On the pro to consumer side, operating level differences are an important concern.
Because consumer inputs rarely include input level controls, the consumer equipment is easily overloaded by pro signal levels.
Again, since the professional reference is +4 dBu or 1.228 V rms and the consumer reference is -10 dBV or 316 mV rms, a loss of about 12 dB is required.
Obviously, the output of the pro equipment could be reduced by 12 dB, but then its level metering would be nearly useless and signal-to-noise performance would be degraded.
Signal attenuation is required for these interfaces.
In most cases, noise rejection is a far more important issue.
For consumer to pro interfaces, the widely used hookup of Figure 1 uses shielded single-conductor cable and an RCA to XLR adapter or ready-made adapter cables built as shown.
Figure 1 - Using an unbalanced cable with an adapter results in zero noise rejection.
Unfortunately, it has 0 dB of ground noise rejection - and wastes all the potential noise rejection of the balanced input! Sadly, the availability of such adapters or cables leads many unwary users to create this noise-prone connection. Performance is especially poor when cables are long, since the entire interface is unbalanced, allowing both audio and ground noise to flow in the cable shield.
A far better hookup shown in Figure 2 uses shielded twisted-pair cable to take advantage of the noise rejection available from the balanced input stage.
Figvre 2 - Using balanced cable wired as shown results in at least 30 dB rejection.
Because ground noise now flows in the shield conductor rather than one of the signal conductors, noise rejection is improved by about 30 dB when the input is a typical “active” differential-amplifier type. If the equipment’s balanced input is truly high-performance, using an input transformer or the InGenius IC, rejection is improved by about 80 dB. [Reference 1]
Figure 3 shows noise rejection for various consumer to pro interfaces. The top plot at 0 dB represents the simple adapter and 2-conductor cable connection.
Figure 3 - Noise rejection in unbalanced to balanced interface. Top to bottom: cable of Figure 1, cable of Figure 2, cable of Figure 2 plus output transformer, cable of Figure 2 plus input transformer.
The plot at -30 dB shows the improvement due to the simple 3-conductor hookup. The next plot shows the effect of an ordinary isolator using an output transformer (no internal Faraday shield!). It improves 60 Hz hum by about 20 dB, but has little effect on buzz artifacts around 3 kHz. A high-quality isolator using an input transformer (with an internal Faraday shield) increases rejection to almost 100 dB at 60 Hz and about 65 dB at 3 kHz.
For the best possible noise rejection, use a 3-conductor cable (wired as in Figure 2) from the unbalanced output to the isolator input. The plots here were measured with a 600-ohm unbalanced output and a 40-kohm balanced input.