Showing papers on "Noise published in 1979"

Enhancement of speech corrupted by acoustic noise

Abstract: This paper describes a method for enhancing speech corrupted by broadband noise. The method is based on the spectral noise subtraction method. The original method entails subtracting an estimate of the noise power spectrum from the speech power spectrum, setting negative differences to zero, recombining the new power spectrum with the original phase, and then reconstructing the time waveform. While this method reduces the broadband noise, it also usually introduces an annoying "musical noise". We have devised a method that eliminates this "musical noise" while further reducing the background noise. The method consists in subtracting an overestimate of the noise power spectrum, and preventing the resultant spectral components from going below a preset minimum level (spectral floor). The method can automatically adapt to a wide range of signal-to-noise ratios, as long as a reasonable estimate of the noise spectrum can be obtained. Extensive listening tests were performed to determine the quality and intelligibility of speech enhanced by our method. Listeners unanimously preferred the quality of the processed speech. Also, for an input signal-to-noise ratio of 5 dB, there was no loss of intelligibility associated with the enhancement technique.

Speech‐reception threshold for sentences as a function of age and noise level

Abstract: For 140 male subjects (20 per decade between the ages 20 and 89) and 72 female subjects (20 per decade between 60 and 89, and 12 for the age interval 90-96), the monaural speech-reception threshold (SRT) for sentences was investigated in quiet and at four noise levels (22.2, 37.5, 52.5, and 67.5 dBA noise with long-term average speech spectra). The median SRT as well as the quartiles are given as a function of age. The data are described in terms of a model published earlier [J. Acoust. Soc. Am. 63, 533-549 (1978)]. According to this model every hearing loss for speech (SHL) is interpreted as the sum of a loss class A (attenuation), characterized by a reduction of the levels of both speech signal and noise, and a loss class D (distortion), comparable with a decrease in signal-to-noise ratio. Both SHLA+D (hearing loss in quiet) and SHLD (hearing loss at high noise levels) increase progressively above the age of 50 (reaching typical values of 30 and 6 dB, respectively, at age 85). The spread of SHLD as a function of SHLA+D for the individual ears is so large (sigma = 2.7 dB) that subjects with the same hearing loss for speech in quiet may differ considerably in their ability to understand speech in noise. The data confirm that the hearing handicap of many elderly subjects manifests itself primarily in a noisy environment. Acceptable noise levels in rooms used by the aged must be 5 to 10 dB lower than those for normal-hearing subjects.

A spectral subtraction algorithm for suppression of acoustic noise in speech

Abstract: Spectral subtraction has been shown to be an effective approach for reducing ambient acoustic noise in order to improve the intelligibility and quality of digitally compressed speech. This paper presents a set of implementation specifications to improve algorithm performance and minimize algorithm computation and memory requirements. It is shown spectral subtraction can be implemented in terms of a nonstationary, multiplicative, frequency domain filter which changes with the time varying spectral characteristics of the speech. Using this filter a speech activity detector is defined and used to allow the algorithm to adapt automatically to changing ambient noise environments. Also the bandwidth information of this filter is used to further reduce the residual narrowband noise components which remain after spectral subtraction.

High resolution spectral estimation for noisy signals

Abstract: A spectral model containing poles and zeros is derived for the high resolution spectral estimation of data containing an auto-regressive (all-pole) signal, interference, and white noise A computationally efficient method for computing the spectrum is introduced Examples of some spectra calculated by this method are presented and compared with the autoregressive spectral estimator

Side-effects of interference with speech by noise.

Abstract: Pairs of housewives performed a proof reading task in conditions of loud and soft office noise. Loud noise produced slower proofreading and an increase in overall errors. In order to assess any after effects of proof reading conditions subjects completed a mood adjective check list and performed a Stroop test. These two items were completed in silence after half an hour of proof reading. Loud noise produced greater feelings of dysphoria, scepticism and deactivation and was accompanied by reports of lower efficiency, euphoria and activation than was soft noise. There were no effects of noise on the Stroop test.

Predicting annoyance from detectability of low‐level sounds

Abstract: The relationship between the predicted detectability and judged annoyance of 24 low‐level sounds heard in three noise backgrounds was investigated by an adaptive paired comparison procedure under free‐field listening cojditions. The predicted detectability of the set of sounds acco nted for almost 90= per−cent)= of the variance in judgments of anjoyance made in the most commonplace (falling spectru ) background noise environment. Conventional methods of predicting annoyance did not yield significantly higher correlations than a detectability‐based method in two other unusually shaped background noise environments.

