Year 11 AQA Computer Science: Speaking & Listening Technology Revision | AQA计算机科学：口语与听力技术备考专项

📚 Year 11 AQA Computer Science: Speaking & Listening Technology Revision | AQA计算机科学：口语与听力技术备考专项

Speaking and listening are essential human skills, and modern computer systems are increasingly capable of processing speech and audio. In the AQA GCSE Computer Science specification, understanding how sound is represented digitally, how speech recognition works, and how machines synthesise speech forms a key part of the Data Representation and Computer Systems topics. This article covers the core concepts, calculations, and exam techniques you need to confidently tackle any question about sound digitisation and speech technology.

口语和听力是人类的基本能力，而现代计算机系统正越来越多地处理语音和音频。在 AQA GCSE 计算机科学考试大纲中，理解声音的数字化表示方式、语音识别的工作原理以及机器如何合成语音，是数据表示与计算机系统主题的重要组成部分。本文将涵盖核心概念、计算方法以及应考技巧，帮助你自信应对有关声音数字化和语音技术的各类试题。

1. Why Sound Matters in Computing | 声音在计算中的重要性

Sound is a continuous (analogue) wave. Computers, being digital devices, can only store and process discrete binary values. To work with real‑world speech, music or any audio, a conversion from the analogue domain to digital form must take place. This process underpins every voice assistant, online call, music streaming service and accessibility tool.

声音是一种连续的（模拟）波形。计算机作为数字设备，只能存储和处理离散的二进制数值。要处理现实世界中的语音、音乐或任何音频，就必须完成从模拟域到数字形式的转换。这一过程是所有语音助手、在线通话、音乐流媒体服务和无障碍工具的基础。

In the AQA exam, you may be asked to explain how an analogue sound wave is converted into binary, calculate audio file sizes, or discuss the impact of sample rate and bit depth on quality. You also need to recognise how speech synthesis and recognition rely on this digital representation.

在 AQA 考试中，可能会要求你解释模拟声波如何转换为二进制、计算音频文件大小，或讨论采样率和位深度对音质的影响。你还需要认识到语音合成与识别如何依赖这种数字表示。

---

2. Analogue vs. Digital Signals | 模拟信号与数字信号

An analogue signal varies continuously over time – imagine the smooth curve of a sound wave. A digital signal, on the other hand, consists of discrete snapshots of that wave at specific moments. The conversion from analogue to digital is called digitisation, and it involves sampling and quantisation.

模拟信号随时间连续变化——想象一条平滑的声波曲线。而数字信号则是由在特定时刻对该波形进行离散快照组成的。从模拟到数字的转换称为数字化，它涉及采样和量化两个步骤。

When you speak into a microphone, the diaphragm vibrates, creating a continuous electrical voltage. The computer’s sound card measures this voltage at regular intervals (sampling) and assigns each measurement a binary number (quantisation). The result is a stream of binary data representing the original sound.

当你对着麦克风说话时，振膜振动产生连续变化的电压。计算机的声卡以固定间隔测量该电压（采样），并为每次测量分配一个二进制数字（量化）。结果就是表示原始声音的二进制数据流。

3. Sampling: Capturing the Wave | 采样：捕捉波形

The sample rate is the number of samples taken per second, measured in hertz (Hz). Common sample rates for audio are 44.1 kHz (CD quality) and 48 kHz. A higher sample rate captures more detail of the original wave, leading to a more accurate digital representation.

采样率是每秒采集的样本数，单位为赫兹 (Hz)。常见的音频采样率有 44.1 kHz（CD 质量）和 48 kHz。采样率越高，捕捉到的原始波形细节越多，数字表示的保真度也越高。

According to the Nyquist theorem, the sample rate must be at least twice the highest frequency you wish to capture. Human hearing ranges from about 20 Hz to 20 kHz, so a sample rate of 40 kHz or above is needed for full‑range audio. This is why 44.1 kHz was chosen for CDs.

根据奈奎斯特定理，采样率必须至少为你想要捕捉的最高频率的两倍。人耳的听觉范围大约在 20 Hz 到 20 kHz，因此全频段音频需要 40 kHz 或更高的采样率。这就是为什么 CD 选择了 44.1 kHz 的采样率。

Sample rate (Hz) = 2 × highest frequency (Hz)

---

4. Bit Depth and Quantisation | 位深度与量化

Bit depth determines how many bits are used to store each sample. It controls the number of possible amplitude levels. For example, an 8‑bit sample can represent 2⁸ = 256 different levels, while 16‑bit audio offers 2¹⁶ = 65,536 levels. Greater bit depth reduces quantisation error (the difference between the true analogue value and its digital approximation), yielding cleaner, more dynamic sound.

