Orientador: Luiz Carlos Cagliari / Banca: Daniel Soares da Costa / Banca: Luciana Mercês Ribeiro Santos / Resumo: Várias vezes ouvimos falar que o português do Brasil é uma língua homogênea. Mas, um território de 8,5 milhões de quilômetros quadrados, com aproximadamente 200 milhões de pessoas, com grandes índices de analfabetismo e com uma pluralidade cultural - não poderia de modo algum apresentar uma língua homogênea. O presente estudo verifica como determinados falantes pronunciam os ditongos nasais -ões, -ães e -ãos, analisando os segmentos [ɐ̃], [õ], [ɪ̃] e [ʊ̃], que compõem estes ditongos nasais. O foco da pesquisa é através de uma investigação acústica. Verificadas as qualidades acústicas investigadas, a pesquisa faz uma análise comparativa entre as qualidades acústicas e articulatórias. A pesquisa utiliza o programa de análise acústica Praat. Com o auxílio do Praat e seguindo a Teoria Acústica de Produção de Fala de Fant (1960), foram feitas as segmentações apropriadas dos enunciados registrados e, para os segmentos relevantes, foi investigada a estrutura acústica dos formantes com as facilidades fornecidas pelo Praat. O objetivo da pesquisa está voltado para a forma como os falantes pronunciam os ditongos nasais. No estudo, foi levada em consideração a percepção do fator, já que surgiram variações inesperadas. Fonologicamente, a forma de plural dos ditongos nasais em estudo pode ser interpretada como /-awNS/. / Abstract: Brazilians usually hear that his language is quite homogeneous. However, a large territory of 8,5 millions square kilometers, and with 200 million inhabitants, with a large contingent of his population not alphabetized, plus a cultural plurality, could not by all means be linguistically homogeneous. The present study verifies how some speakers pronounce the nasal diphthongs -ões, -ães e -ãos, with a particular articulatory and acoustic analysis of the diphthongal segments [ɐ̃], [õ], [ɪ̃] e [ʊ̃]. The focus of the research is the acoustic investigation. Once detected the acoustic patterns the work makes a comparative analysis between the acoustic and the articulatory patterns. The Praat Program is use for the acoustic analysis. With the help of Praat and following Fant speech acoustic theory (1960), there has been made the appropriate segmentations of the recorded utterances and for the relevant segments it has been investigated the formant structure with the facilities provided by Praat. The aim of the research is orientated towards how the speakers pronounce the nasal diphthongs. In the study it has been taken into consideration the factor perception, since there emerged unexpected variations. Phonologically the plural form of Portuguese nasal diphthongs may be interpreted as /-awNS/ / Mestre

by Choy Chi Yan. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1999. / Includes bibliographical references (leaves 111-115). / Text in English; abstract also in Chinese. / by Choy Chi Yan. / ABSTRACT --- p.I / ACKNOWLEDGMENTS --- p.III / TABLE OF CONTENTS --- p.IV / LIST OF FIGURES --- p.VII / LIST OF TABLES --- p.VIII / Chapter 1. --- INTRODUCTION --- p.1 / Chapter 1.1 --- Speech Recognition --- p.1 / Chapter 1.2 --- Development of Speech Recognisers --- p.4 / Chapter 1.3 --- Speech Recognition for Chinese Language --- p.5 / Chapter 1.4 --- Objectives of the thesis --- p.6 / Chapter 1.5 --- Thesis Structure --- p.7 / Chapter 2. --- PHONOLOGICAL AND ACOUSTICAL PROPERTIES OF MANDARIN CHINESE --- p.9 / Chapter 2.1 --- Characteristics of Mandarin Chinese --- p.9 / Chapter 2.1.1 --- Syllabic Structures --- p.10 / Chapter 2.1.2 --- Lexical Tones --- p.11 / Chapter 2.2 --- Basic Phonetic Units for Mandarin Chinese --- p.14 / Chapter 2.2.1 --- Tonal Syllables and Base Syllables --- p.14 / Chapter 2.2.2 --- Initial-Finals --- p.14 / Chapter 2.2.3 --- Phones --- p.16 / Chapter 2.2.4 --- Preme-Core-Finals and Preme-Tonemes --- p.17 / Chapter 2.2.5 --- Summary-The phonological hierarchy of Mandarin Syllables --- p.19 / Chapter 