Check this out:
Here is my two cents, limited in understanding, but for what its worth....
The interface between cable and component may be a source of reflections and the potential noise from reflections and the group delay characteristics of the cable itself could conceivably shift the clock signal detection slightly. However, a slight shift or drift is NOT a problem, as long as it is reasonably constant (a very gradual drift, such as the effect of gradual temperature rise in the circuitry, will not affect the audible signal).
"Jitter" becomes a problem when the shift is no longer consistent and the clock signal or the detection of the clock signal jumps around ever so slightly...sometimes slightly early and sometimes slightly late. Small amounts of random jitter variation are generally not regarded as audible, as they will not produce a consistent output signal much above the noise floor, however, correlated variation in response to some non random signal (an unstable power supply that oscillates for example) may be more detrimental and may indeed be audible (rise above the noise floor).
Jitter is certainly an "active" component problem. It does not come from a source disc or a hard drive or a cable...it comes from the lack of consistency in the clocking out the bits from the DAC circuitry (whatever causes this... even if it might be a result of the variation in power drawn by the transport itself).
Of course, changing cables might have an audible effect (if a cable is faulty or out of specifications or is poorly shielded and is picking up EM or RF non random noise) but the root cause of the problem remains with some active circuitry somewhere, such as power supplies, the clock, detection circuits, or RF/EM interference....as some "non-random" signal MUST influence the timing detection in order for tiny jitter variations to become audible.
Jitter makes a strong case for optical connections between digital devices, as optical cables will be completely immune to 60 Hz noise from power supply transformers or from radio signals (the most obvious sources of non-random noise that might affect clock signals ever so slightly).
Another solution to EMI interference is to deliberately "dither" the clock signal....the idea is that the dither should drown out any consistent non-random jitter effect due to EMI and make the noise uncorrelated and therefore inaudible. See the following link;
Other techniques to tackle "jitter" involve buffering and re-clocking, using more accurate clocks, good shielding and good power supplies. "Jitter" was more a problem of the early 90's, it is much less likely to be an audible problem with reputable equipment today.
IMHO, I would tend to invest in HI-FI components from reputable companies rather than to pour money into cables. (Surely I am not the only one that finds it concerning that cables can cost more than CD players!)