First spike latency has been suggested as a source of the information required for fast discrimination tasks. However, the accuracy of such a mechanism has not been analyzed rigorously. Here, we investigate the utility of first spike latency for encoding information about the location of a sound source, based on the responses of inferior colliculus (IC) neurons in the guinea pig to interaural phase differences (IPDs). First spike latencies of many cells in the guinea pig IC show unimodal tuning to stimulus IPD. We investigated the discrimination accuracy of a simple latency code that estimates stimulus IPD from the preferred IPD of the single cell that fired first. Surprisingly, despite being based on only a single spike, the accuracy of the latency code is comparable to that of a conventional rate code computed over the entire response. We show that spontaneous firing limits the capacity of the latency code to accumulate information from large neural populations. This detrimental effect can be overcome by generalizing the latency code to estimate the stimulus IPD from the preferred IPDs of the population of cells that fired the first n spikes. In addition, we show that a good estimate of the neural response time to the stimulus, which can be obtained from the responses of the cells whose response latency is invariant to stimulus identity, limits the detrimental effect of spontaneous firing. Thus, a latency code may provide great improvement in response speed at a small cost to the accuracy of the decision.

Phase-locked responses to pure tones have previously been described in the primary auditory cortex (AI) of the guinea pig. They are interesting because they show that some cells may use a temporal code for representing sounds of 60-300 Hz rather than the rate or place mechanisms used over most of AI. Our previous study had shown that the phase-locked responses were grouped together, but it was not clear whether they were in separate minicolumns or a larger macrocolumn. We now show that the phase-locked cells are arranged in a macrocolumn within AI that forms a subdivision of the isofrequency bands. Phase-locked responses were recorded from 158 multiunits using silicon based multiprobes with four shanks. The phase-locked units gave the strongest response in layers III/IV but phase-locked units were also recorded in layers II, V and VI. The column included cells with characteristic frequencies of 80 Hz-1.3 kHz (0.5-0.8 mm long) and was about 0.5 mm wide. It was located at a constant position at the intersection of the coronal plane 1 mm caudal to bregma and the suture that forms the lateral edge of the parietal bone.

Stellate cells within the ventral cochlear nucleus (VCN) are a diverse cell group that have been classified according to their size and morphology. Some of these stellate cell types constitute major projection neurones into the brainstem and directly into the inferior colliculus, while others are implicated in more local processing. It is still not clear whether a specific physiological profile is uniquely associated with each distinct type of stellate cell. To investigate such associations, we have analysed 23 units with a battery of physiological stimuli in vivo and then examined their shape and outputs following juxtacellular labelling with biocytin. Five physiologically identified groups of cells were filled. These formed two major response classes: onset cells and chopper cells. The two classes could be separated purely on morphological grounds. The onset cells had large somata, large symmetrical dendritic trees and profuse axonal branches that were restricted to the cochlear nucleus on one (On-L) or both sides (On-C) of the brainstem. The chopper cells had smaller, asymmetric, dendritic trees, which were either planar or marginal, had smaller somata and an output axon that left via the trapezoid body. We have confirmed profuse projections into the dorsal cochlear nucleus from all onset cells, and more focal projections from some members of all three groups of chopper cells.

Sensitivity to changes in the interaural time difference (ITD) of 50 msec tones was measured in single units in the inferior colliculus of urethane-anesthetized guinea pigs. ITD functions were measured with 100 repeats and fine spacing (100 points per cycle). The just noticeable difference (jnd) for ITD was determined using receiver operating characteristic (ROC) analysis of the spike-count distribution at each ITD. The jnd became progressively smaller as the signal frequency increased from 50 to 800 Hz but became unmeasurable above 1 kHz. The lowest jnds (30 microsec) were comparable with human jnds, indicating that there is sufficient information in the firings of individual neurons to permit discrimination without obligatory pooling. ROC analysis requires the choice of a reference ITD from which the jnd may be found by stepping the target ITD through the ITD function. For each neuron the reference was chosen to minimize the jnd. The lowest jnd was usually for ipsilateral leading references, near the minimum of the ITD function where the variance was also low, but where the slope was nearing its steepest. This was despite the peak of the ITD function occurring for contralateral leading stimuli. When the reference ITD was on midline, a jnd could be obtained by looking for firing rates either greater or smaller than the firing rate at midline. The lower jnd was usually obtained by looking for a decrease in firing rate. As duration increased, jnds either decreased or increased, depending on unit type, whereas when level increased, jnds generally increased.