3. --- HIDDEN MARKOV MODELS --- p.20 / Chapter 3.1 --- Introduction --- p.20 / Chapter 3.1.1 --- Speech Data --- p.20 / Chapter 3.1.2 --- Fundamental of HMMs --- p.21 / Chapter 3.2 --- Using Hidden Markov Models for Speech Recognition --- p.22 / Chapter 3.2.1 --- Likelihood of the state sequence of speech observations --- p.22 / Chapter 3.2.2 --- The Recognition Problem --- p.24 / Chapter 3.3 --- Output Probability Distributions --- p.25 / Chapter 3.4 --- Model Training --- p.26 / Chapter 3.4.1 --- State Sequence Estimation --- p.26 / Chapter 3.4.2 --- Gaussian Mixture Models --- p.29 / Chapter 3.4.3 --- Parameter Estimation --- p.30 / Chapter 3.5 --- Speech Recognition and Viterbi Decoding --- p.31 / Chapter 3.6 --- Summary --- p.32 / Chapter 4. --- LARGE VOCABULARY CONTINUOUS SPEECH RECOGNITION FOR MANDARIN CHINESE --- p.33 / Chapter 4.1 --- Introduction --- p.33 / Chapter 4.2 --- Large Vocabulary Mandarin Chinese Recognition System --- p.34 / Chapter 4.2.1 --- Overall Architecture for the Speech Recogniser --- p.34 / Chapter 4.2.2 --- Signal Representation and Features --- p.36 / Chapter 4.2.3 --- Subword Unit Models Based on HMMs --- p.39 / Chapter 4.2.4 --- Training of Subword Units --- p.42 / Chapter 4.2.5 --- Language Model (LM) --- p.43 / Chapter 4.2.6 --- "Transcriptions, Word Networks and Dictionaries for LVCSR System" --- p.44 / Chapter 4.2.7 --- Viterbi Decoding --- p.47 / Chapter 4.2.8 --- Performance Analysis --- p.48 / Chapter 4.3 --- Experiments --- p.48 / Chapter 4.3.1 --- Tasks --- p.48 / Chapter 4.3.2 --- Speech Database --- p.49 / Chapter 4.3.3 --- Baseline Experimental Results --- p.51 / Chapter 4.4 --- Context Dependency in Speech --- p.52 / Chapter 4.4.1 --- Introduction --- p.52 / Chapter 4.4.2 --- Context Dependent Phonetic Models --- p.53 / Chapter 4.4.3 --- Word Boundaries and Word network for context-dependent HMMs --- p.54 / Chapter 4.4.4 --- Recognition Results Using Cross-Syllable Context-Dependent Units --- p.56 / Chapter 4.5 --- Tree-Based Clustering --- p.58 / Chapter 4.5.1 --- Introduction --- p.58 / Chapter 4.5.2 --- Decision Tree Based Clustering --- p.59 / Chapter 4.5.3 --- The Question Sets --- p.61 / Chapter 4.5.4 --- Convergence Condition --- p.61 / Chapter 4.4.5 --- The Final Results --- p.63 / Chapter 4.6 --- Conclusions --- p.65 / Chapter 5. --- APPLICATION1 ISOLATED WORD RECOGNITION FOR MANDARIN CHINESE --- p.67 / Chapter 5.1 --- Introduction --- p.67 / Chapter 5.2 --- Isolated Word Recogniser --- p.68 / Chapter 5.2.1 --- System Description --- p.68 / Chapter 5.2.2 --- Experimental Results --- p.70 / Chapter 5.3 --- Discussions and Conclusions --- p.71 / Chapter 6. --- APPLICATION2 SUBWORD UNITS FOR A MANDARIN KEYWORD SPOTTING SYSTEM --- p.74 / Chapter 6.1 --- INTRODUCTION --- p.74 / Chapter 6.2 --- RECOGNITION SYSTEM DESCRIPTION --- p.75 / Chapter 6.2.1 --- Overall Architecture and Recognition Network for the keyword Spotters --- p.75 / Chapter 6.2.2 --- Signal Representation and Features --- p.76 / Chapter 6.2.3 --- Keyword Models --- p.76 / Chapter 6.2.4 --- Filler Models --- p.77 / Chapter 6.2.5 --- Language Modeling and Search --- p.78 / Chapter 6.3 --- EXPERIMENTS --- p.78 / Chapter 6.3.1 --- Tasks --- p.78 / Chapter 6.3.2 --- Speech Database --- p.79 / Chapter 6.3.3 --- Performance Measures --- p.80 / Chapter 6.3.4 --- Details of Different Word-spotters --- p.80 / Chapter 6.3.5 --- General Filler Models --- p.81 / Chapter 6.4 --- EXPERIMENTAL RESULTS --- p.83 / Chapter 6.5 --- CONCLUSIONS --- p.84 / Chapter 7. --- CONCLUSIONS --- p.87 / Chapter 7.1 --- Review of the Work --- p.87 / Chapter 7.1.1 --- Large Vocabulary Continuous Speech Recognition for Mandarin Chinese --- p.87 / Chapter 7.1.2 --- Isolated Word Recognition for a Stock Inquiry Application --- p.88 / Chapter 7.1.3 --- Keyword Spotting for Mandarin Chinese --- p.89 / Chapter 7.2 --- Suggestions for Further Work --- p.89 / Chapter 7.3 --- Conclusion --- p.91 / APPENDIX --- p.92 / BIBLIOGRAPHY --- p.111

