Same-sex background voices mask speech for listeners of either sex.

Evidence

Figure 6–17 shows a robust benefit from two primary factors: increasing the SNR and increasing the difference between the target and masker voice characteristics. The benefits of presenting the talker with a background a voice of a different sex were huge — at low SNRs performance increased by 40 percentage points! The largest effects were evident at the low SNRs, where performance was worst. The lack of benefit at the highest SNRs either may be due to ceiling effects, as performance cannot be better than 100%, or may reflect that the mechanism responsible for the release from masking is no longer necessary due to the ease of the task. — Jennifer Lentz, Psychoacoustics- Perception of Normal…, 1317