Claim Missing Document
Check
Articles

Found 3 Documents
Search
Journal : Paradigma

Teknik Adaptive pada Modulasi Delta Djadjat Sudaradjat
Paradigma Vol 21, No 2 (2019): Periode September 2019
Publisher : LPPM Universitas Bina Sarana Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (763.856 KB) | DOI: 10.31294/p.v21i2.5632

Abstract

DM (Delta Modulation) is one of the waveform coding methods that converts analog waveforms into digital waveforms, by encoding binary numbers 1 when analog waveforms rise, and encodes binary numbers 0 when analog waveforms drop, and if analog waveforms are flat will be encoded binary numbers 1 and 0 alternately. So that the row of binary numbers 1 and 0 will show the fluctuation of the analog signal form. Problems faced in Delta Modulation when the ups and downs of analog waveforms are too steep or too tilted so that analog waveform changes cannot be followed, there will be slope overload noise, which causes SNR (Signal to Noise Ratio) to shrink because it is unable follow the analog wave changes which are too high to produce a large noise. For this reason, Adaptive techniques are made, so that excess can be avoided, thereby reducing the Noise and SNR expected to be obtained. ADM (Adaptive Delta Modulation) reduces load tilt errors, and seeks to increase dynamic range. Adaptive algorithms adjust the step size (from various step sizes) to the signal power level and thus increase the dynamic range of the coding system. This paper discusses Delta Modulation techniques for increasing SNR with ADM techniques, as well as performance comparisons between DM and ADM.
Perbandingan Prediksi Linear Menggunakan Metoda Autokorelasi Dan Metoda Burg Djadjat Sudaradjat; Andi Rosano
Paradigma Vol 23, No 2 (2021): Periode September 2021
Publisher : LPPM Universitas Bina Sarana Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (715.14 KB) | DOI: 10.31294/p.v23i2.10589

Abstract

Abstrak  - Pada sistem pengenal ucapan manusia digunakan metoda pengenalan parameter-parameter yang dihasilkan oleh ucapan manusia untuk mengenalinya. Parameter-parameter ucapan yang dihasilkan bersifat unik yaitu tidak ada yang sama pada tiap orang seperti yang terjadi pada sidik jari manusia tidak ada yang sama, sehingga parameter-parameter rongga pengubah getaran ucapan manusia dapat digunakan untuk mengenali ucapan seseorang.  Beberapa metoda tiruan rongga pengubah getaran ucapan manusia diantaranya menggunakan metoda Autokorelasi dan metoda Burg. Pada penelitian ini akan diamati perbedaan kinerja metoda Autokorelasi dan metoda Burg dengan mengamati perbedaan spektrum pada tiap panjang frame yang berbeda-beda. Dari hasil pengukuran dapat dilihat bahwa pada perubahan panjang frame yang berbeda-beda pada metoda Burg tidak banyak perubahannya dibandingkan dengan metoda Autokorelasi. Pada metoda Autokorelasi terjadi perubahan spektrum ketika panjang frame makin pendek, sedangkan  pada metoda Burg hampir tidak mengalami perubahan yang banyak. Sedangkan kesalahan rekonstruksi sinyal sering terjadi pada frame yang panjang karena kemungkinan terdapat dua jenis sinyal voiced/unvoiced lebih besar dari pada menggunakan frame yang pendek. Dengan demikian untuk menghindari kesalahan rekonstruksi sinyal lebih baik menggunakan frame yang pendek, dan menggunakan metoda Burg dengan frame yang pendek menunjukkan kinerja spektral yang lebih baik dibandingkan dengan metoda Autokorelasi.
Perbandingan Algoritma Pengkodean Suara Manusia : Adaptive Delta Modulation and Linear Prediction Coding Djadjat Sudaradjat
Paradigma Vol 21, No 1 (2019): Periode Maret 2019
Publisher : LPPM Universitas Bina Sarana Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1548.784 KB) | DOI: 10.31294/p.v21i1.4641

Abstract

Coding algorithms human voice with ADM (Adaptive Delta Modulation) and LPC (Linear Prediction Coding) made to the encoding signals digitally produced has bit rate as low as possible, thus saving bandwidth transmission line if it will be distributed to the transmission medium, and save memory or computer storage if it will be stored as data storage.ADM technique different from the LPC, the difference is generally classified on how it is encoded. If the ADM system which is encoded representation of the waveform so often called "waveform coding", while the system is encoded LPC are the parameters of the source of the human voice, so often called the "parametric coding".Human voice coding algorithms to be compared are the results of my own research when conducting studies Bachelor of Engineering and Master of Engineering at ITB. The results of measurements taken from the ADM system materials Bachelor of Engineering Thesis entitled "Low Bit-Rate Delta Modulation", while the LPC system is taken from the material Engineering Master's thesis with the title "Voice Signal Processing with LPC Method".From the measurement results ADM system can generate the bit rate of 16 Kbps, while the LPC system generates the bit rate of 2.4 Kbps with the sound quality is almost the same output quality human voice that can be understood (intelligible speech).