Analysis and synthesis of oboe tones by the linear prediction method
J. Duane Dudley, William J. Strong
Abstract
Open-access reader
J. Duane Dudley, William J. Strong
Abstract
Open-access reader
The linear prediction method has recently been used with considerable success in the analysis and synthesis of speech sounds. The purpose of this paper is to describe its application to musical tones in general, and to the analysis and synthesis of oboe tones in particular. Preliminary indications are that five predictor coefficients are adequate for the synthesis of good oboe tones, and in some ways are better than 12. Adaptation of the analysis-synthesis algorithms and determination of the optimum number of predictor coefficients for various pitches and tone dynamics is discussed. Tones synthesized with various numbers of predictors coefficients are demonstrated.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The linear prediction method has recently been used with considerable success in the analysis and synthesis of speech sounds. The purpose of this paper is to describe its application to musical tones in general, and to the analysis and synthesis of oboe tones in particular. Preliminary indications are that five predictor coefficients are adequate for the synthesis of good oboe tones, and in some ways are better than 12. Adaptation of the analysis-synthesis algorithms and determination of the optimum number of predictor coefficients for various pitches and tone dynamics is discussed. Tones synthesized with various numbers of predictors coefficients are demonstrated.
Key concepts: Oboe, Musical tone, Tone (literature), Speech recognition, Computer science, Adaptation (eye), Acoustics, Mathematics