Approximation of F0 contours in Cantonese speech is investigated. Multiple approximations are examined and evaluated. The modified speech utterances that carry the approximated contours at syllable, word and sentence levels are perceptually examined with reference to natural speech. It is found that linear approximation can adequately describe all perception-sensitive F0 variations in Cantonese speech. Each tone contour can be represented by one or two linear movements, and the transition between co-articulated tones can be represented by one linear movement. / F0 contours measured from human speech (observed contours) generally vary to a considerable extent. This research attempts to investigate perception-critical variations in these highly varying contours. In particular, F0 contours in Cantonese speech are concerned. Cantonese is a major Chinese dialect that is known of being rich in tones. Psychoacoustic findings suggest that human perception has limitations in perceiving pitch movements. This means that not all of the variations in the observed contours are perceivable. A major problem addressed in this study is to find the simplest acoustic representation of an observed F0 contour that is adequate to attain comparable perception with the natural speech. / F0 variation in speech is known to carry abundant information, both linguistic and paralinguistic. Its impact on speech communication is thus widely concerned. F0 variation in speech, being a major super-segmental acoustic feature, has received a lot of attention, particularly from the perspectives of production-acoustics and perception-acoustics. However, it is noted that perception-acoustic knowledge of F0 variation in association with speech naturalness is quite limited. This is especially the case in the studies of tonal languages, in which most efforts are made on acoustic cues related to tone identification. / The feasibility of using linear approximation greatly simplifies the way to understand and interpret F0 variations in speech processing, by the means of learning the properties of linear movements. Three steps of analysis are carried out on the generated linear approximations. The first one examines the movement slopes in the approximated F0 contours of isolated syllables, in comparison with the perceptual thresholds found in the psychoacoustic studies. The second analysis is performed over a set of linearly approximated F0 contours of polysyllabic Cantonese words. The determining attributes of these linear movements, i.e., movement slopes, movement heights and time locations of turning points are analyzed statistically. The last analysis concerns the evaluation of modified F0 contours. Objective evaluations are compared with perceptual evaluation. These analyses provide knowledge which can improve our understanding on how F0 variations are processed in speech path. / To explore the potentials oflinear approximation in research of speech prosody, a perception-oriented framework of automatic approximation is developed, so as to replace the manual process in the feasibility study. The framework aims to make the process of deriving approximations standardized, consistent and efficient. It is formulated based on the experiences from manual approximations and is also implemented with other perceptual findings. The initial test on polysyllabic words gives promising results. / Li, Yujia. / Adviser: Tan Lee. / Source: Dissertation Abstracts International, Volume: 73-04, Section: B, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (leaves 161-170). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [201-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. Ann Arbor, MI : ProQuest Information and Learning Company, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese; includes Chinese characters.

Thesis (M.S.)--Rochester Institute of Technology, 1990. / Includes bibliographical references (leaves 51-53).

Thesis (Ph. D.)--University of Washington, 2004. / Vita. Includes bibliographical references (p. 393-404).

