difference by changing the vibration pattern of the CD being played, and thereby changing the current transients drawn by the servos.
Besides the obvious servo problem that many designers have tried to address in their premium CD players, there are a host of other, more subtle problem areas where so-called digital circuitry is vulnerable to analog influences. Even small perturbations within the digital circuitry itself can have unwanted influences upon that same circuitry, leading to self or auto modulation distortion in this world where even microscopic unwanted variations of .0001% (one part in a million) can cause distortion. For example, if the music signal just happens to be changing such that, in going from one digital sample to the next, all 16 or 20 bits have to switch their digital state, then that will force the relevant digital circuitry stages to suddenly draw more current (as well as causing more radiated switching transient interference, q.v. discussion below). That perturbation might then in turn affect the exact analog threshold level at which some digital circuitry switches digital states, or might affect the exact analog amount of time some relevant capacitor takes to charge. And, as we now know, such variations or indeterminacy in analog threshold levels and/or analog timing can distort your music.
There are other unwanted external influences, besides the current draw spikes discussed thus far. Consider for example induced noise. Again, the popular misbelief is that digital circuitry is either on or off (thereby representing either a 1 or a 0), and that therefore it is substantially immune to any noise in the system (at least any reasonably moderate amount of noise, say below the voltage differential between the 0 and 1 digital states of the circuit). But now we know that these so-called digital circuits are actually analog in their internal operation. They are supposed to switch their digital states at a precise analog threshold level of voltage (or current), and at a precise analog moment in time. But if the exact analog level of this threshold varies or is unpredictable from one music sample to the next, then distortion of your music can result; and likewise if the exact analog moment in time of switching varies or is unpredictable from one music sample to the next, then distortion of your music can result.
Now, the analog voltage ramp input, which a digital circuit looks at, waiting for the threshold to be crossed to tell it to switch digital states, should be a very steep, clean, straight line, at least near the threshold crossing (in fact, the steeper the better, since then the exact analog moment of threshold crossing is more precisely and repeatably and predictably determined). But suppose instead that there is a small amount of analog random noise that has somehow been induced into the input line to this digital circuit. That noise, riding on the larger waveform of the rising ramp, means that the ramp is no longer a clean, straight line. Rather, it is now a fuzzy line, full of little squiggles, which of course represents the analog noise riding on the analog ramp input that should be a clean, straight line. These squiggles mean that the exact threshold will be crossed when the squiggles riding on the ramp say so, not when the ramp itself was designed and intended to say so. Thus, the ramp itself will not be at the correct analog amplitude level when the analog threshold is reached that causes the digital circuit to switch digital states. Furthermore, the digital circuit will not switch digital states at exactly the correct moment in time that the master so-called digital clock was intending it to. In other words, this digital circuit will switch states at the wrong input amplitude and at the wrong time. As we now know, that error can cause distortion of your music.
To make matters worse, that error will be different for each music sample, since the noise squiggles riding on the ramp waveform keep changing over time. And these errors will be randomly and unpredictably different, assuming the induced noise is random. Indeed, you can imagine a real travesty if you imagine a narrow noise spike occurring just a few split moments before the threshold of the main ramp was supposed to be reached. This spike could easily cause the voltage to drive through the threshold not only at the wrong time, but actually at two different times. The spike could rise above the threshold, then fall below the threshold, and then the main ramp would finally drive up through the threshold again. In theory, this could cause a major distortion disaster by making the digital circuit switch on, then off, and then on again, within the one sampling time period when it was supposed to turn on only once (in practice, the truly analog, so-called digital circuit might well have sluggish analog time constants built in that would hopefully prevent it from being fast enough to execute such tragic false double triggering).
Note that a random error in time, caused merely by induced noise, is just as bad as time jitter caused directly by a poor or unstable master clock. Jitter might get all the publicity as an unwanted evil nowadays, but noise is just as bad.