Electronic gain and noise control for recording of analog information

Abstract: Audio signals with opposite relative phases are recorded on two tracks. A dual track head reproduces the signals from the two tracks and introduces another phase reversal of one signal relative to the other. Noise which was not present in the record circuit, but enters the system at the playback head, is out of phase in the signals which the playback head applies to the playback circuit. Combining the out-of-phase noise components of the two signals substantially cancels the noise.

A frequency domain noise cancelling preprocessor for narrowband speech communications systems

Abstract: Performance of narrowband speech communications systems, such as Linear Predictive Coding (LPC), is often severely degraded by the presence of ambient acoustic noise in the input speech signal. Spectral subtraction techniques show promise in improving the overall performance of LPC in acoustic noise environments, but typically present annoying musical tones at the output. A spectral subtraction technique is described, which includes a biased estimate of the noise, that does not present musical tones at the output. In addition, an automatic speech activity detector is described and used to adapt the noise estimate to changing noise environments.

Audio recognition unit

Abstract: PURPOSE:To reject error recognition due to noises generated during voice production, by eliminating ambient noises generated closely or being isolated from the voice production by detecting a voice period with a real voice level. CONSTITUTION:This unit consists of voice collecting microphone 10, voice level measurement part 20 for the calculation of the strength level of collected voice signals, ambient-noise collecting microphone 30, noise-level measurement part 40 for the calculation of the strength level of collected noise signals, voice detection part 50 which determines a voice period by detecting the voice level compensated at a noise level, and recognition processing part 60 which recognizes an input voice by analyzing and discriminating signals in the voice period. Then, mixing noises from microphone 10 are compensated at the noise level obtained by microphone 30 and measurement part 40 and the detection of the voice period is attained at a real voice level.

Temporal resolution in normal and hearing‐impaired listeners

Abstract: Temporal resolution, defined as the minimum detectable temporal gap (Δt ms) separating equal intensity noise bursts, was investigated with normal‐hearing and cochlear‐impaired listeners. In separate conditions, octave‐band noise signals of 400–800 Hz, 800–1600 Hz, and 2000–4000 Hz were presented against a background of band‐reject broad spectrum noise designed to eliminate the influence of spectral artifact in threshold measures. Gap resolution for both listener groups showed systematic improvement with increased signal frequency, though performance of hearing‐impaired subjects was significantly poorer than normal at each octave‐band region. These findings were evident for listener group comparisons made on the basis of both equivalent sound‐pressure‐level and sensation‐level signal intensities. Results are compared to previous gap resolution estimates reported for wide‐band stimuli, and are discussed in terms of peripheral filter mechanisms thought to influence resolution performance. Implications for temporal processing in the presence of cochlear damage are also considered. [Work supported by NIH Grant No. NS12045.]

Noise and Reading Performance in an Open Space School.

Abstract: Parents and educators have expressed fears that high noise levels in open space schools may interfere with academic achievement. This investigation compared the reading comprehension of sixty fourth grade students in an open space school during periods of quiet and periods of naturally occurring background noise. The reading tasks resembled normal class routine. Noise sessions averaged 13 dbA higher than quiet sessions. Despite this substantial difference, the noise had no significant effect on accuracy or reading speed.

BSR (wave V) and N1 latencies in response to acoustic stimuli with different bandwidths.

Abstract: Shortlatency averaged responses to 0.02–6 kHz broadband (BB) and 2 kHz narrow‐band (NB) filtered clicks were recorded from the ear canal and the vertex of five adults with normal audiograms. Response latencies to the signals presented in quiet and with masking noise were compared. For constant spectrum level, the low‐frequency limit of the masking noise was systematically increased to 6 kHz. The upper‐frequency limit was always 8 kHz. For signals at 40 dB HL, the latencies of wave V and N1 decreased as the lower limit of the noise was raised and nearly equaled the latency for the signals presented in quiet. For signals at 70 dB HL, the latency in quiet was much less than that observed for all noise conditions, i.e., frequencies above 8 kHz would be required to interfere with the signal‐dependent synchronous neural activity. Similar effects were measured for both wave V and N1, but the analysis of wave V–N1 latency intervals suggested a possible signal dependency.

