KS3 Computer Science Key Terms & Vocabulary

Assignment

one-way: x = 5 puts 5 into x; 5 = x is a syntax error.

Python list

an ordered collection of values, written between square brackets, e.g. [10, 20, 30].

Logic error

when the program runs successfully but produces the wrong answer.

Logic errors

trickier than syntax errors because the code runs without complaining; only the result is wrong.

Multiplication operator

*. 2 * 3 returns 6.

Parameter

an input value listed in a function's definition; the function uses it to do its work.

Remainder (modulo) operator

%. 10 % 3 returns 1: the remainder after dividing 10 by 3.

Python string

enclosed in matching single or double quotes.

Syntax error

the code breaks the rules of the Python language; the program will not run at all until the error is fixed.

Variable

a named place in memory that stores a value the program can use later.

def add(a, b): return a + b, add(2, 3) returns 5.

After x = 7, the value of x is the integer 7, not zero, not the letter "x", not undefined.

xs.append(5) adds 5 to the end of the list xs.

In Python, you assign a value to a variable with the = operator, e.g. x = 5. The name goes on the left and the value on the right.

The general name for = storing a value in a variable is "assignment".

Python's bool type stores one of two values: True or False.

True and False are values of the Boolean (bool) type.

You call a function in Python by writing its name followed by parentheses. To call greet, write greet().

The four core Python types taught at KS3 are int, float, str and bool: integer, float, string, Boolean.

Every comparison in Python returns a Boolean: True or False.

'hello' + ' world' gives 'hello world' because the second string starts with a space character.

.count(sub) returns how many times sub appears. 'apple'.count('p') is 2.

In Python the keyword that begins a function definition is def.

In Python the division operator is the forward slash /. 10 / 4 returns 2.5.

def double(n): return n * 2, double(double(3)) is double(6) = 12.

== compares (returns True/False); = assigns (sets a variable). Confusing them is one of the commonest Python mistakes.

In Python the equality (test) operator is == (double equals).

= is assignment, not comparison; using = where you mean == is a common bug.

= is assignment (store a value); == is comparison (test equality and return True/False). They are different operators.

Given xs = [10, 20, 30], xs[0] is 10.

Showing 30 of 101. Practise the full Programming in Python set →

Bit

a single binary digit, either 0 or 1, and is the smallest unit of digital information.

Decimal 10

binary 1010 (8 + 2).

Almost all electronic circuits

designed around two clear voltage states: "on" and "off", because three or more states would be much harder to detect reliably.

Hex

commonly used in memory addresses and HTML/CSS colour codes (e.g. #FF00FF).

Kilobyte

1,000 bytes × 8 bits = 8,000 bits.

Kilobyte

much smaller than a gigabyte (about one millionth the size).

Hex

used because it is a much shorter, easier-to-read way to write long binary numbers.

0001 + 0001 = 0010 (1 + 1 = 2).

0011 + 0011 = 0110 (3 + 3 = 6).

0100 + 0100 = 1000 (4 + 4 = 8).

0101 + 0010 = 0111 (5 + 2 = 7).

1101 + 0011 = 10000 (13 + 3 = 16).

In 4 bits, 1111 + 0001 = 10000: the result is 5 bits, so 4-bit storage overflows.

Binary addition works right-to-left, carrying into the next column when needed, exactly like the long-form denary addition learned in primary school.

Like denary, binary addition aligns columns starting from the right: the least-significant bit first.

Binary 0110 = 4 + 2 = 6 in decimal.

Binary 10 = 2 + 0 = 2 in decimal.

Binary 100 = 4 + 0 + 0 = 4 in decimal.

Binary 1000 = 8 in decimal (the 1 sits in the 8-column; everything else is zero).

Binary 10000001 = 128 + 1 = 129 in decimal.

Binary 1001 = 8 + 1 = 9 in decimal (1-bits in the 8-column and the 1-column).

Binary 1010 = 8 + 2 = 10 in decimal.

Binary 11000000 = 128 + 64 = 192 in decimal.

Binary 1111 = 8 + 4 + 2 + 1 = 15 in decimal (the largest 4-bit value).

Binary 11111111 = 128 + 64 + 32 + 16 + 8 + 4 + 2 + 1 = 255 in decimal (the largest 8-bit value).

Binary digits (0/1) map directly onto Boolean logic (false/true), making it natural for computers to use AND/OR/NOT gates to process binary data.