Thesis (Ph. D.)--Ohio State University, 2003. / Title from first page of PDF file. Document formatted into pages; contains xx, 351 p.; also includes graphics Includes bibliographical references (leaves 344-351). Available online via OhioLINK's ETD Center

Speaker accent influences the accuracy of automatic speech recognition (ASR) systems. Knowledge of accent based acoustic variations can therefore be used in the develop¬ment of more robust systems. This project investigates the differences between first language (L1) and second language (L2) English in South Africa with respect to vowels and diphthongs. The study is specifically aimed at L2 English speakers with a native African mother tongue, for instance speakers of isi-Zulu, isi-Xhosa, Tswana or South Sotho. The vowel systems of English and African languages, as described in the linguistic literature, are compared to predict the expected deviations of L2 South African English from L1. A number of vowels and diphthongs from L1 and L2 speakers are acoustically compared and the results are correlated with the linguistic predictions. The comparison is firstly made in formant space using the first three formants found using the Split Levinson algorithm. The L1 vowel centroids and diphthong trajectories in this three-dimensional space are then compared to their L2 counterparts using analysis of variance. The second analysis method is based on simple hidden Markov models (HMMs) using Mel-scaled cepstral features. Each HMM models a vowel or diphthong from one of the two speaker groups and analysis of variance is again used to compare the L1 and L2 HMMs. Significant differences are found in the vowel and diphthong qualities of the two language groups which supports the linguistically predicted effects such as vowel substitution, peripheralisation and changes in diphthong strength. The long-term goal of this project is to enable the adaptation of existing L1 English recognition systems to perform equally well on South African L2 English. / Dissertation (MEng (Computer Engineering))--University of Pretoria, 2005. / Electrical, Electronic and Computer Engineering / unrestricted

The theory proposed in this thesis, Phonological Pitch, concerns the representation and behavior of the tone feature. It is a formally simple phonological theory constrained by a set of explicit extragrammatical principles. Phonological Pitch contains two major grammatical mechanisms. First, tone is represented with a single multivalued feature (Pitch) whose value can range from 1 to n, where n is a language-specific number with no universal upper limit. Second, the Contiguity Hypothesis states that tone groups in rules must always form contiguous sets, though these groups can vary from rule to rule. Phonological Pitch can be so simple because the power of the grammatical theory is constrained with independently necessary extragrammatical factors. Specifically, limits on the number of tone levels arise from learnability and perceptual constraints, which can be precisely formalized, that also play a role in nonlinguistic domains. Similarly, the Contiguity Hypothesis is derived from psychoacoustic constraints on discriminating between acoustically similar pitches. Other perceptual and physiological constraints explain patterns in the typology of contour tones and in the interactions of tone with other features. The empirical support for Phonological Pitch includes the following. First, languages are attested with as many as five distinct tone levels, and the number of languages with n tone levels gradually decreases as n increases, rather than dropping off abruptly at some point. An analysis using learnability and perceptual constraints can explain this gradual drop better than a universal grammatical upper limit. Second, tone rules can transpose sets of tones up or down by a fixed interval, a fact which is easier to formalize with a single multivalued feature than with a set of binary features. Third, tone groups do not form universal natural classes nor groups with noncontiguous tones, as other tone theories predict. Fourth, tone interacts not only with laryngeal features like voicing, but also with nonlaryngeal features like vowel height, and both the existence and relative rarity of tone-vowel height interactions imply that understanding tone interactions requires reference to extragrammatical physiological factors.

