohms signals.
and sound verry similar.

digital = s/pdif, sony philips digital interface.
has 1 cable for 2 audio signals if used in PCM mode, and if used for RAW/AC3 mode has 6 channels.

digital has some problems like jitter, wordclock accuracy, and cable performance.
becouse digital audio travels in serial, not parallel.
all affects the digital signal, digital has a travel bandwith of 3MHZ up to 25MHZ, but an audio bandwith of 44.1k for both L&R = 22050hz, the other bandwith its used for the 16/18/20/24-Bits or each sample in 1 second.
depending on the resolution used, 44.1-16bits up to 192k-24bits.
digital sound needs aprox. 3million of bits, "1s and 0s" per second for a 44.1khz - 16 bit signal.
spdif was original designed maxium for 20bits audio and other 4bits for unknown/user data.,

the 24khz/24.000Hz analog audio signal must be transformed into serial information represented by gruoups of bits & Bytes,
and for all those 1s and 0s to sound at the same time!!, must travel pretty fast.

digital 1 and 0s, travel at a verry tiny rate.
controlled by a wordclock, or masterclock.
every digital device has a wordclock, from ipod to the soundcard of my pc,
usually its a quats crystal vibrating at a fixed rate.

the wordclock signal is a tiny square waveform with a frequency of minium 3MHZ. = 3000.000Hz digital audio for 44.1-16bits stereo signal vs. 24.000hz of analog audio.
with near "same" information.

also wordclocks have problems, most are not perfect,
have side effects like wrong pitch and jitter.
bad designed wordclocks have lots of jitter.
from bad power supply or crystals contaminated with micro dust at the manufacturing process, or low quality/purity crystals selected, bad or cutting, to fluctuations of the frequency created by unstable internal operating temperatures, that expand and contract the crystal, affecting its vibration.

jitter does not affect digital sound when its in digital format, BUT when its tranformed/converted from analog sound to digital or from digital to analog, affects totally!!! the harmonics, creating harmonic cancelations. or the blurr efect, over the entire frequency range. depends on the jitter and the .wav .mp3 file being listened. or the instrument recorded.
also affects stereo separation becouse both channels are in serial. LRLRLRLRRLRLRLRLRRLRLRLR, jitter blurrs the line of that separation between L&R.
but some people like digital jitter, sounds "more analog."

analog sound its a representation of the air molecules in electrical alternate current impulses/vibrations form.
Analog sound/line level sound emulates exactly as posible the frequency/movement/vibration of the air molecules.

the tiny square waveforms of the wordclock need something called slew rate that is the rise time from 0 to max 0.5v.
at that level 0.5v 75ohms. rise time. bad slew rate = not a square waveform, more like a 50% sine wave form 50% square.

s/pdif digital cables have 2 important values,
velocity of propagation, that should be ideal equally to the speed of light, but in the best cables its arround/less than 85%, and worse cables near 50%,
also cables have something called capacitance.

capacitors store energy and then release it, like cut & paste, a capacitor its made by isolating two electrical currents, ironically thats exactly what a cables is.

in extreme capacitance, cuts more, making holes from the tiny bits&Bytes traveling at 3Mhz. and then paste on top of other bits, creating dissorder, or a blurr effect , creating digital errors, micro data lost. clock sync lost.

also capacitors are used in audio crossovers, usually passive crossovers found in most 2/3 way pasive speakers,
if capacitors are used in serial or parallel to the audio signal becomes a hipass or a low-pass filter at near -3 to -6dB per octave. = 1-pole, also same happens with resitors and inductors.

the capacitance affects the point "cutoff" of the low-pass filter.

if the capacitance its high, has a low frequency cutoff point, for example: lets say 10kHz. ive heard audio XLR cables that have less highs than others. that why i choosed 10Khz.

that means that at a frequency of 20kHz the same cable has -6dB, and at 40kHz has -12dB, and at 80khz has -18dB, at 160kHz has -24dB, 320k-30, 640k-36, 1280k-42dB, 2560khz will have -48dB
and the s/pdif digital signal travels near 3Mhz = 3000Khz.
up to 25MHz.
and digital s/pdif has a power of 0.5volts
capacitance makes a weak signal, weaker.!
measured per x meters. longer distances are worse = more capacitive and resistive.

thats why there are baluns from ets and canare and neutrik to transform AES/EBU to s/pdif and back.
aes signal its stronger than s/pdif 5v vs. 0.5v but impedance of 110ohms vs. 75ohms. and twisted pairs vs. coax cable makes things diferent.

...components:
silver its the most electric conductive material, in pure form. unless its mixed with some other chemicals like artic silver 5 cpu thermal grease, it becomes capacitive, non conductive.

2nd best its 99.9999% OFC Copper, 3rd best its 24k Gold, gold does not oxidate like silver or copper, thats why its used in connectors.
oxidation makes harder for electrons to pass through.

for example, dean connectors for R/C cars, have a blend between silver, copper & gold.
and blow away any other tamiya aluminium connectors, becouse generates less heat, becouse its more conductive, less resistive.

a dirty contaminated silver cable, its worse than a pure high quality copper cable for example.
OFC oxigen free copper, was invented as a measure for the quality/purity of copper.
some cables say OFC, but only true OFC will make s/pdif sound better/transparent/clean/detailed.

a bad quality copper cable, has oxigen micro holes that makes the cable last less years and making electrons harder to travel, also bad copper has diferent micro densities, not uniform, making like a open road, harder to travel for electrons, making harder for electrons to move unobstructed.
bad manufactured copper cables are not enought to be error free digital signal.
bad copper cable can have a Velocity of propagation of 50%
verry resistive!!

also insulation between the center and the shield of the coax s/pdif cable.

