| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 3.76 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A criação de um software aplicado a uma interface destinada a alunos de música e cantores,
nasce da necessidade de um utensílio que reúne diversas ferramentas em apoio à
afinação da voz. O software é destinado tanto a utilizadores sem experiência em teoria
musical, alunos do tema em específico quanto cantores experientes. O método de
aquisição da frequência fundamental da voz é a autocorrelação com center-clipping. Na
presente pesquisa, utilizou-se o software MatLab para o desenvolvimento do sistema. O
sinal de voz adquirido é comparado com um vetor de notas que varia de acordo com uma
das quatro aplicações do software: Ferramenta de apoio à determinação da classificação
vocal, treino da escala musical, treino de músicas pré-escritas e finalmente um escritor
de melodias com a possibilidade de reprodução da melodia composta, antes do treino. O
software foi avaliado por três categorias de alunos sendo: sem nenhuma, pouca e considerável
experiência com canto. O programa apresentou bons resultados avaliativos nos
aspectos geral e específico de cada ferramenta, e um índice evolutivo significante ao longo
dos treinos, com uma diminuição considerável do erro, incluso o grupo com significativa
experiência. Desta maneira, verificou-se a validade da ferramenta e a eficiência do método
do autocorrelação para a determinação da frequência fundamental, numa aplicação
prática de sua determinação em curto espaço de tempo.
The creation of a software applied to an interface intended for music students and singers, comes from the need of a new utensil version that reunites different tools to support voice tuning. The software is intended for both users with no experience in musical theory, theme students in specific and experienced singers. The acquisition method of the fundamental frequency is autocorrelation with center-clipping. In the present research, MATLAB was the choosen software for the development of the system. The acquired voice signal is compared to a vector of notes that vary according to the four applications of the software: Vocal classification support tool, musical scale training, prewritten music training and finally a composer of melodies with a possibility of chearing the created melody before training. The software was evaluated by three categories of students: without any, little and good experience with singing. The program presented good evaluative results in the various exercises, and an evolutionary index throughout the training, with a greater diminuition of the error, including the group with significant experience. In this way, the projects validates the efficiency of the autocorrelation method for the fundamental frequency determination, and proposes an option for a practical application of aquisition in short time.
The creation of a software applied to an interface intended for music students and singers, comes from the need of a new utensil version that reunites different tools to support voice tuning. The software is intended for both users with no experience in musical theory, theme students in specific and experienced singers. The acquisition method of the fundamental frequency is autocorrelation with center-clipping. In the present research, MATLAB was the choosen software for the development of the system. The acquired voice signal is compared to a vector of notes that vary according to the four applications of the software: Vocal classification support tool, musical scale training, prewritten music training and finally a composer of melodies with a possibility of chearing the created melody before training. The software was evaluated by three categories of students: without any, little and good experience with singing. The program presented good evaluative results in the various exercises, and an evolutionary index throughout the training, with a greater diminuition of the error, including the group with significant experience. In this way, the projects validates the efficiency of the autocorrelation method for the fundamental frequency determination, and proposes an option for a practical application of aquisition in short time.
Description
Keywords
Matlab Autocorrelação Frequência fundamental Pitch Afinação vocal