γ-LMS and its use in a noise-compensating adaptive spectral analysis technique

Abstract: The application of an adaptive prediction error filter to autoregressive spectral analysis has been proposed by Lloyd Griffiths and is in use for the measurement of the frequencies of sinusoids It is well known that additive observation noise can bias these frequency estimates, thus limiting the useful applications of Griffiths technique to low noise environments. In the limit of white observation noise this paper develops a modified adaptive algorithm which incorporates a priori or measured noise power information so that the generated frequency estimates are unbiased. Unlike other bias removal techniques recently discussed, this technique can work with multiple sinusoids. The performance of this algorithm will be demonstrated through computer simulations and its convergence behavior will be discussed.

Reduction of nonstationary acoustic noise in speech using LMS adaptive noise cancelling

Abstract: Nonstationary acoustic noise with energy possibly equal to or greater than the speech is suppressed using a two microphone implementation of adaptive noise cancellation. The primary noise added to the speech is reduced by subtracting a filtered version of the second microphone reference noise. The reference noise filter is adaptively up dated using the Widrow-Hoff LMS algorithm [1]. The effectiveness of noise suppression depends directly on the ability of the filter to estimate the transfer function relating the primary and reference noise channels. A study of the filter length required to achieve a desired noise reduction level in a hard-walled room is presented. Results demonstrating noise reduction in excess 10dB in an environment with 0dB signal noise ratio are presented.

Noise generator for a polyphonic tone synthesizer

Abstract: In an electronic musical instrument apparatus is provided for producing a noise-like signal suitable for a variety of musical effect such as the imitation of percussive musical instruments. A noise master data set is created repetitively and independently of tone generation by computing a Fourier algorithm using random values for the Fourier coefficients. The noise master data set is transferred to a noise tone register whose output is sequentially and repetitively read and converted to an analog noise signal. Formant circuitry is used to vary the noise signal's spectrum in a time variant manner and the frequency of an assigned variable frequency clock can be used to vary the spectral bandwidth of the output noise signal.

Estimation of distances of recorded sounds presented through headphones.

Abstract: Intermittent sounds generated at 270° azimuth and from distances ranging from 2 to 10 feet were recorded on magnetic tape and played back to listeners via headphones. Loudness cues for relative distance were eliminated at the time of recording. Listeners were required to estimate the apparent distance of the recorded sounds when heard monoaurally and binaurally. Most subjects estimated the order of distances correctly. Distance estimations were as proficient when listening monaurally as when listening binaurally. Performance was more accurate for high-pass (>4.0 kHz) noise bursts than for low-pass (<1.0 kHz) noise bursts. In a second study, broadband noise bursts were recorded from azimuthal positions of 360°, 330°, 300° and 270° again at distances ranging from 2 to 10 feet. Estimations of the distances of the sounds, presented via headphones, were most proficient when azimuthal position of the original stimuli was 330°.

Laboratory study of annoyance to combined airplane and road-traffic noise

Abstract: The annoyance of noise, which consisted of both separate and combined airplane and road-traffic noises, was studied. The subjects judged each session as to how annoyed they were in the simulated living room laboratory environment and as to how annoyed they were if they heard the noise in their home during day, evening, and night periods. The airplane noises, for equal session levels were judged significantly more annoying than the road traffic noises for the separate sessions. For the combined sessions, an interaction was found between the airplane noise and traffic noise levels, which was not adequately assessed by the total energy concept. Significant differences were found between the projected home responses for the day, evening, and night periods.

Fourier-autoregressive spectral estimation

Abstract: The Fourier-Autoregressive spectral estimator allows one to perform autoregressive spectral estimation on a narrowband basis as opposed to the conventional broadband approach. This capability is particularly valuable when the desired time series is a narrowband process, which is embedded in broadband observation noise. To generate the narrowband time series, the original data is passed through a bank of bandpass filters which is conveniently implemented by an FFT. It is shown that for a sinusoidal process in white noise, a tradeoff may be effected between the narrowband signal-to-noise ratio, which is principally responsible for spectral resolution, and the useable data record length.

Investigation of the noise generated by "quiet" design bias-ply and radial truck tires with cross-bar tread patterns

Abstract: v

Permanent changes in psychophysical tuning curves following exposure to weak, steep‐sided noise

Abstract: Two of three listeners participating in an experiment involving the use of a high‐pass noise with a very steep skirt (about −400 dB/octave below about 3000 Hz) now yield forward‐masking tuning curves that are considerably less steep on the high‐frequency side than was true prior to exposure to the steep‐sided noise. The exposures were monaural and intermittent both within daily sessions and across an eight‐week period, the “multiplied” noise was weak (spectral level 35 dB SPL, overall level 65 dB SPL), and there is no evidence of threshold shift (PTS). The greatest detuning occurs in the ipsilateral ear at a signal frequency of 2400 Hz, and there is little or no effect at 1700 Hz. There are also contralateral effects, but the patterns are different—the primary change is an elevation in the master level necessary when the masker and signal are the same frequency. There was some initial recovery, but deficits persist three months after the last exposure. These preliminary results indicate that noises with steep skirts should be treated with caution until their damage potential is better understood. [Work supported by NINCDS.]