also the material of the shield of the coax s/pdif cable.
most manufacturers use aluminium shield becouse its cheap, but its not 100% Radio Frequency shielded.

almost anything generates external RF interference.
from the sun, your tv, to the electric wires in your house inside the walls, the invisible AM/FM radio waves, cell phone waves, microwave ovens, fluorescent bulbs, etc...
true RF elimination its from a copper mesh.
that was discovered by someone hundred years ago.
cant remember its name.
anyway,
the insulation material affects also.
a bad insulation material between center and shield can make incredible BIG distortions in the digital signal.

some manufacturers use esoteric meterials, and hi tech fabrication process.in the search for the best insulation, low loss cable.
= cables are far from perfect.
like:
abbey road and empirical audio, that has pure silver cables, $800usd. meter. holosound2bp, or something. with teflon or another hi-tech esoteric insulation material.
cant remember if the shield its silver or copper.

zaolla, and others has mixed silver & copper center.
vovox digital has pure copper ofc, silver coated/plated. center.

etc..etc..etc.. thers lots of esoteric brands out there.
if you search yahoo/google.
but few make velocity propagation and capacitance tests above 1Mhz.

also stranded vs. non stranded, vs. diferent kind of stranded forms to avoid the skin effect.
that make the high frequencies travel at the outside of the cable, or so they say.
most tv & radio stations use Belden 1694a or gepco.
anyway.
i think apogee wyde-eye its gepco with canare connectors.
abbey road studios use custom pure silver cables. etc..
lots other brands...
also RCA connectors make a big diference.
plastic vs. metallic, aluminium with gold plated, vs. copper vs. gold platted, vs. pure copper OFC.
etc..
also the connectors must be OFC., thers no point in OFC cable without OFC connectors.
seems that size of the shield of the interconnects also affects.

the more V.P. the cable has, and less capacitance, the electrons organized in tiny square wave forms with a rate/vibration of 3mhz, those bits can travel easier = more accurate.
if you have a <1ppm wordclock like Drawmer M-Clock, apogee big ben, dcs995,verona, mutec iclock, etc.. or more accurate wordclock like 0.03ppb Antelope Atomic 10M clock, digital will sound more transparent, detailed and real, blurr will disapear completly,
but the highs will sound "strange" 10101010 at near 22050hz, aka.44.1khz.
if you record low jitter and reproduce with jitter will sound better. than recording with jitter and reproducing with jitter, becouse its twice jitter! and its not 100% exactly the same.

some cables adds a kind of jitter that makes the digital sound more "analog" in the highs, but blurr in the mids, and weaker in the lows.

BUT the realism you get in the mid frequencies, and the added loudness, and stronger bass with a jitter-less system, makes you forget the ear piercing 1010101, in the verry highs.

the "solution" for bad s/pdif cables was introduced i think by toshiba, the Toslink, = optical s/pdif.
specs where added later for raw signal 6 channels AC3 and for 8 channel ADAT 48-24bits & MADI optical formats. and later updated again for smux 96k and smux2 192k, BUT... optical has problems too.

has no capacitance but has added Jitter. in some cases multiplied by x8 or x64 times more jitter.
electric to optic to electric adds jitter to the signal.
external wordclock must be used.
becouse wordclock is multiplied.by the optical interface and divided again at the reciving unit, creating jitter.
also fiber optis have something similar to velocity of propagation, that measures the energy loss of the 640nm laser beam light.
so an external wordclock also must be used to avoid optical added jitter.
= again back to cables...
BNC Wordclock cable its "the same as s/pdif cable"
75ohm,

s/pdif can be used for BNC WC cable with optional copper center BNC male to RCA female adapters, easy.

anyway...
i have found a "secret" cable that blows away any other s/pdif or bnc cable!!! i have tested so far.
it does not add jitter!!. and its not silver.

finally i could hear my wordclock what trully sounds like.
1010101010101010.

remember that digital audio its in real time, not like downloading/uploading stuff in the internet, that does not need to be 100% realtime accurate.
in a online radio station maybe they have a super internet connection and you also, thats why buffer was invented for, but if the soundcard has jitter...

in the end, all you have to do its find the s/pdif & BNC cable that sounds good to you.<---------------
its verry fascinating to heard diferent cables,
most cables will sound in the same level/league.
some people cant hear the diference, anyway.
some others dont care.
Juanpablocuervo -

Electric current is not a "travell of electrons" - they move only about 1/2" per second and is also not a charge moved from electron to electron (wave) since they are spaced 10000 time their size apart. One theory says that electrons emit photons but this doesn't work either. Check Wikipedia!!!

While you describe properties of the cables why don't you mention inductance of the cable? It would be nice to mention characteristic impedance.

"in extreme capacitance, cuts more, making holes from the tiny bits&Bytes traveling at 3Mhz. and then paste on top of other bits, creating dissorder, or a blurr effect , creating digital errors, micro data lost. clock sync lost." - You must be kidding!!!!
In my experience, not only to different digital cables sound different, they actually have distinctive sounds. For example, I have a Goldmund cable that is consistently bright sounding, and a transparent Audio reference cable that is consistently warmer, smoother and more organic sounding.

I actually took a break from the hobby for several years and became skeptical of a lot of my audiophile beliefs, including my belief in the potential efficacy of cables and accessories. I was auditioning two components using both of the above digital cables, and the one using the Goldmund was simply bright and harsh. So, with much skepticism, I decided to switch the cables and wouldn't you know, the component that previously sounded harsh now sounded smooth and rich with the Transparent cable. It was like a different system.
although a digi cable only trasmits "0" and "1" - a beat of music may take 10k digits so the pace and accuracy will imprint themselves to the dac which will convert these to differing sounds. A simple a/b of different cables will produce much different sounding examples.