I can't envision a reason why the length of the cable between source and preamp would be any more critical than with an active preamp.
With passive controls, the cable length is quite critical as is the construction of the cable. The interaction is between the source impedance, the installed variable of the passive control, the characteristics of the cable that goes from the passive to the amplifier, and the input impedance of the amp. Its a bit complicated, but with passives what you will universally hear is that as you turn down the control, the bass and impact in the system will be diminished. It does not matter the quality of the control itself- that part is really not a variable, although the *value* of the control is.
The problem there is that a lot of sources may not be so happy with a 10K passive control hung of its output; even then you will still hear effect.
TVCs overcome some of these disadvantages. They can present a much lower impedance to the interconnect cable at their output- this reduces the effect of the cable and can reduce the variables in that rather complex interaction I pointed to above. The price of TVCs is that they must be designed to be loaded properly at each tap on the transformer, else they can 'ring' or distort. In addition, if improperly loaded, the inter-winding capacitances of the device will come into play. This is more or less like saying a random capacitance is being used to bypass the TVC. This will be a path for high frequencies.
So great care has to be exercised in the design of a TVC. Even so, If you want the system to play bass, the TVC must have bandwidth to 2 Hz if the system is the play to 20Hz without loss. If the unit cuts off at 10Hz, effects that will be heard as losses will extend to 100Hz. Making transformers that go that low without problems in the highs is a really challenge!
Active preamps have advantages of cable control as well, although if the designer of the active line stage does not recognize that that advantage may not be realized. Now f things are done right the effect is that the cable will have no sonic contribution at all, and the length of the cable will be irrelevant. I think you can see from this last statement that many line sections fall well short of this ability! Quite often in the use of a line stage the hidden cost of the line stage is the cost of the cable that makes it sound right. In such cases its easy to see how a passive or TVC would be preferred.
But if an active line stage is properly executed, the length of the cable and its cost will have no bearing on the sound of the system at all. The advantage here is you can avoid the colorations of a passive, and potentially get greater bandwidth and lower distortion (read: smoother and more detailed) than is possible with a TVC.
Now in the case of DACs and CDPs, quite often the output of such units is prodigious, far more than is needed to overload even the least sensitive amplifier made. This requires that the output of most DACs and CDPs must be reduced (attenuated) by some means. A digital control is not a good way to do that- such controls work by essentially subtracting bits from the signal. By the time you get below about 85% of the range of the control, a significant loss of resolution is audible.
A simple solution would be for digital equipment manufacturers to go back to the 1V peak standard that has been in place for 40 years. It pretty obvious now that LP will not be going away as the digital community had predicted (I remember seeing on TV an RCA executive saying that they were going to discontinue LP production in 1987... lol!). At any rate its not to the advantage of anyone using an all-digital setup to have the output voltage be so high that all amps made are overloaded by it.
Active buffers are a good choice in such a case. They avoid interactions with the cable by keeping the volume control intimate with the rest of the circuit and have the possibility of cable control just like an active preamp (which in reality is what they really are- active, but with no gain). We've built a few such devices by getting rid of the gain stage in our preamps and they have worked quite well, but in such a case you really do need the extra output that DACs and CDPs provide. The problem there is that if you want to use a phono section as well, it will either have to have a very high output or it will not be usable...
So actives remain my first choice but you knew that by now :)