Thus, these so-called digital circuits are in fact just as vulnerable to induced noise as purely analog circuits are. Indeed, they are even more vulnerable, since the consequences can be more degrading to your music. If a purely analog circuit is fed random noise in addition to the music signal, it pretty much just keeps the noise as simple background noise added to the music. But if a so-called digital circuit is fed random noise in addition to its analog input ramp commanding it to switch on, this can easily cause distortion of your music. Distortion is far worse than background noise, especially since this distortion is not harmonically related to the music. There is evidence that high frequency time jitter creates an FM distortion that sounds like ugly smearing, especially of music's high frequencies, with considerable degradation of what should be black intertransient silence. If white noise or high frequency noise were induced into digital circuit lines, the distortion this could produce should sound similarly ugly.
A third lesson is that the signals traveling throughout your digital playback unit are even more vulnerable to unwanted external influence because they actually travel not through conductors or wires, but rather through the space outside the conductors. This space outside the conductors contains the insulators (called dielectrics) for the conductors, and a whole lot of air inside your CD player. This space may well be contaminated, especially in a CD player environment where there are many unwanted sources of pollution, radiating high frequency noise and spikes into the space surrounding the sensitive, vulnerable circuitry (see further discussion of this in IAR issue 45).
The model of electricity you were taught in school, and that many design engineers still believe in, is a hydraulic model. Electricity flows inside a wire like water flows inside a pipe. And, just as water securely ensconced inside a pipe is protected from external influences, so electricity securely flowing inside a wire is relatively safe from external influences that are in the air outside the wire. But the truth is very different. In truth, as discussed in IAR issues 11, an electrical signal does not travel inside a wire. Instead, it travels in the space outside a wire. A wire merely serves as a guide, telling the electromagnetic wave carrying the electrical signal where in space or air to go. An apt analogy is a train traveling on railroad tracks. The train does not travel inside the rails of the track, but instead travels in the air outside the rails, and the rails merely serve to guide where in the air the train will travel.
Both analog and digital signals travel in the space outside the wires of your CD player's circuitry. And, as we saw above, the so-called digital signals are actually analog signals that are as vulnerable to corruption as purely analog signals are (or indeed are even more vulnerable).
Now, the space inside your CD player, just outside the wires of the circuitry, is highly polluted, full of noise radiated by the player's own circuitry. This circuitry (especially the digital circuitry) handles RF and radiates RFI, generates noise, and generates overshoot switching spikes that also radiate polluting noise. Thus, there's a witches' stew of polluted electromagnetic garbage in the space inside your CD player, most of it generated by the digital circuitry itself.
Some premium CD players make a big selling point of a feature whereby they separate and shield the purely analog output section from all the prior digital circuitry. This undoubtedly accomplishes some good. But it is not nearly enough. For, as we have just seen above, it is the digital circuitry which is just as or even more vulnerable than the analog circuitry, vulnerable to unwanted external influences. So it is the digital circuitry that needs the most protection. Protection from what? From itself.
We see now that the vulnerable signals for the digital circuitry, that we want to keep safe from external influences, are actually traveling unprotected in the space outside the wires, not safely ensconced inside the wires. We saw above that these signals for the digital circuitry are actually analog, and are just as vulnerable to external influences as purely analog signals. And we also see now that the space in which these vulnerable signals are traveling is polluted by the very circuitry that these signals are serving, the very circuitry that needs these signals to stay pure and uncorrupted. The digital circuitry generates and radiates garbage that travels as electromagnetic waves in the space just outside the digital circuitry, within your CD player's box. And the wanted musical signal information, being processed by this same digital circuitry, also travels as electromagnetic waves in the space just outside the digital circuitry. The two kinds of electromagnetic wave information, unwanted garbage and wanted musical information, are mixing it up in the same space. Naturally, the wanted musical information in the air gets corrupted by the garbage in the same air. And, since so-called digital circuits are actually vulnerable in an analog fashion to degradation by noise and garbage, they can produce distortion of your music when the wanted signal they receive has been corrupted. For example, as we discussed above, noise added to what should be a clean, straight voltage ramp can switch a digital circuit at the wrong instant of analog time, thereby effectively causing time jitter, and ultimately causing distortion of your music when this digitally switched signal reaches the DAC chip at not precisely the right sampling time (remember that, when the CD player re-creates the two dimensional music waveform, the right amplitude re-created at the wrong time will distort your music waveform as surely as the wrong amplitude re-created at the right time).
