Low-pass Cutoff Frequency Research Articles

Tracking in a ground-to-satellite optical link: effects due to lead-ahead and aperture mismatch, including temporal tracking response

Establishing a link between a ground station and a geosynchronous orbiting satellite can be aided greatly with the use of a beacon on the satellite. A tracker, or even an adaptive optics system, can use the beacon during communication or tracking activities to correct beam pointing for atmospheric turbulence and mount jitter effects. However, the pointing lead-ahead required to illuminate the moving object and an aperture mismatch between the tracking and the pointing apertures can limit the effectiveness of the correction, as the sensed tilt will not be the same as the tilt required for optimal transmission to the satellite. We have developed an analytical model that addresses the combined impact of these tracking issues in a ground-to-satellite optical link. We present these results for different tracker/pointer configurations. By setting the low-pass cutoff frequency of the tracking servo properly, the tracking errors can be minimized. The analysis considers geosynchronous Earth orbit satellites as well as low Earth orbit satellites.

Detection of B-FSK signals using schmidt trigger

Possibility to record B-FSK signals with use of stochastic resonance is studied. A Schmidt trigger is employed as bistable stochastic resonator. A digital model of the detector is proposed and quantitative results are obtained by means of simulations. Dependence of the probability of correctly detected bits on the added noise power and on the low-pass filter cutoff frequency is studied. Also the stability of the detector with respect to channel-introduced phase shifts is revealed.

The rule of 1 s for padding kinematic data prior to digital filtering: Influence of sampling and filter cutoff frequencies

The influence of signal sampling frequency and the low-pass digital filter cutoff frequency on the minimum number of padding points when applied to kinematic data are factors often absent in data processing descriptions. This investigation determined a relationship between the number of padding points and the ratio of filter cutoff to signal sampling frequencies ( f c/ f s). Two kinematic recordings were used which represented signals with high and low deterministic variation magnitudes at the signals’ beginning. Signal sampling rates (40–128 Hz) were generated at intervals of 1 Hz. Filter cutoff frequency was iterated from 2 to 10 Hz at 0.5 Hz intervals. Data extrapolation was performed using three different techniques (first order polynomial, third order polynomial, and data reflection). A maximum of 2 s of padding points were added to the beginning of each test signal which was then dual-pass filtered using a second order Butterworth filter. For each successive increase in the number of padding points, the filtered test signal was compared to a criterion signal and the root mean square difference (RMSD) over the first second was calculated. The number of padding points required to attain a constant RMSD was recorded as the minimum number of padding points needed for that ratio of filter cutoff to sampling frequency. As f c/ f s increased, the number of padding points decreased non-linearly. More padding points were required for the signal with higher deterministic variation at the beginning than the signal with lower deterministic variation. Additional padding points (beyond the determined minimum) did not further reduce the RMSD. The largest temporal extrapolation determined by the algorithm to produce a stable RMSD was 1 s. It is suggested that a minimum of 1 s of extraneous data be used when using a low-pass recursive digital filter to remove noise from kinematic data.

Spectral and temporal cues for speech recognition: Implications for auditory prostheses

Features of stimulation important for speech recognition in people with normal hearing and in people using implanted auditory prostheses include spectral information represented by place of stimulation along the tonotopic axis and temporal information represented in low-frequency envelopes of the signal. The relative contributions of these features to speech recognition and their interactions have been studied using vocoder-like simulations of cochlear implant speech processors presented to listeners with normal hearing. In these studies, spectral/place information was manipulated by varying the number of channels and the temporal-envelope information was manipulated by varying the lowpass cutoffs of the envelope extractors. Consonant and vowel recognition in quiet reached plateau at 8 and 12 channels and lowpass cutoff frequencies of 16 Hz and 4 Hz, respectively. Phoneme (especially vowel) recognition in noise required larger numbers of channels. Lexical tone recognition required larger numbers of channels and higher lowpass cutoff frequencies. There was a tradeoff between spectral/place and temporal-envelope requirements. Most current auditory prostheses seem to deliver adequate temporal-envelope information, but the number of effective channels is suboptimal, particularly for speech recognition in noise, lexical tone recognition, and music perception.

