--- title: "Source coders used in speech coding" subject: "Wireless Communication" chapter: 6 key: "semester_7/wireless_communication:speech-source-coders" source: https://www.bctengineer.com/q/semester_7/wireless_communication/speech-source-coders site: BCT Engineering --- # Source coders used in speech coding Wireless Communication · Chapter 6: Speech and Channel Coding - Asked in: 2074 Bhadra, 6 marks - Video: [Speech coding](https://www.youtube.com/watch?v=QYcU8cfwR2U) — PiEdX - Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14 - Page: https://www.bctengineer.com/q/semester_7/wireless_communication/speech-source-coders - Whole chapter: https://www.bctengineer.com/md/semester_7/wireless_communication/chapter-6 ## Question Describe the operation of any two source coders used in speech coding. ## Short answer ADPCM predicts each sample, quantizes only the prediction error with adaptive step size, and achieves 32 kbps. LPC models speech as excitation through an all-pole filter; it sends filter coefficients, gain, pitch and voiced/unvoiced flag, working at 2.4 kbps. ## Answer Speech **source coders** remove redundancy in speech to reduce the bit rate. Two widely used ones are described below. #### 1. Adaptive Differential PCM (ADPCM) – a waveform coder ADPCM uses the high correlation between speech samples. It predicts each sample from past samples and quantizes only the **prediction error**, using a quantizer whose step size adapts to the signal level. ```text s(n)->(+)-e(n)->[adaptive quantizer]-+-> e_q(n) out ^ - ^ | | [step adapt] v +-- s^(n) <--[predictor]<-----(+)<-- s^(n) s~(n) = s^(n) + e_q(n) (reconstructed) ``` - The predictor forms ŝ(n) from previous reconstructed samples. - Error e(n) = s(n) − ŝ(n) has a much smaller range than s(n), so fewer bits are needed. - Step size and predictor coefficients adapt to the changing speech level. - The decoder contains the same predictor, adding e_q(n) to ŝ(n). - ITU G.726 ADPCM gives toll quality at **32 kbps** (4 bits/sample) – half of 64 kbps PCM. Used in DECT. #### 2. Linear Predictive Coder (LPC) – a vocoder LPC models speech as an excitation passed through an all-pole vocal-tract filter, and sends only model parameters. ```text Tx: speech->[20 ms frames]->[LPC analysis a1..a10, G] ->[pitch, V/UV] -> encode Rx: [pulses | noise]->(×G)->[1/A(z) filter]->speech ``` - Each frame: about 10 predictor coefficients aₖ (from minimising prediction error, ŝ(n) = Σ aₖs(n−k)), gain G, pitch and voiced/unvoiced flag. - The receiver excites the filter 1/A(z) with a pulse train (voiced) or noise (unvoiced). - LPC-10 works at **2.4 kbps**; quality is synthetic. Hybrid versions (CELP, RPE-LTP) send a better excitation and are used in mobile systems. ## Check yourself The answer's key points as multiple-choice questions, with the correct option and why. 1. What does ADPCM quantize? A. The raw sample B. The prediction error with an adaptive step size C. The pitch D. The spectrum Answer: B. The prediction error has a smaller range than the signal, so fewer bits suffice. (Key point: ADPCM principle) 2. What bit rate does G.726 ADPCM use for toll quality? A. 2.4 kbps B. 13 kbps C. 64 kbps D. 32 kbps Answer: D. 4 bits/sample at 8 kHz = 32 kbps, half of PCM. (Key point: ADPCM rate) 3. How does the ADPCM decoder rebuild the sample? A. Using the same predictor and adding the received e_q(n) to the prediction B. By inverse FFT C. By noise generator D. By pitch pulses Answer: A. The decoder contains the same predictor and adds e_q(n) to s^(n). (Key point: ADPCM decoder) 4. In the LPC receiver, what is the synthesis filter? A. An FIR comb B. A band-pass bank C. An all-pole filter 1/A(z) driven by pulses or noise D. A differentiator Answer: C. Speech is produced by exciting the all-pole vocal-tract filter. (Key point: LPC receiver) 5. What is the bit rate of LPC-10? A. 2.4 kbps B. 32 kbps C. 16 kbps D. 64 kbps Answer: A. LPC-10 works at 2.4 kbps with synthetic quality. (Key point: LPC-10 rate) ## Related in Wireless Communication - [Need for speech coding; vocoder concept](https://www.bctengineer.com/question.md?key=semester_7%2Fwireless_communication%3Aspeech-coding-need-vocoder) — Wireless Communication, chapter 6, asked 2× - [Speech characteristics in coder design](https://www.bctengineer.com/question.md?key=semester_7%2Fwireless_communication%3Aspeech-characteristics-coder-design) — Wireless Communication, chapter 6, asked 3× - [Vocoder and predictive coders](https://www.bctengineer.com/question.md?key=semester_7%2Fwireless_communication%3Avocoder-predictive-coders) — Wireless Communication, chapter 6 - [Speech characteristics and linear predictive coder](https://www.bctengineer.com/question.md?key=semester_7%2Fwireless_communication%3Aspeech-lpc) — Wireless Communication, chapter 6, asked 5× - [Formant vocoder and speech characteristics](https://www.bctengineer.com/question.md?key=semester_7%2Fwireless_communication%3Aformant-vocoder-speech) — Wireless Communication, chapter 6 - [Waveform/voice coding and GSM codec](https://www.bctengineer.com/question.md?key=semester_7%2Fwireless_communication%3Awaveform-voice-coding-gsm-codec) — Wireless Communication, chapter 6 ## Related in other subjects - [Source coding vs channel coding](https://www.bctengineer.com/question.md?key=semester_5%2Fdata_communication%3Asource-vs-channel-coding) — Data Communication, chapter 8 - [Source coding; fixed vs variable length codes](https://www.bctengineer.com/question.md?key=semester_5%2Fdata_communication%3Asource-coding-fixed-variable) — Data Communication, chapter 8 - [Source coding aim; Huffman "Kun Mandir Ma Janchhau Yatri"](https://www.bctengineer.com/question.md?key=semester_6%2Fcommunication_systems%3Ahuffman-kun-mandir) — Communication Systems, chapter 6 - [Importance of source coding; Shannon-Fano vs fixed length (seven symbols)](https://www.bctengineer.com/question.md?key=semester_6%2Fcommunication_systems%3Asource-coding-sf-fixed-seven) — Communication Systems, chapter 6