Two clearly distinct voltage states are easy for hardware to detect even when electrical signals are noisy.

Binary addition only uses the digits 0 and 1: every "2" in a column becomes "10" (write 0, carry 1).

Binary uses only two digits: 0 and 1.

A bit (0 or 1) is the smallest unit of digital information.

Showing 30 of 75. Practise the full Binary and Number Systems set →

Web browser

a program used to view web pages (e.g. Chrome, Firefox, Edge, Safari).

Google Chrome

an example of a web browser.

Client

a computer (or program) that requests a service from a server.

Client

a device that requests data from a server.

Domain names

used because they're easier for humans to remember than IP addresses.

Ethernet

the most common wired LAN standard.

HTTP

the protocol used to transfer web pages from a web server to your browser.

Internet

a global network of networks: many networks connected together worldwide.

Internet

the largest WAN: in fact a global "network of networks".

Internet

not the same as the World Wide Web: the Web is just one of many services that run on the Internet.

Internet

a worldwide collection of interconnected networks.

IP address

a unique number that identifies a device on a network.

IP address

what uniquely identifies a device online.

IPv6

created because the Internet was running out of IPv4 addresses: IPv4 allows only about 4.3 billion (2³²) addresses.

Internet Service Provider (ISP)

a company that connects homes and businesses to the Internet.

LAN

usually owned and managed by one organisation (a school, an office, a home); a WAN typically links networks owned by many organisations.

Networks

usually classified by scale: a LAN (Local Area Network) covers a building or campus; a WAN (Wide Area Network) spans cities, countries, or the world.

Network

best described as devices connected so they can share data and resources.

Computer network

two or more computers (or devices) connected together so they can share data and resources.

Networks

used to share files, resources (printers, storage), and an internet connection.

Packet

a small chunk of data sent across a network as part of a larger message.

Peer-to-peer (P2P)

the alternative model where every device is an equal node, with no central server.

School computer network

typically a LAN.

Server

a computer that provides a service or data to clients.

Network switch

the device that connects multiple computers together in a LAN.

URL (Uniform Resource Locator)

the address of a web page.

URL

the address that locates a particular web page on the Web.

URL

a web-page address, not a piece of physical hardware.

World Wide Web

a system of linked pages (documents) accessed over the Internet.

Web

one of many services that run on the Internet (others include email and online gaming).

Showing 30 of 71. Practise the full Networks and the Internet set →

Attribution

naming the original creator when you reuse their work.

CEOP

the UK service for reporting online child sexual exploitation or uncomfortable communication from an adult.

Cookie

a small file in your browser that can track your activity across a website (and sometimes across sites).

Copyright

the legal right of the creator of an original work (song, book, software, art) to control how it is used.

Creative Commons

a set of share-friendly licences that let creators allow reuse under defined terms.

Creative Commons

the name of the licence family that lets others reuse work under defined conditions.

Cyber-bullying

bullying that uses digital technology (messages, social media, gaming).

Digital citizenship

behaving responsibly and respectfully online.

Long passwords

generally stronger than short ones because each added character makes guessing exponentially harder.

Malware

malicious software (any program designed to cause harm).

Password manager

the app that stores many strong passwords safely.

Password manager

the safest place to store passwords because the data is encrypted.

Phishing

when criminals send fake messages designed to trick you into revealing information or installing malware.

Online privacy

having control over your personal data and who can see it.

Reporting concerns

a positive action: it helps protect you and others.

Strong password

long and uses a mix of varied characters (letters, numbers, symbols, mixed case).

Trojan

malware disguised as legitimate / useful software, tricking the user into installing it.

Trojan

malware that pretends to be legitimate software (named after the wooden horse).

Computer virus

malware that spreads by attaching itself to other programs or files.

Virus

malware that replicates by attaching itself to other files / programs.

Worm

standalone malware that spreads itself across networks without attaching to a host program or file (unlike a virus, which needs a host).

Two-factor authentication (2FA) adds a second proof of identity beyond the password.

Anti-virus software helps by detecting and removing known malware.

Anti-virus software ITSELF needs to be regularly updated to recognise new threats.

Banks do not routinely ask for full passwords by email.

Blocking someone on a platform stops them from sending you further messages.

Cyber-bullying can have a serious effect on mental health, including anxiety and depression.

One example of cyber-bullying is sending hurtful messages to someone online.

Creative Commons material can usually be reused, but only by following the specific licence's terms (attribution, non-commercial, share-alike, etc.).