At the level of the brainstem, precise temporal information is essential for some aspects of binaural processing, while at the level of the cortex, rate and place mechanisms for neural coding seem to predominate. However, we now show that precise timing of steady-state responses to pure tones occurs in the primary auditory cortex (AI). Recordings were made from 163 multi-units in guinea pig AI. All units increased their firing rate in response to pure tones at 100 Hz and 46 (28%) gave sustained responses which were synchronised with the stimulus waveform (phase-locking). The phase-locking units were clustered together in columns. Phase-locking was generally strongest in layers III and IV but was also recorded in layers I, II and V. Good phase-locking was observed over a range of 60-250 Hz: some units (30%) were narrow band while others (37%) were low-pass (33% were not determined). Phase-locking strength was also influenced by sound level: some units showed monotonic increases in strength with level and others were non-monotonic. Ten of the units provided a good temporal representation of the fundamental frequency (270 Hz) of a guinea pig vocalisation (rumble) and may be involved in analysing communication calls.

We report the spectrotemporal response properties of single units in the primary (A1) and dorsocaudal (DC) fields, and the ventrorostral belt of the urethane-anaesthetised guinea pig auditory cortex. Using reverse correlation analysis, spectrotemporal receptive fields (STRFs) were constructed and subsequently classified according to a novel qualitative scheme that was based on the duration and bandwidth of excitatory and inhibitory regions within the STRF. The STRFs of units in both A1 and DC showed either broad-band (> or = 1 octave) or narrow-band (< 1 octave) excitatory and inhibitory regions occurring either alone or together. The excitatory regions were of short duration (lasting for <50 ms) or more sustained (up to about 100 ms) and inhibitory areas either followed excitation or were located as inhibitory sidebands along the high- and low-frequency edges of the excitatory regions. Inhibitory areas that followed excitatory regions were found to be either short lasting (10-20 ms) or longer lasting (up to 200 ms or more). The STRFs recorded from each cortical area indicated temporal response properties consistent with those shown by traditional peristimulus time histogram analysis. Overall, fields A1 and DC showed no significant differences in the distribution of STRF types. Thus, it appears that both fields display similar spectrotemporal sensitivities to auditory stimuli and therefore, appear to process such stimuli in a parallel fashion. Single units recorded in the ventrorostral belt area showed STRF types similar to those recorded in A1 and DC. However, the proportions of STRF types were significantly different, suggesting a difference in spectrotemporal processing between the ventrorostral belt and the core areas.

The binaural masking level difference (BMLD) for pure tones masked by broadband noise is a robust and large (up to 15 dB) release from masking due to differences in the signals reaching the two ears. Physiological experiments have demonstrated that neurones in the auditory midbrain nucleus (the inferior colliculus) respond to stimulus configurations that evoke BMLDs in ways that are consistent with their sensitivity to the interaural delay of the masking noise and signal. The responses of single neurones are entirely consistent with models of binaural processing based upon cross correlation of the signals at the two ears. The lowest masked thresholds appear to be due to desynchronisation of the activity in response to the masking noise. Different populations of neurones may contribute to detection under different binaural configurations. The values of BMLD and the within channel signal-to-noise ratios found in single neurones in the guinea pig are smaller than found psychophysically in humans.

Sound localization in humans depends largely on interaural time delay (ITD). The ability to discriminate differences in ITD is highly accurate. ITD discrimination (Δ ITD) thresholds, under some circumstances, are as low as 10–20 μs. It has been assumed that thresholds this low could only be obtained if the outputs from many neurons were combined. Here we use Receiver Operating Characteristic analysis to compute neuronal Δ ITD thresholds from 53 cells in the inferior colliculus in guinea pigs. The Δ ITD thresholds of single neurons range from several hundreds of μs down to 20–30 μs. The lowest single-cell thresholds are comparable to human thresholds determined with similar stimuli. This finding suggests that the highly accurate sound localization of human observers is consistent with the resolution of single cells and need not reflect the combined activity of many neurons.

