The secrets of spill: Crosstalk - electronic spill
Although spill - as we shall find out later - can be a desirable thing, its close relation crosstalk is wholly negative in its effect. Crosstalk occurs in electronic systems where there is a path between one channel and another, not intended by the designer.
Electricity is pretty well behaved (certainly in comparison to magnetism) and tends to follow cables and circuit tracks fairly obediently. But it isn’t impossible that there could be an electrically conductive path that leads the signal astray, or a path that should be blocked isn’t entirely closed off. An example of this is where you pull a fader down all the way, but there is still a faint ghost of the signal present.
A good fader should give at least 85-90dB of attenuation, but this still isn’t equal to fully off. More likely however is that there will be capacitative coupling between channels. This means that there isn’t a direct conductive connection but the signal jumps from one path to another via its electric field. Inductive coupling would be possible also but in practice doesn’t seem to cause so much of a problem. The difference between resistive and capacitative coupling is that the former retains the full frequency response of the signal, the latter allows mainly high frequencies to pass.
Inductive coupling would allow mainly low frequencies to pass. Probably the worst problem that could occur due to electrical crosstalk is timecode leaking into audio channels. With digital equipment, this shouldn’t be a problem. I suspect it would have to be a major design or operational error to allow timecode to crosstalk in an all-digital system. In the analogue domain however it is possible to develop a very keen ear for timecode, and its sonic signature makes it clearly audible even when it is only present at very low levels.
The key to preventing timecode crosstalk is simply to keep it physically as far away from audio as possible. Course Directorly this means don’t let it into the mixing console, and preferably don’t allow it onto the patchbay either. Timecode circuits should have their own wiring loom, and I would probably go so far as to say that I would give it its own conduit or trunking for a lengthy run.
Of course this is fine in theory, but it is important that a studio should be flexible, and it is quite possible with analogue recording equipment that some users would want to use all the tracks and not give up any for timecode, therefore track 24 has to be wired to the console. It’s an imperfect world, and many of the otherwise very professional recordings that come my way remind me of this.