Submitted by Filipe dos Santos (fsantos@pucsp.br) on 2018-08-01T11:33:52Z No. of bitstreams: 1 Nathália Dos Reis.pdf: 6098992 bytes, checksum: 8086672011bbdc448fbdfde6bb8719ca (MD5) / Made available in DSpace on 2018-08-01T11:33:52Z (GMT). No. of bitstreams: 1 Nathália Dos Reis.pdf: 6098992 bytes, checksum: 8086672011bbdc448fbdfde6bb8719ca (MD5) Previous issue date: 2018-06-15 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / This research integrates the field of the investigations of the alaryngeal speech with support of the field of Acoustic Phonetics. The aim of this study was to analyze, from an acoustic point of view, the quality of the tracheoesophageal speech and the characteristics of the stops consonants and the adjacent vowels produced by the APE. From an initial group of 22 subjects (17 total laryngectomized and 05 not submitted to treatments for laryngeal cancer), 15 comprised the groups studied as follows: ALA group (10 albedo speakers with PTE, distributed in groups: ALAa: five speakers, age ± 64 years and ALAb: five speakers, age ± 57 years) and LAR group (five laryngeal speakers, age ± 47 years). The data collection procedure (audio recordings) was performed at the Laryngectomized Outpatient Clinic of the Head and Neck Surgery Sector of the Medical School of Ribeirão Preto and at the Integrated Laboratory of Acoustic Analysis and Cognition of PUC-SP. The corpus of the research consisted of six words with the structure CVCV (Consonant - Vowel - Stop Consonant - Vowel), disyllables and paroxitones inserted in a vehicle sentence. In order to meet the objectives, two instruments of analysis were elaborated: a System of Acoustic Classification of the Quality of the Tracheoesophageal Speech and a script, applicable to the PRAAT software, for extracting acoustic measures of stop consonants, highlighting the information regarding voicing contrast (set of 44 measures related to duration, fundamental and formants frequency, and the intensity of adjacent consonants and vowels). The results were analyzed by means of descriptive statistics and multivariate analysis (Discriminant Analysis and Principal Component Analysis). The Quality Acoustic Classification System of Tracheoesophageal Speech revealed three relevant dimensions, in descending order of influence: measures characteristic of the sonority source (harmonics); the vocal tract resonances (formants) and the source of transient noise of stop consonants and adjacent vowels. These dimensions were the subdivision of the ALA speakers group (ALAa: acoustic quality of upper tracheoesophageal speech and ALAb: mean and lower limits). Speech samples from the ALAa and ALAb subgroups revealed a limitation of the maximum harmonic frequency, in relation to the LAR group, in addition to an increase in durations and voicing percentage in non-voiced segments, in a higher proportion in the ALAb subgroup. The influential variables referred to harmonic measurements and those related to the occlusion phase. For discriminant analysis of voicing contrast, considering pairs per point of articulation, the percentages of classification were 89.29% in bilabial sounds, 93.06% in alveolar sounds and 86.75% in velar sounds. The acoustic findings revealed a wide range of mobilizations developed by the alarygeal speakers in the pursuit of the goal of articulated and intelligible speech production / Esta pesquisa integra o campo das investigações da fala alaríngea com respaldo do campo da Fonética Acústica. Apresentou como objetivos analisar, do ponto de vista acústico, a qualidade da fala traqueoesofágica e as características das consoantes oclusivas, e das vogais adjacentes, produzidas por falantes alaríngeos com PTE. De um grupo inicial de 22 sujeitos (17 laringectomizados totais e 05 não submetidos a tratamentos para o câncer de laringe), 15 compuseram os grupos estudados da seguinte maneira: grupo ALA (10 falantes alaríngeos com PTE, distribuídos em grupos: ALAa: cinco falantes; idade ± 64 anos e ALAb: cinco falantes; idade ± 57 anos) e grupo LAR (cinco falantes laríngeos; idade ±47 anos). O procedimento de coleta dos dados (audiogravações) foi realizado no ambulatório de Laringectomizados do Setor de Cirurgia de Cabeça e Pescoço da Faculdade de Medicina de Ribeirão Preto e no Laboratório Integrado de Análise Acústica e Cognição da PUC-SP. O corpus da pesquisa consistiu de seis palavras com a estrutura CVCV (Consoante – Vogal – Consoante oclusiva- Vogal), dissílabas e paroxítonas inseridas em sentença veículo. Para atender aos objetivos, foram elaborados dois instrumentos de análise: um Sistema de Classificação Acústica da Qualidade da Fala Traqueoesofágica e um script, aplicável ao software PRAAT, para extração de medidas acústicas de consoantes oclusivas, com destaque para as informações relativas ao contraste de vozeamento (conjunto de 44 medidas relativas à duração, à frequência fundamental e de formantes, e à intensidade das consoantes e das vogais adjacentes). Os resultados foram analisados por meio de procedimentos de estatística descritiva e de análise multivariada (Analise Discriminante e Analise de Componentes Principais). O Sistema de Classificação Acústica da Qualidade da Fala Traqueoesofágica