Skip to content
CampusPrep
EC-803 (C) · Speech Processing/Important Questions

Speech Processing (EC-803 (C)) - Important Questions

  1. Unit 514 Marks High Priority

    Give an overview of speech synthesis methods (concatenative and waveform methods). Explain the main blocks of a text-to-speech (TTS) system and discuss the role of prosody in achieving naturalness.

    Core topic: frequent unit topic (speech synthesis methods) — repeated in past papers and central to Unit 5.

  2. Unit 57 Marks High Priority

    Compare concatenative and waveform (parametric) speech synthesis methods. For concatenative systems describe unit types (e.g., diphone, syllable, word), unit selection criteria and concatenation smoothing techniques. For waveform/parametric systems describe the synthesis pipeline and trade-offs in naturalness and flexibility.

    Core derivation from Unit 5: detailed comparison and internal mechanisms of concatenative versus waveform synthesis.

  3. Unit 57 Marks Medium Priority

    Explain the PSOLA (Pitch Synchronous Overlap and Add) algorithm. Describe how PSOLA performs time-scale modification and pitch modification of speech, and outline its advantages and limitations in TTS.

    Important algorithmic question addressing a key synthesis technique used for pitch and duration modification in Unit 5.

  4. Unit 57 Marks Medium Priority

    Describe the source-filter model of speech production and explain its application in parametric speech synthesis. Discuss how excitation (source) and spectral envelope (filter) are parameterized and re-synthesized in a vocoder-based system.

    Covers the foundational source-filter model as used in parametric synthesis and vocoder design; core to Unit 5 understanding.

  5. Unit 57 Marks Medium Priority

    Write short notes on common vocoders used in speech synthesis (for example LPC vocoder, STRAIGHT, WORLD). Compare them in terms of synthesis quality, controllability of parameters and computational cost.

    Covers practical vocoder comparisons and their relevance to synthesis quality — often asked to test comparative knowledge in Unit 5.

  6. Unit 57 Marks Medium Priority

    Discuss objective and subjective evaluation methods for text-to-speech systems. Explain Mean Opinion Score (MOS), Comparative MOS (CMOS), and objective measures such as Mel-Cepstral Distortion (MCD) and intelligibility metrics.

    Evaluation methods for TTS systems are essential Unit 5 topics and are frequently examined in practical and theoretical questions.

  7. Unit 510 Marks Medium Priority

    Explain statistical parametric speech synthesis approaches (HMM-based synthesis) and contrast them with modern neural TTS methods. Describe the training and synthesis pipelines, and discuss advantages and disadvantages compared to concatenative synthesis.

    Covers modern statistical approaches to synthesis (HMM and neural TTS) which are current trends in Unit 5 and likely examination topics.

  8. Unit 57 Marks Medium Priority

    Discuss prosody modelling in TTS. Explain methods for predicting and generating fundamental frequency (F0), duration and energy contours from linguistic features, and describe how prosody affects perceived naturalness and intelligibility.

    Prosody modelling is a recurring theme in synthesis questions — tests ability to link linguistic analysis with acoustic generation.

Go to where you left off?

Quick Add to Notes

Save questions, your own notes and screenshots into notes filed by unit. It takes a free account.

Create free account

Have an account? Log in