Phase-locked responses to pure tones are a characteristic of most auditory cells at the level of the brain stem and allow sophisticated analyses based on coincidence detection. Phase-locking to tones has not previously been shown at the level of the auditory cortex in single unit studies. We have now identified phase-locked responses in 10% of low-frequency (< 1 kHz) units in the ventrorostral belt, a strip of cortex immediately ventral to the primary auditory area. All of these units showed phase-locking in their response to binaural tone pips of 60-200 Hz and showed narrow band pass characteristics within this range.

Convergent input from cells in the medial superior olive (MSO) and lateral superior olive (LSO) onto a single inferior colliculus (IC) cell explains many findings that are not compatible with a simple coincidence detector mechanism. Here this explanation is tested using a physiologically accurate computer model of the binaural pathway in which the input to the IC cell is either from two MSO cells or a MSO and a LSO cell. Auditory nerve (AN) spike trains are formed by a stochastic hair cell model following a basilar membrane simulation using a gammatone filter. In subsequent cells input spikes cause post-synaptic potentials (PSPs) which are summed causing the cell to fire when the sum crosses a threshold. The individual cells are matched to the physiology by varying the number of inputs, the magnitude and duration of the PSPs and the firing threshold. Non-linear best-phase-versus-frequency functions arise if the two IC inputs have different best frequencies and different characteristic delays. One input can be selectively suppressed by turning on an additional tone at the worst phase of that input. Non-zero characteristic phases arise if the characteristic frequencies of the AN fibres feeding into a single superior olive cell are mismatched.

This study investigated the nature and topography of binaural interactions in the primary auditory field (AI) and dorsocaudal field (DC) of the urethane anaesthetised guinea pig auditory cortex. Single and multi-units were classified by their responses to monaural and binaural stimulation. In both AI and DC, units displayed binaural facilitation, binaural inhibition, or a level dependent mixture of facilitation and inhibition. There was a significant difference in the distribution of binaural response types between the two fields. Facilitated units predominated in DC (facilitated: 58%; inhibited: 24%; mixed: 6%; non-interacting: 12%), while inhibited units were the most common class in AI (facilitated: 15%; inhibited: 44%; mixed: 18%; non-interacting: 22%). It has previously been suggested that inhibited and facilitated units are concerned with processing different areas of space suggesting a possible separation of function between the two core fields. Topographically, the binaural response properties in AI and DC varied along isofrequency bands, with neurones displaying similar interactions aggregating in clusters. These clusters were similar in size for the two fields and often overlapped neighbouring isofrequency bands. However, their shape and position varied between different animals. This clustered organisation of binaural interactions is similar to that reported in recent studies of AI in other mammals.

Responses to sound stimuli that humans perceive as moving were obtained for 89 neurons in the inferior colliculus (IC) of urethan-anesthetized guinea pigs. Triangular and sinusoidal interaural phase modulation (IPM), which produced dynamically varying interaural phase disparities (IPDs), was used to present stimuli with different depths, directions, centers, and rates of apparent motion. Many neurons appeared sensitive to dynamic IPDs, with responses at any given IPD depending strongly on the IPDs the stimulus had just passed through. However, it was the temporal pattern of the response, rather than the motion cues in the IPM, that determined sensitivity to features such as motion depth, direction, and center locus. IPM restricted only to the center of the IPD responsive area, evoked lower discharge rates than when the stimulus either moved through the IPD responsive area from outside, or up and down its flanks. When the stimulus was moved through the response area first in one direction and then back in the other, and the same IPDs evoked different responses, the response to the motion away from the center of the IPD responsive area was always lower than the response to the motion toward the center. When the IPD was closer at which the direction of motion reversed was to the center, the response to the following motion was lower. In no case did we find any evidence for neurons that under all conditions preferred one direction of motion to the other. We conclude that responses of IC neurons to IPM stimuli depend not on the history of stimulation, per se, but on the history of their response to stimulation, irrespective of the specific motion cues that evoke those responses. These data are consistent with the involvement of an adaptation mechanism that resides at or above the level of binaural integration. We conclude that our data provide no evidence for specialized motion detection involving dynamic IPD cues in the auditory midbrain of the mammal.

