Blackness - how quiet does it need to be?

Jitter in playback of digital source is basically noise proportional to amplitude of the signal and not detectable without signal.

Csontos - CDP is also an equipment and can be very noisy when playing but quiet when not. That way noise is undetectable without music (proportional to volume). In addition amplifier might create noise of its own by intermodulation etc. (speakers can also intermodulate).

Last_lemming, your amp might appear as quiet but in reality can be very noisy.

Many recordings were made with pre-emphasis to greatly reduce tape noise in playback. CDP or DAC has processor for it (de-emphasis) and every CD has bit switching it on or off in CDP or DAC. Today almost all recordings are digital and pre-emphasis is never used. Because of that many CDPs or DACs don't even have de-emphasis. None of Lavry DACs, for instance, has de-empasis but my Benchmark DAC1 does. If your CDP or DAC doesn't have de-emphasis and you listen to one of few recordings with pre-emphasis it will sound noisy. I'm not sure if tapes were recorded with pre-empasis (10db increase between 3kHz-10kHz) before invention of CD. Some people tested rippers and found that EAC does not correct for pre-emphasis (leaves recording intact) while Itunes does apply correction. Also some MP3 compressors like for instance LAME have de-emphasis option.

Last_Lemming I'm hesitating if I should even start since Al can explain it much better and more coherent way. If he does follow his explanation.

In both cases, digital and analog, we're talking about modulation that creates new frequencies.

Two frequencies in presence of nonlinear element (nonlinear transistor, nonlinear motion of speaker's membrane) produce additional signals of frequencies that are sum and the difference of original two frequencies. You can think of it as noise since it wasn't in the music and amplifier or speaker manufactured it. Great amplifiers or speakers modulate very little but we also "learned to listen" and can detect smallest differences in clarity.

Time jitter of your transport digital signal when playing single frequency also creates two additional frequencies (sidebands), but this time they are spaced on both sides of original/root frequency by modulation frequency distance (frequency of time jitter). Al mentioned once, that this modulation frequency can be completely random at the moment (random noise affecting digital signal timing) and we call it uncorrelated or can be cause by particular frequency like noise from switching power supply - then it is correlated. Usually it is mix of both. When complex signal (music) contains a lot of frequencies there is a lot of byproducts - basically a hash proportional to amplitude of the music. When music stops hash stops. It is detectable only as a lack of clarity and as such affects pretty much everything from tonality to imaging. Jitter usually has very small amplitude and byproducts (sidebands) are very, very small but still audible since they are not harmonically related to root frequency (like in harmonic distortion) and our ears are sensitive to it.

We're dealing wit two aspects of jitter: amplitude (modulation index) and the frequency. Small modulation frequency (like 60Hz) creates sidebands near root frequency that are easily masked therefore is less dangerous but high frequencies of the jitter create byproduct far apart from the root frequency and very audible. Amplitude of the jitter affects amplitude of the byproduct up to the point where even more than two sidebands can be created. Bad CDP have often a few nanoseconds jitter amplitude while good transports perhaps one tenth of it. I shouldn't really attribute it to transport but rather to system (transport, cable, DAC, ambient noise, power noise etc.)

Al, did I miss something?

Bryon, Jitter modulation produces all sorts of garbage across all frequencies raising noise floor. Normally you would not be able to hear it as a hiss because it would be covered/masked by loud signals and not audible during gaps or soft passages. Let's assume that jitter product is for instance, always -60dB down from the loudest music level. When you play softer passage jitter products are -60dB down from that being inaudible. The louder you play the more apparent it is showing as a lack of clarity (loud trumpet is not as pure as it could be etc.)

One possibility is that they suppress microphones' hiss before and after but I would agree with Al, that what you hear might be differential of high inaudible frequencies modulated on the tweeter or amplifier. These frequencies come from many sources including digital playback itself (quantization noise). New techniques of oversampling push quantization noise outside of audible range but it might come back when system is not perfectly linear and has additional noise sources (poorly filtered switching power supplies, RFI pickup etc). These high frequencies can have extremely high amplitudes. I remember that some of SACD players have bandwidth limiting switch (filter) to protect weaker tweeters from overheating just from inaudible quantization noise. Stereophile review mentioned that it was left by manufacturer to customers with warning of possible risk of turning off the filter switch. Tweeter membrane does not move much at very high inaudible frequencies but as long as it does it will modulate because its displacement vs. voltage is never perfectly linear.