|
|
Variable segmentation for sound signal processing
Garai, Szabolcs ; Míča, Ivan (referee) ; Balík, Miroslav (advisor)
This paper describes the methods used mainly in the filtration of audio signals -- noise reduction. It realizes segmentation with variable lenght of segment, constant overlap add and segmentation with variable lenght of segment and overlap add. These methods are then compared with comon methods of segmentation in dependance of the lenght of segment and used window function by the thresholding method. For this purpose it uses the database of audio records. In the first part it desrcibes the technique of audio signal processing with comon method of segmentation. According to this method it continues in design of signal processing by variable segmentation method, in which it is needed to modify the shape of window function, which influences and attributes are explained in next chapter. In practical part it describes the implemented methods in MATLAB programming language with each steps of testing. It continues with chart of enclosed files and the evaluation of the results of hearing tests.
|
|
|
Advanced method of signal time-stretching based on tonal and non-tonal segment detection
Vomáčka, Jan ; Trzos, Michal (referee) ; Balík, Miroslav (advisor)
This thesis deals with the method for time conversion of an audio signal based on a different approach to groups tonal and non-tonal segment signal implemented in Matlab. The first part deal with methods for musical signal processing based on time compression or expansion and describes the problems associated with the current way of implementing these methods. The following describes the methods of variable segmentation and tonal and non-tonal components detection in the audio recording. At the second part are these methods of variable segmentation, tonal component detection and time compression or expansion implemented in MATLAB. At the third part is performed testing on sound recordings and evaluation of results.
|
|
|
Advanced method of signal time-stretching based on tonal and non-tonal segment detection
Vomáčka, Jan ; Trzos, Michal (referee) ; Balík, Miroslav (advisor)
This thesis deals with the method for time conversion of an audio signal based on a different approach to groups tonal and non-tonal segment signal implemented in Matlab. The first part deal with methods for musical signal processing based on time compression or expansion and describes the problems associated with the current way of implementing these methods. The following describes the methods of variable segmentation and tonal and non-tonal components detection in the audio recording. At the second part are these methods of variable segmentation, tonal component detection and time compression or expansion implemented in MATLAB. At the third part is performed testing on sound recordings and evaluation of results.
|
|
|
Variable segmentation for sound signal processing
Garai, Szabolcs ; Míča, Ivan (referee) ; Balík, Miroslav (advisor)
This paper describes the methods used mainly in the filtration of audio signals -- noise reduction. It realizes segmentation with variable lenght of segment, constant overlap add and segmentation with variable lenght of segment and overlap add. These methods are then compared with comon methods of segmentation in dependance of the lenght of segment and used window function by the thresholding method. For this purpose it uses the database of audio records. In the first part it desrcibes the technique of audio signal processing with comon method of segmentation. According to this method it continues in design of signal processing by variable segmentation method, in which it is needed to modify the shape of window function, which influences and attributes are explained in next chapter. In practical part it describes the implemented methods in MATLAB programming language with each steps of testing. It continues with chart of enclosed files and the evaluation of the results of hearing tests.
|
guest ::
