opaqueice;175906 Wrote: > OK, I guess that wasn't my last post... > > Take the actual waveform, subtract the ideal, and that's the noise > spectrum (or distortion, or choose whatever term you prefer - from now > on I'll refer to it as noise). Because everything here is linear, you > can think of the actual signal as the noise plus (overlaid on) the > ideal signal. Some of that noise comes from jitter, and some comes > from thermal noise in the analogue stage, and some comes from EMI, etc. > > Now, noise due to jitter is correlated with the signal in the following > sense: if you add some jitter with frequency x to a digital signal > representing an analogue tone with frequency y, you get noise at the > analogue output concentrated at frequency y-x and y+x, as well as some > other components. If you add that same interference to an analogue > signal, you just get noise at x (this is what I said two posts back), > so that could be said to be uncorrelated with the signal. Of course if > you add full spectrum noise to the digital signal you will get full > spectrum (possibly with a different shape) noise at the analogue out, > so there the difference is qualitatively much less. > > My point, again, is that the noise due to jitter is going to be much > harder to hear than the noise directly added to the analogue stage if > the source of both is the same old EMI. So if you don't hear any > change in the noise floor when you unplug the wallwart I'm very > skeptical there can be any change when music is playing. > > It would be interesting to see if there's some DAC design for which > this isn't true - maybe you could build a DAC which greatly amplifies > jitter at certain frequencies or something... what DACs are you two > using?
Your reasoning is only partially complete. In reality HF may be rejected by analogue circuits, and if they make it through you won't hear them anyway (unless mixed down into the audible spectrum by some non-linearity). In contrast HF injected into a DAC's clock (one way or another) will mix down into the audible spectrum. The resulting noise is correlated to the audio signal, and hence should be referred to as distortion -even if its properties are not a simple to pin down as normal non-linearities in analogue amplification. AND you don't get to hear this distortion separately from any signal. Comparing a low-jitter signal to a high-jitter one, just listening to the difference signal should be interesting. -- P Floding ------------------------------------------------------------------------ P Floding's Profile: http://forums.slimdevices.com/member.php?userid=2932 View this thread: http://forums.slimdevices.com/showthread.php?t=32231 _______________________________________________ audiophiles mailing list audiophiles@lists.slimdevices.com http://lists.slimdevices.com/lists/listinfo/audiophiles