I’m currently setting up a small home studio for songwriting and demos and I’d like to use split-input monitoring, so I don’t have to worry about latency. My initial plan was to split the output of my vocal mic at my patchbay, and send one output to my Focusrite Scarlett interface for recording, and another to my Yamaha MG06X mixer for monitoring. I realise that phantom power is a problem, so I’ve got an ART Phantom II Pro [standalone phantom power supply]. What I’m not sure about is how this will affect impedance. Will it be a problem for the mic, as it will ‘see’ two input loads? And what’s the best way to do this: a Y-cable, a passive mic splitter, or an active mic splitter?
SOS Forum post
Technical Editor Hugh Robjohns replies: Two thoughts spring immediately to mind. First, the Scarlett interface already offers latency-free monitoring of its inputs, which makes me wonder whether you really need the external monitoring mixer. Second, while the short converter/DSP delay can be a problem for some people when singing and monitoring over headphones, it doesn’t bother many at all — so before you get involved in the complications of splitting signals it would be worthwhile trying a simpler setup! If latency does cause a performance problem for you when singing, arranging latency-free monitoring is the right way to go, and there are many possible solutions — none of which require a separate phantom supply unit.
The easiest, and technically the best, approach is to use a standalone mic preamp for your vocal mic. The mic sees a single input, from which it receives its phantom power, so the mic is happy. The preamp’s line output can be split passively to as many destinations as you like, without losing any significant signal level or loading the output drivers, because the output has a very low impedance (typically 100Ω) and line inputs have a relatively high impedance (>10kΩ). The only potential problem is of ground loops between the various destination devices, but that can be resolved easily, by isolating the screen connection at the destination end of any affected output cable.
Splitting at mic level can be done, but it’s inherently a (slightly) more compromised approach. The best way, by far, is to use a decent transformer-based mic splitter, which typically provides one direct and one or more electrically isolated output(s). That arrangement neatly avoids the phantom power issue you were concerned about by ensuring that only one preamp is connected directly to the mic and can pass phantom power; the other preamp is electrically isolated from it by the transformer. So, if still possible, I suggest that you’d be much better off swapping your ART phantom supply for the same company’s ProSplit transformer mic splitter!
Turning now to your impedance concerns, if you were to split the mic signal with a simple Y-cord the mic would indeed ‘see’ the two input destinations effectively in parallel with each other, and that will result in the effective load for the microphone being reduced below the impedance of either. The formula, where Za and Zb are the input impedances of the destinations, is:
Z total = 1/(1/Za+1/Zb)
The input impedance of most Scarlett interfaces is around 3kΩ, and that of the Yamaha probably around 2kΩ. So the total in a parallel connection is roughly 1.2kΩ. Modern microphones work in a voltage-matched environment, in which the preamp input impedance needs to be between five and 10 times the output impedance of the mic. So, if the mic has a nominal output impedance of 150Ω, it should see a destination impedance of at least 750Ω, and ideally about 1.5kΩ. In the example above, the mic probably won’t notice anything amiss — because we’re comfortably towards the high end of that range.
In fact, transformerless capacitor and active-dynamic mics won’t have a problem at all, because the output electronics aren’t very fussy about load impedances anyway. But if the impedance gets too low, you might possibly notice a small change in the tonal character or transient response with traditional dynamic mics and transformer-output capacitor mics.
So, what’s the best solution in your situation? An active splitter is basically a low-gain mic preamp with multiple outputs, all at mic level, and these devices are generally designed for live-sound purposes, where low noise and distortion isn’t as critical as it might be for high-quality recording. It has the advantage that the mic is connected to a single phantom-providing preamp, but the disadvantage that it puts more active electronics in the mic’s signal path. A similar but better approach, as I described earlier, would be to use a good stand-alone mic preamp, and split the output at line level.
The next best option would be a passive splitter which uses a transformer to generate one or more isolated mic-level outputs, again as described above (and shown in the image above). The quality is determined by the quality of the transformer, so really good ones (that can handle very high levels of very low frequencies with negligible distortion) are expensive. But in most typical applications, and especially with vocal recording, I doubt you’d notice any significant issues, even with budget splitters.
A simple Y-cord (parallel split) can certainly work, and I’ve used that technique myself on occasion, but it does open the door to conflicting phantom-power supply problems, and makes ground loops very likely. The way around the ground-loop issue — this applies to active splitter/preamp techniques too, of course — is to disconnect and isolate the cable screens for all but one of the outputs at the destinations. Balanced connections don’t employ the cable screen as a ground reference, so disconnecting it won’t affect the signal transfer (but it is required for the return phantom current, so needs to be present in the connection to the preamp providing phantom power).