Speech reception thresholds (SRTs) were measured at a central focus of attention and at 20°, 40°, and 60° locations distant in azimuth. Measurements were taken with one target collocated with two maskers, or with maskers flanking the target by ±20°. For 80% of trials, the target was played from the attended location (“expected”), and 20% came from another (“unexpected”) location. For collocated stimuli, SRTs worsened with increasing distance from the expected location by 2.1 dB over the 60° azimuth and by 5.1 dB for spatially separated target and maskers. In spatially separated conditions, a 2.9-dB change was still found when gaze was away from the attended location. Spatial attention appears to increase speech intelligibility against interferers, with gain decreasing with distance from the focus of attention. Spatial release from masking (RFM) was only found for the attended location. Further experiments suggest that target location uncertainty prevented RFM at unattended locations.

中文翻译：

---

语音清晰度会降低与与会者所在位置的距离：注意力的听觉空间梯度的证据。

语音接收阈值（SRT）是在注意力的中心焦点以及方位角相距20°，40°和60°的位置处测量的。测量是在一个目标与两个掩膜同时放置的情况下进行的，或者与目标相距±20°的掩膜一起进行。在80％的试验中，目标是从参加人员的位置（“预期”）进行的，而20％是从另一个（“意外”）位置进行的。对于并置的刺激，SRT随着距预期位置的距离（在60°方位角上）增加了2.1 dB，而在空间上分离的目标和掩蔽物则增加了5.1 dB，从而使SRT恶化。在空间上分开的条件下，当视线离开参与人员的位置时，仍然发现2.9 dB的变化。空间注意力似乎增加了针对干扰者的语音清晰度，并且增益随着距注意力焦点的距离而减小。仅在有人参与的位置发现了从遮罩（RFM）释放的空间。进一步的实验表明，目标位置的不确定性阻止了无人值守位置的RFM。