Ability to detect tones in noise was measured without defined observation intervals. Latency density functions were estimated for the first response following a signal and, separately, for the first response following randomly distributed instances of background noise. Detection performance was measured by the maximum separation between the cumulative latency density functions for signal-plus-noise and for noise alone. Values of d′, estimated by this procedure, were approximately those obtained with a 2-dB weaker signal and defined observation intervals. Simulation of defined- and nondefined-interval tasks with an energy detector showed that this device performs very similarly to the human listener in both cases. Subject Classification: [43]65.58, [43]65.75, [43]65.50, [43]65.68.
No takes yet. Share an insight, caveat, or question.
Watson et al. (1976) studied this question.