Spectral and temporal cues for phoneme recognition in noise

Cochlear implant users receive limited spectral and temporal information. Their speech recognition deteriorates dramatically in noise. The aim of the present study was to determine the relative contributions of spectral and temporal cues to speech recognition in noise. Spectral information was manipulated by varying the number of channels from 2 to 32 in a noise-excited vocoder. Temporal information was manipulated by varying the low-pass cutoff frequency of the envelope extractor from 1 to 512 Hz. Ten normal-hearing, native speakers of English participated in tests of phoneme recognition using vocoder processed consonants and vowels under three conditions (quiet, and +6 and 0 dB signal-to-noise ratios). The number of channels required for vowel-recognition performance to plateau increased from 12 in quiet to 16-24 in the two noise conditions. However, for consonant recognition, no further improvement in performance was evident when the number of channels was > or =12 in any of the three conditions. The contribution of temporal cues for phoneme recognition showed a similar pattern in both quiet and noise conditions. Similar to the quiet conditions, there was a trade-off between temporal and spectral cues for phoneme recognition in noise.

Quality of reporting masticatory muscle electromyography in 2004: a systematic review

This study evaluated the quality of reporting electromyography in studies evaluating the masticatory muscles published during 2004. Several electronic databases were searched. Abstracts and later articles were selected by a consensus from the five reviewers. An adaptation of the methodological checklist published by the International Society of Electrophysiology and Kinesiology (ISEK) was used. The following information regarding electrodes was reported on the 35 finally selected articles: location (94.3%), interelectrode distance (48.6%), and material (42.9%); detection and amplification: amplification type (54.3%), gain (25.7%), low high pass filters (60%) and cut-off frequencies (60%); electromyography (EMG) processing: sampling rate (74.2%), rectification (46.6%), root-mean square (RMS) (39.2%); number of bits and model of A/D card (17.1%); and normalization procedure (40%). Reasons for the poor reporting are discussed. Because of the general poor quality of reporting of the analysed studies, findings of studies using surface electromyography of masticatory muscles should be interpreted with caution.

A 1.2- V 30.4-dBm OIP3 Reconfigurable Analog Baseband Channel for UMTS/WLAN Transmitters

This paper presents a reconfigurable universal mobile telecommunication systems [(UMTS) and wireless local area network (WLAN) mode)] analog baseband transmitter channel composed of a current steering digital-analog converter (DAC), a transimpedance stage, and a low-pass reconstruction filter. The device operates from a single 1.2-V supply voltage while guaranteeing the high-linearity UMTS/WLAN standard requirements. It can be digitally programmed to process WLAN 802.11a/b/g and UMTS signals, by adjusting the DAC conversion frequency and the low-pass filter cutoff frequency. For the WLAN mode, the DAC operating frequency and the filter bandwidth are set to 100 and 11 MHz, respectively, while for the UMTS mode, they are equal to 50 and 2.11 MHz. The device is realized in a 1.2-V 0.13-mum standard CMOS technology. The die area occupation, equal to 0.9mm2, has been minimized by optimizing the component sharing for the two operation modes. The proposed circuit achieves a 30.4 dBm third-order output referred intermodulation intercept point (OIP3) for the WLAN mode, and a 31.5 dBm-OIP3 when configured for the UMTS mode, while the spurious-free dynamic range is 58 dB for WLAN mode, and 60 dB for UMTS mode. The power consumption is optimized according to the operation mode and is 19.44 mW in WLAN mode and 16.8 mW in UMTS mode

1.2-V Low-Power Multi-Mode DAC+Filter Blocks for Reconfigurable (WLAN/UMTS, WLAN/Bluetooth) Transmitters

Two versions of a baseband block composed by a 8-bit current-steering DAC and a fourth-order low-pass reconstruction filter are realized in a 0.13-mum CMOS technology to be embedded in multistandard wireless transmitters. In order to satisfy the specifications of WLAN IEEE 802.11a/b/g, UMTS, and Bluetooth standards, the proposed devices can be digitally programmed, adjusting the DAC conversion frequency and the low-pass filter cut-off frequency. For the WLAN case, the DAC operating frequency and the filter bandwidth are set to 100 MHz and 11 MHz, respectively, for the UMTS case, they are equal to 50 MHz and 2.5 MHz, and for the Bluetooth case, they are equal to 50 MHz and 1 MHz. The first device is reconfigurable between WLAN and UMTS, and the second one between WLAN and Bluetooth. The two fabricated devices operate from a single 1.2-V supply voltage and occupy a 0.8 mm 2 and 0.7 mm2 die area, respectively. The power consumption is optimized according to the operation mode and is 8 mW in WLAN mode, 8.4 mW in UMTS mode, and 5.4 mW in Bluetooth mode. For all the considered standards, the measured OIP3 is larger than 28 dBm, while the SFDR is 54 dB for WLAN, 61 dB for UMTS, and 63 dB for Bluetooth

