Hasil Pencarian  ::  Simpan CSV :: Kembali

Abstrak :
ABSTRACT
Spatial Audio Object Coding SAOC merupakan standar pengkodean audio yang diluncurkan oleh Moving Picture Expert Group MPEG yang dapat melakukan kompresi dan koding audio berbasis objek dan telah diaplikasikan ke banyak bidang telekomunikasi, multimedia, dan hiburan. Salah satu kelemahan dari SAOC adalah kurang dapat diandalkan untuk menjamin kualitas audio yang baik untuk aplikasi pemisahan sumber audio. Hasil beberapa penelitian mengungkapkan bahwa ternyata SAOC memiliki struktur sistem yang mirip dengan algoritma pemisahan sumber suara yang bernama Informed Source Separation, Informed Source Separation ISS, yaitu algoritma sistem yang dapat memprediksi sinyal objek audio masukan untuk memisahkan objek audio dengan kualitas yang baik. Pada penelitian ini, telah dirancang algoritma pemisahan sumber audio di MATLAB yang diadopsi dari persamaan matematis standar SAOC serta ditambahkan filter Wienerpada algoritma tersebut. Algoritma yang diajukan diuji dengan memisahkan beberapa rekaman musik profesional dan kualitas audio rekonstruksinya akan dibandingkan dengan algoritma pemisahan sumber audio lain secara objektif. Hasil pengujian menunjukkan bahwa adanya peningkatan nilai Source-to-Distortion Ratio SDR dalam hasil rekonstruksi pemisahan sinyal dengan algoritma SAOC yang ditambahkan filter Wiener dari algoritma pemisahan sumber suara yang lain sebesar maksimum 10,42232904 dB untuk objek audio bass, 13,95175919 dB untuk drum, 17,73005926 dB untuk vokal, dan 8,266838319 dB untuk instrument lain.

---

ABSTRACT
Spatial Audio Object Coding SAOC is an audio coding standard launched by MPEG Moving Picture Expert Group MPEG that can perform audio compression and codingand has already applied to many areas of telecommunications, multimedia, and entertainment. One of the disadvantages of SAOC is low reliability in ensuring good audio quality for audio separation. The results of several studies have found that SAOC has the same structure with an audio source separation algorithm, namely Informed Source Separation, which is an algorithm that can predict audio object signals to separate audio objects while ensuring good output audio quality. In this research, an audio source separation algorithmwhich is adopted from SAOC standard mathematical equation andWiener filter addition has been designed. The proposed algorithm is tested by separating several professional music recordings and thereconstructed audios rsquo quality werecomparedwith other audio source separation algorithms. The results show that there is an increase in Source to Distortion Ratio SDR value of reconstructed audio object that is separated with SAOC algorithmwith addition of Wiener filter compared tootheraudiosource separation algorithmsbymaximum of 10.42232904 dB for bass, 13,95175919 dB for drums, 17,73005926 dB for vocals, and 8.266838319 dB for other instruments.

Abstrak :
This textbook presents an introduction to signal processing for audio applications. The authors approach posits that math is at the heart of audio processing and that it should not be simplified. He thus retains math as the core of signal processing and includes concepts of difference equations, convolution, and the Fourier Transform. Each of these is presented in a context where they make sense to the student and can readily be applied to build artifacts. Each chapter in the book builds on the previous ones, building a linear, coherent story. The book starts with a definition of sound and goes on to discuss digital audio signals, filters, The Fourier Transform, audio effects, spatial effects, audio equalizers, dynamic range control, and pitch estimation. The exercises in each chapter cover the application of the concepts to audio signals. The exercises are made specifically for Pure Data (Pd) although traditional software, such as MATLAB, can be used. The book is intended for students in media technology bachelor programs. The book is based on material the author developed teaching on the topic over a number of years. Presents a comprehensive introduction to audio processing for students in media technology and signal processing Builds a foundation for audio applications based on mathematical equations, presented in a way understandable to students without a math background Includes a full suite of classroom material including end of chapter exercises and companion Youtube video tutorials on the authors channel.