Obviously, the digital circuitry needs to be shielded, not just from the outside world, but also from itself. Digital circuitry (which is really analog) should be better designed, so it generates and thus radiates less RFI, noise, switching spikes, etc. Then the digital circuitry (including wiring and supporting parts) should be meticulously subdivided so the parts can be effectively shielded from one another. In many cases, this might even require re-designing digital IC chips, subdividing them and incorporating internal shielding within the chip.
Until such drastic corrective measures are taken, we have to endure digital circuitry that lives in a music-corrupting polluted environment. And you can see why any tweaks that we can invent to modify this polluting environment will doubtless make a sonic difference. For example, some people have found that they can change the sound of CD players by putting a layer of rubber on top of digital circuit IC chips. These people theorize that the rubber is damping mechanical vibrations that somehow affect the relationships among the elements inside the IC chip. But these elements are anchored in place, and the spatial relationships among them cannot be changed by mechanical vibrations. The correct explanation for why the rubber changes the sound probably has nothing to do with mechanics, or mechanical vibrations. Instead, the correct explanation is probably that the rubber as a dielectric affects the electromagnetic fields just outside the chip top (every dielectric placed in an electromagnetic field affects that field). The rubber dielectric would modify some of the garbage electromagnetic fields just above the IC chip, where they are closest to and therefore most likely to corrupt the wanted signals traveling (most intensely) inside the IC chip (just outside the conducting wires within the IC chip). Also, the rubber dielectric would conversely modify some of the electromagnetic garbage radiated upward by the IC chip's internal switching spikes, etc., and might thereby ameliorate some of the garbage radiated into the air by this IC chip, garbage that might well adversely corrupt other nearby digital circuitry. Thus it is that this rubber dielectric (or indeed any dielectric) placed just outside a digital IC chip can affect the sound of a CD player.
Of course, what's most important here, and most remarkable, is that there is any sonic change at all from merely placing a dielectric close to a circuit that is supposedly digital and supposedly immune to such external influences. This sonic change in itself (whether it is for the better, for the worse, or a mixed bag) is strong corroboration of our argument that so-called digital circuits are actually analog, and are vulnerable in analog ways to external influences that they should not be vulnerable to at all if they were truly digital circuits as popular misbelief supposes.
Incidentally, it is not really sonically advisable to put rubber sheets on top of IC chips, be they analog or digital. You see, all dielectric materials influence and alter electromagnetic waves that travel through them, thereby coloring the signal with a signature coloration and time constant pattern characteristic of that material. You can hear this signature coloration of dielectric materials imposed as artificial foreign colorations on the music. Hard plastics impose one kind of characteristic coloration upon music, typically a hard glare in the upper midrange and lower treble (even the vaunted Teflon dielectric, a pretty hard plastic, degrades the music signal with a hard plastic sound). Rubber imposes another kind of characteristic coloration upon music, typically a long time constant hangover that makes the music sound, yes, rubbery. If you add a rubber sheet on top of a hard plastic IC chip case, you are merely adding a soft rubbery hangover to music that sounds too hard and glary from the hard plastic case. The soft rubbery coloration might seem to offset the hard, glary coloration, but two wrongs don't make a right, and you're really creating a complex stew of compound colorations that's taking you farther away from the original music signal, rather than getting closer to the original music signal. Note that the rubbery sonic coloration imposed by rubber makes it unwise to use rubber anywhere near any electronic circuits carrying music. For example, it is unwise to mechanically anchor capacitors in place with silicone rubber compounds.