revelou três dimensões relevantes, em ordem decrescente de influência: medidas características da fonte de sonoridade (harmônicos); das ressonâncias do trato vocal (formantes) e da fonte de ruído transiente das consoantes oclusivas e de vogais adjacentes. Tais dimensões pautaram a subdivisão do grupo de falantes ALA (ALAa: qualidade acústica da fala traqueoesofágica nos limites superiores e ALAb: limites médios e inferiores). As amostras de fala dos subgrupos ALAa e ALAb revelaram limitação da frequência máxima de harmônicos, em relação ao grupo LAR, além de aumento de medidas duracionais e de percentual de vozeamento em segmentos não vozeados, em maior proporção no subgrupo ALAb. As variáveis influentes referiram-se a medidas de harmônicos e aquelas relativas à fase de oclusão. Para análise discriminante do contraste de vozeamento, considerando-se os pares por ponto de articulação, os percentuais de classificação foram 89,29% nos sons bilabiais, 93,06% nos sons alveolares e 86,75% nos sons velares. Os achados acústicos revelaram uma vasta gama de mobilizações desenvolvidas pelos falantes alaríngeos na busca do cumprimento da meta de produção de fala articulada e inteligível

Made available in DSpace on 2014-08-27T14:36:47Z (GMT). No. of bitstreams: 0 Previous issue date: 2013-08-26Bitstream added on 2014-08-27T15:57:14Z : No. of bitstreams: 1 000731266.pdf: 2249574 bytes, checksum: 40b26d7fb5eb25c7c6d095bb7962b473 (MD5) / O presente trabalho versa sobre a presença de contrastes encobertos na fala de crianças com transtorno fonológico. Na literatura, os contrastes encobertos são descritos como produções que, apesar de apresentarem resultados perceptivo-auditivos idênticos/semelhantes, revelam, a partir da análise acústica, diferenças sutis. De maneira mais específica, este estudo busca observar se há preferência, por parte das crianças com transtorno fonológico, pela manipulação de pistas acústicas que não são robustas para o Português Brasileiro, na tentativa de distinguir os fones fricativos. Para tanto, foram utilizadas cinco gravações em áudio, da fala de cinco crianças entre 4 e 5 anos com transtorno fonológico, que apresentavam as chamadas “substituições fônicas” envolvendo a classe de sons das fricativas. Os dados foram coletados utilizando-se o Instrumento para Avaliação de Fala para Análise Acústica – IAFAC, gravados em cabine acústica, socilitando a cada criança cinco repetições das 96 palavras que compõem o instrumento. Os dados foram editados e analisados com o uso do software PRAAT. Foi realizada uma transcrição fonética da primeira repetição (R1) de cada criança, por três juízes, e considerada a concordância de 66%. A partir desta transcrição, foi realizado o cálculo do grau de severidade do transtorno fonológico por meio do PCC-R. Em seguida, realizaram-se a análise fonológica contrastiva da produção das cinco crianças e a análise acústica de todas as “substituições” envolvendo a classe de sons das fricativas. Para a análise acústica, os seguintes parâmetros foram adotados: limite inferior do pico espectral, centróide, variância, assimetria, curtose e duração. Após a análise acústica, verificou-se a existência de contrastes encobertos nas produções tidas como homófonas auditivamente, representando um total de 54% do total das ... / The present study focuses on the presence of covert contrasts in the speech of children with phonological disorder. The covert contrasts are described in the literature as productions that, despite showing auditory perception results identical/similar reveal, from the acoustic analysis, subtle differences. More specifically, this study to observe whether there is a preference, on the part of children with phonological disorders, on the manipulation of acoustic cues that are not robust to Brazilian Portuguese in an attempt to distinguish the fricative phones. Five audio recordings of the speech of five children with speech disorder between 4 and 5 years old who presented the so-called phonic substitution involving the sound class of the fricatives were used. These data were collected using the Speech Assessment Instrument for Acoustic Analysis – SAIAA (IAFAC), recorded in a soundproof booth, requesting each child five repetitions out of the 96 words that make up the instrument. The data were edited and analyzed using the software PRAAT. A phonetic transcription of the first repetition (R1) of each child were performed by three judges and considered the agreement of 66%. From this transcription the degree of severity of phonological disorder was calculated through the PCC-R. Posteriorly, contrastive phonological analysis of the production of the five children was carried out and, finally, acoustic analysis of all the substitutions was performed involving the sound class of the fricatives. For the acoustic analysis the following parameters were used: the lower limit of the spectral peak, centre of gravity, variance, skewness, kurtosis and duration. After acoustic analysis, we could verify the existence of covert contrast in the productions as homophones aurally taken by the judges, representing a total of 54% of total substitutions identified through impressionistic approach by the judges ...