Audio analysis and synthesis unit

Abstract: PURPOSE:To output the audio having naturality audibly, by measuring the short time average power of audio signal mixing surrounding noise and that for surrounding noise, subtracting the average power of surrounding noise from the average power of audio signal, and determining the amplitude of the excited power supply from the subtracted power. CONSTITUTION:The audio signal from the audio signal source 205 including the noise from the noise sound source 206 is measured to measure the parameter including frequency spectrum and audio information, and the audio is synthesized from the parameter value. In such an audio analysis and synthesis unit, the input signal of the signal source 205 including noise and the noise source 206 is fed to the short time average power measuring units 211 and 215 provided with the analyser side 202, the short time average power is measured, and the measuement value of noise is subtracted from the measurement value of audio signal with the audio power 213 and it is fed to the coder 212. The output at the analyzer side 2 with this constitution is fed to the composite coder at the synthesis side 203 and the output of the composite coder 216 is fed as the exciting sound source information of the audio synthesis filter 223.

Loudness and the acoustic reflex: normal-hearing listeners.

Abstract: The relationships among most comfortable listening level (MCL), loudness discomfort level, and the acoustic reflex to speech were studied on normal-hearing listeners using earphones and sound field test conditions. Recorded sentence materials were presented monaurally in quiet and, in the sound field, in the presence of 55 dB SPL cafeteria noise. The results indicate the MCL in the quiet sound field at approximately 70 dB SPL with the acoustic reflex occurring at 16 dB higher intensity. The earphone MCL was 7 dB lower than in the sound field, a finding that may reflect a real reverse in usual earphone/sound field results or simply calibration factors particular to the speakers and test room used in this study. The AR to speech seems to occur at approximately equal intensities between the MCL and LDL tested in quiet. The MCL is elevated by noise whereas the acoustic reflex remains at a constant level, indicating that no absolute relationship exists between loudness and the AR.

Accuracy of speaker verification via orthogonal parameters for noisy speech

Abstract: An important problem in speaker verification systems arises, when the speech inputs are noise corrupted with signal to noise ratios in the range of 10 to 30 dB (noise assumed to be zero mean and white). This paper deals with the accuracy of speaker verification algorithms derived from an orthogonal parameter representation of speech. Initially, the investigations are directed to evaluate the sensitivity of orthogonal parameters to the level of noise in the speech signal. The accuracy of verification is then determined, using only those parameters that are least sensitive to additive noise. The influence of the order of the linear prediction model on verification is also studied. The verification algorithm is based on the distance measure used by Sambur. Finally thresholds are established to determine the proper choice of orthogonal parameters (used in distance computation) and the order of the linear prediction model for a given signal to noise ratio in the speech signal. The above study is then used to evaluate the accuracy of verification when speech inputs are obtained from a noisy telephone channel.

Squelch and message storage system and method

Abstract: A system and method for squelching noise and storing messages. Many receiving systems produce noise bursts at the end of the reception of a message and prior to the operation of the muting circuit. Such noise bursts are especially severe in frequency modulation systems using limiters. By delaying the audio signal until the squelch circuit's hang time is completed the noise bursts are muted. One application of this invention is in vehicles such as police cars and taxicabs. For such an application one of the essential parts of almost every message is the street address where the vehicle is needed. Since it is easy to confuse addresses, and thereby require a repeat of the message, a recording of at least the address portion of the message would be most useful. A storage system is provided in order to allow the repetition of messages.

Method for the improvement of gramophone records

Abstract: Gramophone records made of a vinyl chloride based resin are subjected to a treatment with low temperature plasma of a gas having no polymerizability in the plasma condition. A markedly improved anti-static effect is imparted to the surface of the gramophone record which also becomes highly resistant to wearing so that the signal/noise ratio in playing of the gramophone record is remarkably improved even after many times of repeated playing of the record.

Audio tone adjusting circuit

Abstract: PURPOSE:To eliminate the feeling of unpleasantness due to low sound or high sound at greater sound or noise, by providing the automatic audio tone adjusting circuit automatically attenuating the low and high band components in the frequency characteristics of audio signal when audio volume is increased.