The Accuracy of Determining Onset of Muscle Activity from the Surface Electromyogram

Amplitude cancellation is a characteristic of the surface EMG that reduces the estimated amplitude due to the overlap of opposing phases of motor unit potentials. Presumably, this attenuation occurs at the onset of a muscle contraction as well, and likely leads to errors in detecting the onset of motor unit activity. PURPOSE: To quantify the influence of amplitude cancellation on the accuracy of detecting the onset of motor unit activity in the rectified interference EMG. METHODS: A generic lower-limb muscle with 350 motor units in 20 random locations was simulated using the Fuglevand model (Fuglevand et al J Neurophysiol 70: 2470–2488, 1993). The force profile for the Achilles tendon during the stance phase of walking was used to validate the simulated net muscle force. The surface EMG signal was simulated by summing the motor unit potentials before (cancellation EMG) and after (no-cancellation EMG) they had been rectified. The two EMGs were compared across 20 trials at four levels of excitation: 10, 20, 40, and 100% of maximum. The influence of amplitude cancellation was examined by comparing the mean amplitude of the cancellation and no-cancellation EMGs. The experiment was repeated with median filtering (window sizes of 10, 20, 50, 80, and 100 ms, incremented by 1 ms) and low-pass filtering (cutoff frequency of 13, 20, 40, 60, and 100 Hz) applied to the cancellation EMG signal. Onset time was determined by visual discrimination as well as statistically as the last sample of the current window when the window average exceeded 3 SDs above the baseline mean. RESULTS: The no-cancellation EMG exhibited a greater rate of increase at the onset of motor unit activity for all four levels of excitation. Onset time differed by up to 12 ms between the cancellation and no-cancellation EMGs and was greatest for window sizes above 50 ms, and for low-pass filter cutoff frequencies below 60 Hz. For example, the comparisons between the cancellation and no-cancellation EMG at 40% excitation indicated that the error in detecting the onset of the contraction from the cancellation EMG was 8–26 ms, and that from the no-cancellation EMG was 7–17 ms. CONCLUSION: Amplitude cancellation did not influence the accuracy of visually detecting the onset time of a contraction in the surface EMG. However, errors were introduced when the EMG signal was processed with median and low-pass filtering techniques; the errors were most pronounced for larger window sizes (median filtering) and lower cutoff frequencies (low-pass filtering). Supported by NINDS NS043275.

Relative contributions of spectral and temporal cues for phoneme recognition

Cochlear implants provide users with limited spectral and temporal information. In this study, the amount of spectral and temporal information was systematically varied through simulations of cochlear implant processors using a noise-excited vocoder. Spectral information was controlled by varying the number of channels between 1 and 16, and temporal information was controlled by varying the lowpass cutoff frequencies of the envelope extractors from 1 to 512 Hz. Consonants and vowels processed using those conditions were presented to seven normal-hearing native-English-speaking listeners for identification. The results demonstrated that both spectral and temporal cues were important for consonant and vowel recognition with the spectral cues having a greater effect than the temporal cues for the ranges of numbers of channels and lowpass cutoff frequencies tested. The lowpass cutoff for asymptotic performance in consonant and vowel recognition was 16 and 4 Hz, respectively. The number of channels at which performance plateaued for consonants and vowels was 8 and 12, respectively. Within the above-mentioned ranges of lowpass cutoff frequency and number of channels, the temporal and spectral cues showed a tradeoff for phoneme recognition. Information transfer analyses showed different relative contributions of spectral and temporal cues in the perception of various phonetic/acoustic features.