The UK law protecting copyrighted works is the Copyright, Designs and Patents Act 1988.

Showing 30 of 71. Practise the full Online Safety and Cyber Security set →

Bubble sort

easy to learn because it uses only "compare and swap", no extra data structures.

Bubble sort

slow on large data because it makes many passes and many swaps (O(n²) comparisons).

Sorted phone book

the classic case where binary search beats linear by a large margin.

Bubble sort

slow on a large list: it can need many passes, each comparing every adjacent pair, so the time taken grows quickly as the list gets longer.

Faster programs

nicer to use: efficiency matters for real-world software.

Worst case

the biggest / hardest input for an algorithm, the one that takes the most steps.

Flowchart

a diagram that shows the steps of an algorithm in order.

Linear search

not restricted to sorted lists. That's a common confusion with binary search.

Linear search

simple to understand and write, but slow on large lists.

Linear search

easy to write because it just walks through the items one at a time, in order, until it finds the target or reaches the end.

Pseudocode

an English-like description of the steps of an algorithm.

Pseudocode

written for humans to read, not for computers to run.

Pseudocode

not tied to any particular programming language.

Pseudocode

used to plan an algorithm in plain language before writing real code.

Trace table

a record of what an algorithm does. It never changes the algorithm itself.

Trace table

used to follow how an algorithm's variables change, step by step.

Trace tables

most helpful when debugging an algorithm that uses variables (especially loops).

Arrows between shapes show the direction of flow: they're how a reader follows the algorithm from start to stop.

Arrows (flowlines) join the shapes and show the direction the algorithm moves through them.

Binary search starts by checking the middle item of the list.

On a large sorted list, binary search is much faster than linear search.

At each step, binary search halves the remaining search space (eliminates the half that can't contain the target).

On a list of 1,000 items, binary search needs at most 10 comparisons (log₂1000 ≈ 10), versus up to 1,000 for linear search.

For a sorted list of 16 items, worst-case binary search makes 5 comparisons: log2(16) = 4 halvings, plus the comparison on the last remaining item (floor(log2 n) + 1).

Binary search cannot be used on an unsorted list: it relies on the ordering to decide which half to discard.

Binary search needs a sorted list to work.

Each step removes half of the remaining data. This is why binary scales so well to large data.

If the target is smaller than the middle item, binary search next looks in the lower half of the list.

Binary search halves the list at each step until it finds the target or the range is empty.

Bubble sort works on a list of any size.

Showing 30 of 70. Practise the full Algorithms set →

Clock speed

the rate at which the CPU's clock generates pulses (cycles) per second, measured in hertz; a higher clock speed means more cycles, so more work, per second.

Cache

the small, fast memory inside or very close to the CPU that stores recently-used data and instructions.

CPU

commonly described as the "brain" of the computer.

HDDs

cheaper per gigabyte than SSDs, which is why they are still used for bulk backup storage.

Headphones

a personal sound output device.

Input device

a piece of hardware that sends data into the computer.

Keyboard

an input device used for typing.

Microphone

an input device for sound (recording audio).

Microphone

not an output device: it is an input device.

Monitor

a visual output device: it displays images on a screen.

Monitor

not an input device: it is an output device.

Motherboard

the main printed circuit board inside a computer; it connects the major components together.

Mouse

an input device used for pointing.

Monitor

an output device; keyboards, mice and microphones are input devices.

Printer

the output device that makes a hard (paper) copy of digital data.

Printer

an output device: it produces printed (hard copy) output on paper.

Projector

an output device: it projects the screen image onto a larger surface.

RAM

primary memory, not secondary storage; RAM is volatile, secondary storage is non-volatile.

RAM

volatile: its contents are lost when the computer is switched off.

ROM

non-volatile: it keeps its contents with no power.

Scanner

an input device: it digitises paper documents and images.

Secondary storage

the long-term, non-volatile storage that keeps user files when the computer is switched off.

Secondary storage

non-volatile: it keeps data even when the power is off.

Speaker

an output device because it takes data from the computer and produces sound; a keyboard, scanner and microphone are input devices.

Speakers

an output device for sound.

SSDs

typically much faster than HDDs for reading and writing.

Touchscreen

both an input device (sensing touch) and an output device (showing the image).

Webcam

a video input device.

The arithmetic logic unit (ALU) is a part of the CPU.

CPU stands for Central Processing Unit.

Showing 30 of 60. Practise the full Computer Hardware set →

Abstraction

useful when writing programs because it lets you focus on the important parts of the problem.