Since the electromagnetic waves actually carrying the signals for and within any IC chip actually travel in the space outside the wiring of that IC chip, these signal carrying waves are traveling in the dielectrics, including the hard plastic case of the IC chip, and then also any additional dielectrics you might add on top of the IC chip's hard plastic case. Since the so-called digital signals within even purely digital IC chips are actually analog signals, even digital IC chips are vulnerable to unwanted coloration by the poor quality dielectrics of the hard plastic case, and then further unwanted coloration by any additional dielectrics you might add on top of the case.
In our lab research, we found that we could obtain the best sonic results, and get back as close as possible to the original music, by critically damping the problem in the hard plastic case dielectric, but using the absolute minimum of additional dielectric material to do so, in order that we would not be adding new colorations from any significant new amount of added dielectric. In our lab, we developed a micro-thin coating that could be applied to the hard plastic cases of IC chips, and that accomplished sonic benefits by reducing colorations of the hard plastic case dielectric without adding new colorations of its own. It proved to be sonically beneficial for both analog and digital IC chips, for video circuitry, and indeed for all electronic parts that have hard plastic cases (transistors, regulators, resistors, and capacitors in hard plastic cases). We turned the results of this lab research over to our sister company, TRT, and they manufacture this special fluid under the name MusiCoat (you can visit TRT's website at trt-wonder.com). Others have praised the sonic effectiveness of this micro-thin fluid treatment, and this is not an appropriate place for self-praise anyway, so we refer you to other websites for impartial third party evaluations (e.g. SMc audio, audioasylum, etc.).
Analog Eye Pattern --
A fourth lesson is that the sound of a CD player is apparently very susceptible to the quality of the analog eye pattern. The "eye pattern" refers to the raw analog input signal picked up by your CD player as a reflection from the laser bouncing off the CD. The waveform of this input signal that is read off the CD is actually very much analog. It is smooth and rounded (looking much like a sine wave but with flatter tops and bottoms), nothing at all like the sharply cornered square waves that represent the digital information throughout the later stages of your CD player.
Ideally, this smoothly undulating analog waveform should have a high amplitude and steep zero crossings. And it should also ideally be very stable temporally, with the zero crossings coming at precisely optimum moments of periodicity, and without temporal jitter, wander, or drift. This also implies that ideally it should be very clean waveform, without added noise (which, as we saw above, causes temporal indeterminacy).
Incidentally, this smoothly undulating analog waveform is called an eye pattern as a slight misnomer. Only when successive iterations of this waveform are overlaid on an oscilloscope do diamond shaped openings appear which look something like eyes. In a higher quality analog input waveform from the CD, the diamond eye will be taller, with steeper zero crossings, and with zero crossings that always cleanly and sharply occur at the exact same temporal points of periodicity, without fuzziness on the oscilloscope (that fuzziness would indicate temporal jitter or wander).
If the analog input waveform coming off the CD has all these ideal qualities, then your CD player's circuitry can interpret this waveform more easily and more accurately, in order to figure out where and exactly when the 1's and 0's are supposed to be placed in the digital signal created by your CD player. Note emphatically that the data being read off the CD are not digital, contrary to popular misbelief. There are no digital data being read off the CD. The input signal from the CD is a purely analog waveform that must be interpreted by your CD player, which then generates a digital bit stream that hopefully corresponds to the digital bit stream that existed in the recording studio. The creation and generation of the digital data stream actually occurs entirely within your CD player, and is fully the responsibility of your CD player.
There are many factors that can degrade the quality of this analog input waveform. If the CD is less than ideally reflective, perhaps because there is still mold release compound on a brand new CD or because an old CD is dirty or scratched, then the amplitude will suffer and the zero crossings won't be as steep and well defined temporally. If some of the laser light is scattered into the disc and somehow later finds its way back to the photo detector, that will make the input signal corrupted
(Continued on page 56)