Effects of Low-Pass Noise Masking on Auditory Event-Related Potentials to Speech

This study investigated the effects of decreased audibility in low-frequency spectral regions, produced by low-pass noise masking, on cortical event-related potentials (ERPs) to the speech sounds /ba/ and /da/. The speech sounds were presented to normal-hearing adults (N = 10) at 65- and 80-dB peak-to-peak equivalent SPL while they were engaged in an active condition (pressing a button to deviant sounds) and a passive condition (ignoring the stimuli and reading a book). Broadband masking noise was simultaneously presented at an intensity sufficient to mask the response to the 65-dB speech sounds and subsequently low-pass filtered. The conditions were quiet (no masking), low-pass noise cutoff frequencies of 250, 500, 1000, 2000, and 4000 Hz, and broadband noise. As the cutoff frequency of the low-pass noise masker was raised, ERP latencies increased and amplitudes decreased. The low-pass noise affected N1 differently than the other ERP or behavioral measures, particularly for responses to 80-dB speech stimuli. N1 showed a smaller decrease in amplitude and a smaller increase in latency compared with the other measures. Further, the cutoff frequency where changes first occurred was different for N1. For 80-dB stimuli, N1 amplitudes showed significant changes when the low-pass noise masker cutoff was raised to 4000 Hz. In contrast, d', MMN, N2, and P3 amplitudes did not change significantly until the low-pass noise masker was raised to 2000 Hz. N1 latencies showed significant changes when the low-pass noise masker was raised to 1000 Hz, whereas RT, MMN, N2, and P3 latencies did not change significantly until the low-pass noise masker was raised to 2000 Hz. No significant differences in response amplitudes were seen across the hemispheres (electrode sites C3M versus C4M) in quiet, or in masking noise. These results indicate that decreased audibility, resulting from the masking, affects N1 in a differential manner compared with MMN, N2, P3, and behavioral measures. N1 indexes the presence of audible stimulus energy, being present when speech sounds are audible, whether or not they are discriminable. MMN indexes stimulus discrimination at a pre-attentive level. It was present only when behavioral measures indicated the ability to differentiate the speech sounds. N2 and P3 also were present only when the speech sounds were behaviorally discriminated. N2 and P3 index stimulus discrimination at a conscious level. These cortical ERP in low-pass noise studies provide insight into the changes in brain processes and behavioral performance that occur when audibility is reduced, such as with low frequency hearing loss.

Experimental Assessment of the Non-Characteristic Harmonics Generated by Adjustable Speed Drives

Adjustable speed drives (ASD) are widely used in industry. A typical voltage-source ASD is comprised of a three-phase rectifier, a low-pass filter and an inverter stage, as shown in Fig. 1. This figure also shows, from left to right, the typical current waveforms at the a.c. input, the d.c. link, and the output. An important issue in this kind of apparatus is the harmonic content in the current drawn from the a.c. line, since the maximum amplitudes that can be tolerated are tightly regulated by standards.It is a common practice to analyse the harmonic generation considering isolated stages but, due to the complexity of the expression involved, the analysis is performed under several simplifying assumptions. Quite often, however, the analysis fails to predict accurately the harmonic content obtained in a particular application. Especially troublesome are the so-called non-characteristic harmonics (that is, harmonics whose frequency is not an integer multiple of the mains frequency). For example, harmonics related to the output waveform that propagate through the filter, to the mains. This paper is aimed at the experimental investigation of the non-characteristic harmonics generated by a typical voltage-fed ASD. The investigation is performed taking into account parameters such as the modulation index and the low-pass filter cutoff frequency. Three different cases are considered: a single ASD, two inverters sharing the same d.c. bus, and two ASD’s connected to the grid at the same point.The test performed show that, although the generation of non-characteristic harmonics does occur, in most cases the amplitudes involved are small. An exception to this occurs when the low-pass filter has not been properly designed, or the supply is unbalanced.

Spectral contributions to the benefit from spatial separation of speech and noise.

Speech recognition in noise improves when speech and noise sources are separated in space. This benefit has two components whose effects are strongest in different frequency regions: (1) interaural level differences (e.g., head shadow), which are largest at higher frequencies, and (2) interaural time differences, which have their greatest contribution at lower frequencies. Binaural interactions enhance the separation of signals from noise through the use of these interaural differences. Here, the benefit attributable to spatial separation was measured as a function of the low- and high-pass cutoff frequency of speech and noise. Listeners were younger adults with normal hearing, older adults with normal hearing, and older adults with hearing loss. Binaural thresholds for narrowband noises were measured in quiet and in a speech-shaped masker as a function of masker low-pass cutoff frequency. Speech levels corresponding to 50% correct recognition of sentences from the Hearing in Noise Test (HINT) were measured in a 65-dB SPL speech-shaped noise. Thresholds for narrowband noises and for speech were measured with two loudspeaker configurations: (1) signals and speech-shaped noise at 0 degrees azimuth (in front of the listener) and (2) signals at 0 degrees azimuth and speech-shaped noise at 90 degrees azimuth (at the listener's side). The criterion measure was spatial separation benefit, or the difference in thresholds for the two conditions. Benefit of spatial separation for unfiltered speech averaged 6.1 dB for younger listeners with normal hearing, 4.9 dB for older listeners with normal hearing, and 2.7 dB for older listeners with hearing loss. Benefit was differentially affected by low-pass and high-pass filtering, suggesting a trade-off of the contributions of higher frequency interaural level differences and lower frequency interaural timing cues. As expected, older listeners with hearing loss benefited little from the improved signal-to-noise ratios in the higher frequencies resulting from head shadow, but showed some benefit from lower frequency cues. Spatial benefit for older listeners with normal hearing was reduced relative to benefit for younger listeners. This result may be related to older listeners' elevated thresholds at frequencies above 6.0 kHz.