Algorithm

a set of step-by-step instructions for solving a problem.

Decomposition

breaking a problem into smaller parts.

Decomposition

useful because smaller parts are easier to solve than the whole problem at once.

Algorithm design

planning the steps needed to solve a problem before writing real code.

Algorithm

not tied to one programming language: it can be written in pseudocode, Python, Scratch, English, or a flowchart.

Pattern recognition

spotting similar features between problems.

Pattern recognition

one of the four parts of computational thinking (with decomposition, abstraction, and algorithms).

Recipe

an everyday algorithm: it lists ordered steps to produce a result.

Ignoring details that don't matter is called abstraction.

Abstraction in computing means ignoring detail that is not needed to solve the problem.

Abstraction does not add extra detail to a problem: it strips detail away.

Abstraction helps you focus only on what is relevant to the problem.

Abstraction removes detail that isn't important.

Step-by-step instructions for solving a problem are called an algorithm.

A blank piece of paper is not an algorithm: there are no instructions.

When abstracting a chess game in code, physical details such as the wood grain of each piece would be ignored: they are not relevant to playing chess.

Clear, ordered steps make an algorithm easier for others to follow and check.

An algorithm must be a sequence of clear (unambiguous) steps.

Breaking a big problem into smaller parts is called decomposition.

Big, complex programs benefit most from decomposition: that's exactly when one person can't hold the whole solution in their head.

Decomposition makes large problems easier to tackle.

Decomposition does not mean making problems bigger: it does the opposite.

Decomposing a game like Pac-Man gives named parts: player movement, ghost movement, scoring, level design.

"Make a sandwich" decomposes into ordered subtasks: get bread → add filling → close → slice.

To decompose a big task, split it into smaller subtasks that you can solve one at a time.

Designing an algorithm before coding helps avoid mistakes that would be costly to fix later.

Learning to drive is an everyday abstraction: a learner is taught that the accelerator speeds the car up and the brake slows it down, without needing to understand how the engine or braking system works.

The first step in designing an algorithm is to understand the problem that needs to be solved.

A model keeps only the details the problem needs: in a flight simulator a plane's weight is essential because it affects handling, while the material covering the seats is irrelevant and is left out.

Showing 30 of 53. Practise the full Computational Thinking set →

Broadcast message

typically given a short descriptive name (e.g. start, game over).

Broadcasts

used to coordinate actions across sprites: one sprite signals, others react.

Events

used to respond to user actions and to messages from other parts of the program.

Scratch

a visual programming language using drag-and-drop blocks.

Sprite

an on-screen character or object in Scratch.

Sprite

an object on the stage that has its own costume(s) and code.

Sprites

useful because they give the characters for interactive programs (games, stories, animations).

Stage

the rectangular background area where sprites appear and act.

The shape in Scratch that snaps to other lines of code is called a block.

Each sprite's blocks live in that sprite's own code (scripts) area.

In Scratch, code is built by snapping blocks together: drag-and-drop, no typing required.

The broadcast () and wait block pauses the sending script until every receiving script has finished; the plain broadcast () block lets the sender continue immediately.

Scratch cannot limit where a broadcast goes: every script with a matching when I receive hat runs. Sending a message to just one sprite needs a workaround, such as a variable that each sprite checks.

A Scratch message sent to other sprites is called a broadcast.

A broadcast block sends a message to other sprites (and the stage).

When sprite A broadcasts start, every sprite with a when I receive start hat runs that script.

Broadcasts trigger other sprites to react to events in the program.

A broadcast can be received by any sprite listening for it, not only the sender.

Broadcasts let one sprite trigger code in another.

The Scratch variable block that changes by an amount is change [variable] by.

The change [score] by 1 block changes a variable's value by a given amount.

Code that runs in response to a user action is triggered by an event.

An event in Scratch is something that triggers code to run.

Events do not only run when the program ends: they run when their trigger happens (start, mid-program, on input).

Events let Scratch programs react to input (key presses, clicks, etc.).

The forever block runs its inner blocks endlessly.

A forever block does not stop after one run: it runs its contents endlessly.

The forever block repeats the code inside it endlessly.

When green flag clicked is a Scratch event.

When green flag clicked is the block that starts the program running.

Showing 30 of 52. Practise the full Visual Programming (Scratch) set →

Deeply nested code

harder to read and debug, so authors usually refactor (split out functions, use early returns) once nesting goes more than two or three deep.