位深度决定了每个样本用多少比特来存储，它控制着振幅级别的数量。例如，8 位样本可以表示 2⁸ = 256 个不同级别，而 16 位音频则提供 2¹⁶ = 65,536 个级别。更大的位深度能减少量化误差（真实模拟值与数字近似值之间的差异），从而产生更干净、动态更丰富的声音。

CD audio uses 16‑bit depth; professional studio recordings often use 24‑bit. When the bit depth is too low, you hear noticeable ‘graininess’ or quantisation noise, particularly in quiet passages of speech or music.

CD 音频使用 16 位深度；专业录音室常使用 24 位。当位深度过低时，尤其是在语音或音乐的安静段落中，你会听到明显的“颗粒感”或量化噪声。

Number of levels = 2^{bit depth}

5. Calculating Audio File Size | 计算音频文件大小

Exam questions frequently ask you to estimate the size of an uncompressed audio file. The formula is straightforward:

考试中经常要求估算未压缩音频文件的大小。公式很简单：

File size (bits) = sample rate (Hz) × bit depth × number of channels × duration (s)

To convert to bytes, divide by 8; for kilobytes, divide by 8 × 1024; for megabytes, divide by 8 × 1024². Make sure you show all steps in your working.

转换为字节需除以 8；转换为千字节需除以 8 × 1024；转换为兆字节需除以 8 × 1024²。确保在解题过程中展示所有步骤。

Example: a 30‑second stereo recording at 44.1 kHz, 16‑bit per sample.
File size = 44,100 × 16 × 2 × 30 = 42,336,000 bits ≈ 5.04 MB. Always state units clearly.

示例：一段 30 秒立体声录音，采样率 44.1 kHz，每样本 16 位。文件大小 = 44,100 × 16 × 2 × 30 = 42,336,000 位 ≈ 5.04 MB。请始终清晰标明单位。

Parameter	Description	Common Values
Sample rate	Samples per second	44.1 kHz, 48 kHz, 96 kHz
Bit depth	Bits per sample	8, 16, 24
Channels	Mono (1) or stereo (2)	1, 2

---

6. Audio File Formats and Compression | 音频文件格式与压缩

Uncompressed formats like WAV and AIFF store raw PCM (Pulse Code Modulation) data. They preserve full quality but result in large files. Lossless compression (e.g. FLAC) reduces file size without discarding any audio information, making it ideal for archiving. Lossy compression (e.g. MP3, AAC) removes sounds that are psychoacoustically less noticeable, achieving much smaller sizes at the cost of some fidelity.

WAV 和 AIFF 等未压缩格式存储原始 PCM（脉冲编码调制）数据。它们保留了完整的音质，但文件体积很大。无损压缩（如 FLAC）在不丢弃任何音频信息的同时减小文件大小，非常适合存档。有损压缩（如 MP3、AAC）则会移除心理声学上不太明显的声音，在牺牲一定保真度的前提下获得小得多的文件体积。

Speech codecs used in telephony (e.g. G.711, Opus) are specifically tuned for voice, using low bit rates while maintaining intelligibility. Understanding compression is vital when explaining how voice assistants and phone calls work efficiently over networks.

电话通信中使用的语音编解码器（如 G.711、Opus）专门针对语音进行了优化，使用低比特率却保持可懂度。在解释语音助手和电话通话如何在网络上高效工作时，理解压缩至关重要。

7. Speech Recognition: From Audio to Text | 语音识别：从音频到文本

Speech recognition (or automatic speech recognition, ASR) converts spoken words into digital text. The process begins with the same digitisation steps: sampling and quantisation. The digitised audio is then analysed to extract acoustic features such as Mel‑frequency cepstral coefficients (MFCCs). These features are matched against acoustic and language models, often using hidden Markov models or deep neural networks, to determine the most probable word sequence.

语音识别（或自动语音识别，ASR）将说出的话语转换为数字文本。这一过程同样从采样与量化这两个数字化步骤开始。随后对数字化音频进行分析，提取诸如梅尔频率倒谱系数（MFCC）等声学特征。这些特征会与声学模型和语言模型进行匹配（常用隐马尔可夫模型或深度神经网络），以确定最可能的词语序列。

In an AQA context, you are not expected to detail neural networks, but you should explain that the system compares patterns in the digital audio to stored templates or models, and that accuracy is influenced by background noise, accent, and the size of the vocabulary.