Responses of low-frequency neurons in the inferior colliculus (IC) of anesthetized guinea pigs were studied with binaural beats to assess their mean best interaural phase (BP) to a range of stimulating frequencies. Phase plots (stimulating frequency vs BP) were produced, from which measures of characteristic delay (CD) and characteristic phase (CP) for each neuron were obtained. The CD provides an estimate of the difference in travel time from each ear to coincidence-detector neurons in the brainstem. The CP indicates the mechanism underpinning the coincidence detector responses. A linear phase plot indicates a single, constant delay between the coincidence-detector inputs from the two ears. In more than half (54 of 90) of the neurons, the phase plot was not linear. We hypothesized that neurons with nonlinear phase plots received convergent input from brainstem coincidence detectors with different CDs. Presentation of a second tone with a fixed, unfavorable delay suppressed the response of one input, linearizing the phase plot and revealing other inputs to be relatively simple coincidence detectors. For some neurons with highly complex phase plots, the suppressor tone altered BP values, but did not resolve the nature of the inputs. For neurons with linear phase plots, the suppressor tone either completely abolished their responses or reduced their discharge rate with no change in BP. By selectively suppressing inputs with a second tone, we are able to reveal the nature of underlying binaural inputs to IC neurons, confirming the hypothesis that the complex phase plots of many IC neurons are a result of convergence from simple brainstem coincidence detectors.

The relative importance of voice pitch and interaural difference cues in facilitating the recognition of both of two concurrently presented synthetic vowels was measured. The interaural difference cues used were an interaural time difference (400 μsec ITD), two magnitudes of interaural level difference (15 dB and infinite ILD), and a combination of ITD and ILD (400 μsec plus 15 dB). The results are analysed separately for those cases where both vowels are identical and those where they are different. When the two vowels are different, a voice pitch difference of one semitone is found to improve the percentage of correct reports of both vowels by 35.8% on average. However, the use of interaural difference cues results in an improvement of 11.5% on average when there is a voice pitch difference of one semitone, but only a non-significant 0.1% when there is no voice pitch difference. When the two vowels are identical, imposition of either a voice pitch difference or binaural difference reduces performance, in a subtractive manner. It is argued that the smaller size of the interaural difference effect is not due to a “ceiling effect” but is characteristic of the relative importance of the two kinds of cues in this type of experiment. The possibility that the improvement due to interaural difference cues may in fact be due to monaural processing is discussed. A control experiment is reported for the ITD condition, which suggests binaural processing does occur for this condition. However, it is not certain whether the improvement in the ILD condition is due to binaural processing or use of the improvement in signal-to-noise ratio for a single vowel at each ear.

Plaids composed of two orthogonal sine-wave gratings appeared to be of lower contrast than single gratings of the same Michelson luminance contrast. This effect for plaids was obtained at all spatial frequencies (1–16 c/deg) and contrast levels (2–32%). Contrast-matching data plotted as a function of the angle between plaid components (0–90 deg) and as a function of spatial frequency and standard contrast level were consistent with a model in which the response of each orientation-tuned spatial frequency channel is a threshold-corrected power function of contrast, and is followed by quadratic summation of responses across all channels. The best-fitting contrast-response exponent in the main experiment was 0.63. Analysis of several other data-sets suggested a slightly higher value, 0.80. The same model gave a good account of contrast-matching between simple and compound (two-component) one dimensional gratings, accounting in particular for the apparent increase in contrast summation exponent at low contrasts reported by Quick, Hamerly and Reichert [(1976) Vision Research, 16, 351–355). The model can, with one further assumption, account for the finding that contrast-matching between sine-wave and square-wave gratings depended only on the amplitude at the fundamental frequency. Comparison with contrast discrimination studies suggests that internal noise (variance of a channel's contrast-response) is not constant, but increases approximately in proportion to the mean response.

The ability of subjects to identify both components of a concurrent vowel pair was determined when vowel fundamental frequency difference and an interaural time difference (ITD) in one of the vowels were introduced. In agreement with earlier studies it was found that introduction of a fundamental frequency difference significantly improved detection performance (by 22% for a 1 semitone difference). Identification performance improved as the ITD was made larger but this improvement was small compared to the fundamental frequency difference effect (7% improvement between 0 and 400 μs).