Features of stimulation affecting tonal-speech perception: implications for cochlear prostheses.

Tone languages differ from English in that the pitch pattern of a single-syllable word conveys lexical meaning. In the present study, dependence of tonal-speech perception on features of the stimulation was examined using an acoustic simulation of a CIS-type speech-processing strategy for cochlear prostheses. Contributions of spectral features of the speech signals were assessed by varying the number of filter bands, while contributions of temporal envelope features were assessed by varying the low-pass cutoff frequency used for extracting the amplitude envelopes. Ten normal-hearing native Mandarin Chinese speakers were tested. When the low-pass cutoff frequency was fixed at 512 Hz, consonant, vowel, and sentence recognition improved as a function of the number of channels and reached plateau at 4 to 6 channels. Subjective judgments of sound quality continued to improve as the number of channels increased to 12, the highest number tested. Tone recognition, i.e., recognition of the four Mandarin tone patterns, depended on both the number of channels and the low-pass cutoff frequency. The trade-off between the temporal and spectral cues for tone recognition indicates that temporal cues can compensate for diminished spectral cues for tone recognition and vice versa. An additional tone recognition experiment using syllables of equal duration showed a marked decrease in performance, indicating that duration cues contribute to tone recognition. A third experiment showed that recognition of processed FM patterns that mimic Mandarin tone patterns was poor when temporal envelope and duration cues were removed.

Interactive Effects of Low–Pass Filtering and Masking Noise on Word Recognition

A word recognition in noise paradigm was employed to examine temporal resolution in individuals with simulated hearing loss. Word recognition scores were obtained for low-pass filtered speech (i.e., cutoff frequencies of 1000,1250, and 1500 Hz) presented in continuous and interrupted noise at signal-to-noise ratios (SNRs) of -10, 0, and 10 dB. Performance improved with increasing SNR and low-pass frequency filter settings. Generally, word recognition performance was better in the interrupted noise condition than the continuous noise condition. This effect was greatest in the -10 dB SNR condition. Since the continuous/interrupted performance difference steadily declined as a function of low-pass filter cutoff frequency, these findings suggest that one factor leading to poorer speech recognition in individuals with high-frequency hearing impairment may be their dependence on low-frequency hearing channels that are inherently poorer than high-frequency channels for temporal resolution.

Integration of optimized low-pass filters in a bandpass filter for out-of-band improvement

In this paper, we discuss the control and suppression of spurious resonances commonly encountered with distributed bandpass filters. The basic idea consists of introducing low-pass structures within bandpass topologies. By adjusting low-pass filter cutoff frequencies, harmonic resonances are attenuated, while maintaining in-band performances. In addition, transmission losses may be reduced as the insertion technique leads to optimum designs in terms of overall size. The approach described here is based upon the combination of classical synthesis techniques and optimization procedures provided by conventional computer-aided-design tools. Experimental results are obtained in the case of /spl lambda//sub g//4 shunt-stub filters implemented in classical semilumped low-pass architectures. Then, we investigate alternative integration technologies and techniques to improve the out-of-band rejection over very wide operating bandwidth. To overcome some design limitations, i.e., workable characteristic impedance value, we propose and validate a multilayer solution. Finally, new low-pass topologies are suggested to improve low-pass filter band rejection.