在 AQA 考试情境下，不需要详述神经网络，但你应该解释系统会将数字音频中的模式与存储的模板或模型进行比较，并且识别的准确性会受到背景噪声、口音和词汇量大小的影响。

Common exam question: “Explain two reasons why a speech recognition system might misinterpret a word.” Answers could include homophones, noisy environments, or ambiguous pronunciation.

常见考题：“解释语音识别系统可能误判词语的两个原因。”答案可包括同音异义词、嘈杂环境或发音模糊。

---

8. Speech Synthesis: Turning Text into Speech | 语音合成：将文本转换为语音

Speech synthesis, often called text‑to‑speech (TTS), is the reverse process. It takes a string of text and produces an audio waveform. Early systems used concatenative synthesis – stitching together pre‑recorded fragments of human speech. Modern systems employ parametric or neural synthesis, generating waveforms from scratch based on linguistic rules and prosody models.

语音合成，通常称为文语转换（TTS），是反向的过程。它将一串文本转换为音频波形。早期系统使用拼接合成——将预先录制的人类语音片段拼接起来。现代系统则采用参数合成或神经合成，基于语言学规则和韵律模型从零生成波形。

In your exam, you could be asked to describe the steps a TTS system performs: text normalisation (expanding abbreviations, numbers), phonetic transcription (converting text to phonemes), and waveform generation. An understanding of how bit rate and sample rate affect the clarity of synthesised speech is also valuable.

在考试中，你可能会被要求描述 TTS 系统的执行步骤：文本规范化（展开缩写、数字等）、音标转换（将文本转为音素），以及波形生成。理解比特率和采样率如何影响合成语音的清晰度同样很有价值。

9. Applications and Ethical Considerations | 应用与伦理考量

Speech technology powers voice assistants (Siri, Alexa), real‑time translation, dictation software, accessibility tools for visually impaired users, and automated call centres. It brings convenience but also raises privacy issues – devices are constantly listening for wake words, and voice data may be stored on cloud servers.

语音技术驱动着语音助手（Siri、Alexa）、实时翻译、听写软件、视障用户无障碍工具以及自动呼叫中心。它带来了便利，但也引发了隐私问题——设备持续监听唤醒词，语音数据可能存储在云端服务器上。

From an ethical standpoint, you should be able to discuss concerns such as voice data being used without informed consent, potential bias in recognition accuracy across different accents or dialects, and the digital divide that excludes those without access to such technology.

从伦理角度，你应该能够讨论以下问题：在未获得知情同意的情况下使用语音数据、不同口音或方言间识别准确性的潜在偏见，以及因无法获取此类技术而产生的数字鸿沟。

• Privacy: Who has access to voice recordings? (隐私：谁能访问语音录音？)
• Bias: Are all user groups treated equally? (偏见：所有用户群体是否受到平等对待？)
• Accessibility: Does the technology help or hinder those with disabilities? (无障碍性：该技术是帮助还是妨碍了残障人士？)

---

10. Exam Technique for Sound and Speech Topics | 声音与语音专题的应考技巧

When answering questions on sound representation, always write down the formula first and substitute numbers clearly. Watch out for unit traps: sample rate given in kHz must be converted to Hz (×1000), and durations in minutes need to be expressed in seconds. Show all conversion steps to secure method marks even if the final answer is slightly off.

在回答声音表示类题目时，务必先写下公式，再清晰地代入数字。注意单位陷阱：以 kHz 给出的采样率需转换为 Hz（×1000），以分钟计的时长需换算为秒。展示所有转换步骤，即便最终答案略有偏差，也能确保拿到方法分。

For longer explanation questions about speech recognition or synthesis, structure your answer using a step‑by‑step approach. Use technical vocabulary correctly – sample rate, bit depth, quantisation, acoustic model, text normalisation. The examiner is looking for precise, logical sequencing.

对于关于语音识别或合成的长答题，采用分步式结构组织答案。正确使用技术词汇——采样率、位深度、量化、声学模型、文本规范化。考官看重的是精确且逻辑条理分明的表述。

Practice comparing lossy and lossless compression for speech recordings. Be prepared to justify why a particular format or sample rate is chosen for a given scenario, such as a phone line versus a high‑fidelity music platform.

练习对比语音录音的有损与无损压缩。准备好解释在特定场景下（例如电话线路与高保真音乐平台）为何选择特定的格式或采样率。

Key Exam Reminder: Bit rate = sample rate × bit depth × channels

---

Published by TutorHao | Computer Science Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）