Nested loops

ideal for working through 2D structures: rows × columns, e.g. printing a multiplication table or processing the pixels of an image.

Nested IF

an if statement placed inside the body of another if statement.

Nested loop

a loop placed inside another loop.

Selection

one of the three structured programming constructs (sequence, selection, iteration).

Sequence

the simplest of the three basic programming constructs; the other two are selection and iteration.

elif (else-if) lets you test another condition when the previous if was false; the first elif that matches runs and the rest are skipped.

A chain of `elif`s tests several alternatives without deep nesting; rewriting them as separate `if`s changes the meaning because the rest of the chain no longer skips.

The else block runs only when the if condition is False.

A for loop is the right choice when you know in advance how many times the code should repeat.

The counter (the loop variable) keeps track of how many times the loop has run so far.

A for loop repeats its body a known number of times, controlled by a counter.

FOR i in range(10) print(i) is a definite loop; WHILE x > 0 … and REPEAT … until quit are indefinite.

A for loop is a definite (count-controlled) loop: the number of iterations is fixed before the loop starts.

In for i in range(10):, the loop variable is i. range is a function, for is a keyword, and 10 is the upper bound: none of those is the loop variable.

A for loop does not run forever; it stops as soon as the counter has worked through every value in the range.

for i in range(2, 7): runs the loop body five times, with i = 2, 3, 4, 5, 6: the upper bound (7) is excluded.

for i in range(5): runs the loop body five times, with i = 0, 1, 2, 3, 4.

Use a for loop when the count is known; use a while loop when the number of repeats depends on a condition.

In a single if … else, exactly one branch runs: never both, never neither.

if is not a loop: it runs its block at most once per check, not many times.

An if statement runs the code inside it only when its condition evaluates to True.

In Python, indentation is what shows which block is inside which: deeper indent means deeper nesting.

for i in range(3): for j in range(2): print(i, j) prints 3 × 2 = 6 lines of output.

Two nested loops with counts m and n run the inner body m × n times in total.

Nesting does not make code faster; it usually does more work (e.g. m × n iterations instead of m or n).

Selection lets a program choose between possible actions depending on a condition.

Selection lets a program respond differently to different inputs (e.g. show "adult" or "child" based on age).

IF age >= 18 print 'adult' ELSE print 'child', with age = 12, the else branch runs and "child" is printed.

Every selection is driven by a condition that evaluates to True or False.

Showing 30 of 48. Practise the full Programming Basics: Sequence, Selection, Iteration set →

Booleans

used in IF statements: the condition is evaluated as either true or false.

Boolean operators

used inside if conditions and while loop conditions.

AND/OR/not

combined to handle more complex decisions in code.

Logic gate

a piece of hardware that performs a Boolean operation on its input(s).

Logic gates

the building blocks of CPUs and other digital circuits.

Truth tables

used to check every case of a logic expression in one go.

For A=1, B=0, the value of A AND (not B) is 1 AND 1 = 1 (true).

IF age >= 18 AND has_ticket runs the body only when both conditions are true.

When both inputs of AND are 1, the output is 1 (true).

An AND gate's output is true only when both inputs are true.

With x = 5, the expression (x > 0) and (x < 10) evaluates to true (both conditions hold).

An AND gate's output is 1 only when every single input is 1.

An AND gate outputs 1 only when all inputs are 1.

true AND (false OR true) = true AND true = true.

An AND gate outputs true only when both inputs are true.

true AND false = false.

true AND true = true (not false).

The data type whose value is true or false is called a Boolean.

Booleans appear in if conditions and while loop conditions.

NOT NOT true = true (double negation cancels).

A 4-input truth table has 16 rows (2⁴).

A basic gate has only two possible outputs: 0 or 1.

Logic gates do not output decimal numbers like 7 or 12: outputs are always a single binary value (0 or 1).

An inverter (NOT gate) flips a 1 to a 0 (and a 0 to a 1).

NOT (x == 5) is true when x does not equal 5.

NOT false = true.

A NOT gate inverts its single input.

A NOT gate has exactly one input: it inverts that input.

NOT (true OR false) = NOT true = false.

NOT true = false.

Showing 30 of 42. Practise the full Boolean Logic set →

Android

a mobile operating system made by Google.

Android

the most common mobile OS globally by user share.

Anti-virus scanner

an example of utility software.

Antivirus

the utility that scans files for malware.

Anti-virus

a type of utility software.