A modular, low-cost, digital signal processor-based lock-in card for measuring optical attenuation

A modular, low-cost, digital signal processor (DSP)-based lock-in card is described for measuring optical attenuation. By transferring the lock-in operation from the analog domain to the digital domain, the nonlinearities gain, and offset errors and drifts are virtually eliminated. The dual phase lock-in operation has been implemented on the low-cost DSP56002 evaluation module (DSP56002EVM) of Motorola that is widely used in audio signal processing. This evaluation board contains a 24 bit DSP56002 DSP and a stereo CD-quality audio codec that makes the board ideal for implementing signal processing algorithms. Due to the maximum sampling rate of the codec embedded on the DSP56002EVM, the frequencies of the processed signals must be below 20 kHz. This specification is enough for the most common applications in the field of optics, where low or very low frequency (<1 kHz) references are frequent. The software algorithm implementing the lock-in amplifier can be particularized by the user on the basis of the needed performances. The effects of finite word length in the digital filter implementation are analyzed. This analysis reveals that a 24 bit word length is not enough to ensure the filter stability and the required frequency response. To overcome this problem, the double precision multiply mode must be used. When the DSP56002 enters this mode, double precision 48 bit by 48 bit multiplication can be performed. The lock-in performance has been tested. The measured amplitude variations of the reference sine signal are about 0.003%, which do not affect the signal measurement. The lock-in behaves like a band-pass filter centered on the reference frequency whose bandwidth is related to the low-pass filter cutoff frequency. The measured frequency response shows that the lock-in performs as theoretically predicted. The DSP56002EVM can be used as a lock-in for electrical signals in stand-alone operation. Besides, we have designed a card that interconnects to the DSP56002EVM and allows the ensemble to act as an optical attenuation detector that measures optical losses over 70 dB. This range is similar to that achievable by commercially available optical loss testers and makes it suitable for optical return loss measurements of all kinds of commercially available optical connectors.

Multi-level image quantization with object-dependent spectral control

The iterative Fourier transform algorithm (IFTA) allows quantization with control of the spectrum of the output image, and therefore offers the possibility to shape the quantization noise to conform to the low-pass properties of the visual system. The windowed IFTA (WIFTA) extends the IFTA by processing on local windows, and therefore allows an adaptation to local object dependent parameters. The extension of this algorithm from two-level to multi-level output offers additional free parameters. We discuss the possibilities and limitations of an adaptation of the low-pass cut-off frequency and the local histogram distribution to the local gray value.

Effect of stimulation rate on phoneme recognition by nucleus-22 cochlear implant listeners.

This study investigated the effect of pulsatile stimulation rate on medial vowel and consonant recognition in cochlear implant listeners. Experiment 1 measured phoneme recognition as a function of stimulation rate in six Nucleus-22 cochlear implant listeners using an experimental four-channel continuous interleaved sampler (CIS) speech processing strategy. Results showed that all stimulation rates from 150 to 500 pulses/s/electrode produced equally good performance, while stimulation rates lower than 150 pulses/s/electrode produced significantly poorer performance. Experiment 2 measured phoneme recognition by implant listeners and normal-hearing listeners as a function of the low-pass cutoff frequency for envelope information. Results from both acoustic and electric hearing showed no significant difference in performance for all cutoff frequencies higher than 20 Hz. Both vowel and consonant scores dropped significantly when the cutoff frequency was reduced from 20 Hz to 2 Hz. The results of these two experiments suggest that temporal envelope information can be conveyed by relatively low stimulation rates. The pattern of results for both electrical and acoustic hearing is consistent with a simple model of temporal integration with an equivalent rectangular duration (ERD) of the temporal integrator of about 7 ms.

Auditory evoked responses to rhythmic sound pulses in dolphins.

The ability of auditory evoked potentials to follow sound pulse (click or pip) rate was studied in bottlenosed dolphins. Sound pulses were presented in 20-ms rhythmic trains separated by 80-ms pauses. Rhythmic click or pip trains evoked a quasi-sustained response consisting of a sequence of auditory brainstem responses. This was designated as the rate-following response. Rate following response peak-to-peak amplitude dependence on sound pulse rate was almost flat up to 200 s-1, then displayed a few peaks and valleys superimposed on a low-pass filtering function with a cut-off frequency of 1700 s-1 at a 0.1-amplitude level. Peaks and valleys of the function corresponded to the pattern of the single auditory brain stem response spectrum; the low-pass cut-off frequency was below the auditory brain stem response spectrum bandwidth. Rate-following response frequency composition (magnitudes of the fundamental and harmonics) corresponded to the auditory brain stem response frequency spectrum except for lower fundamental magnitudes at frequencies above 1700 Hz. These regularities were similar for both click and pip trains. The rate-following response to steady-state rhythmic stimulation was similar to the rate-following response evoked by short trains except for a slight amplitude decrease with the rate increase above 10 s-1. The latter effect is attributed to a long-term rate-dependent adaptation in conditions of the steady-state pulse stimulation.