Editing a user's document

a typical application-software task.

Application software

software designed for user tasks like writing, browsing, and gaming.

Web browser

not utility software: it's application software.

Disk defragmenter

an example of utility software.

Device driver

system software that lets the OS talk to a specific hardware device.

Linux

a free, open-source operating system.

Linux

an open-source desktop OS (and also widely used on servers).

Operating system's job

to manage the hardware and the programs running on the computer.

Operating system

not application software: it's system software.

Managing memory between programs

a typical system-software task (done by the OS).

System software

software that runs the computer system itself (OS, utilities, drivers).

User interface (UI)

the part of the OS that lets users interact with the computer.

Microsoft Windows

a desktop operating system made by Microsoft.

Windows

a desktop operating system.

Microsoft Word

not an operating system: it's application software.

Android and iOS are both mobile operating systems.

Application software runs on top of the operating system.

A backup tool that copies files for safekeeping is utility software.

Backup utility software copies files for safekeeping in case of data loss.

A web browser (e.g. Chrome) is an example of application software.

Compression utility software reduces file size on disk.

Defragmenting speeds up a hard disk because each file's pieces are stored together (contiguously), so the read/write head moves less to read a file.

The OS manages files so the user can organise, store, and find their data.

iOS is Apple's mobile operating system, used on iPhones.

macOS is Apple's desktop operating system, used on Mac computers.

Showing 30 of 42. Practise the full Software and Operating Systems set →

Bit depth

the setting that determines how many colours per pixel are possible (same idea as colour depth).

Bitmap image

made of a rectangular grid of coloured pixels.

Colour depth

the number of bits used to encode the colour of each pixel.

Colour depth

not the width of the image in pixels: width is part of the image resolution.

Higher colour depth

more possible colours per pixel.

Higher resolution

a sharper image but a larger file size.

Pixel

the smallest controllable dot of an image.

Pixel

the smallest unit of a digital image.

Image resolution

the number of pixels in width × height.

1-bit colour depth gives 2 colours, typically black and white.

A 100 × 100 image at 1-bit colour depth = 100 × 100 × 1 = 10,000 bits.

A 100 × 100 image at 8-bit colour depth = 100 × 100 × 8 = 80,000 bits.

A 1920 × 1080 image has roughly 2 million (about 2.07 megapixel) pixels.

A 200 × 200 image at 4-bit colour depth = 200 × 200 × 4 = 160,000 bits.

24-bit colour depth ("True Color") gives about 16.7 million colours, with 8 bits each for red, green, blue.

A 50 × 50 image at 8-bit colour depth = 50 × 50 × 8 = 20,000 bits.

8-bit colour depth allows 2⁸ = 256 different colours per pixel (n bits per pixel give 2ⁿ colours, not n colours).

At 8 bits per pixel an image can represent 256 distinct colour values.

8-bit colour depth gives 256 colours per pixel.

Some older systems use the binary kilobyte (1 KB = 1,024 bytes); modern SI convention uses 1,000.

Bitmap file size (in bits) = width × height × colour depth.

To convert a bitmap file size from bits to bytes, divide by 8 (since 1 byte = 8 bits).

Colour depth is, by definition, the number of bits used per pixel.

File compression (JPEG, PNG) can reduce the stored file size below the raw bitmap calculation by spotting redundancies.

Colour depth DOES affect file size: the formula multiplies depth in directly.

Increasing colour depth grows the file because each pixel needs more bits.

Doubling colour depth roughly doubles the image file size (each pixel takes twice the bits).

Doubling resolution in both dimensions roughly quadruples the file size (×2 width × ×2 height = ×4 pixels).

Higher resolution gives more pixels and so more visible detail.

1 kilobyte (SI definition) = 1,000 bytes = 8,000 bits.

Showing 30 of 35. Practise the full Representing Images set →

Bit depth

the number of bits used per audio sample.

CD audio

sampled at about 44 kHz (44,100 samples per second).

Higher bit depth

each sample stores finer detail of the sound's amplitude.

Lower sample rates

sometimes used to get a smaller file at the cost of quality.

Sample rate

the number of samples taken per second.

Sampling

measuring a sound wave at regular intervals to record it as numbers.

The device that converts a sound wave into digital numbers is called an ADC (analog-to-digital converter).

CD-quality audio uses a sample rate of 44,100 Hz and a bit depth of 16 bits, which is enough to cover the human hearing range.

Sampling makes sound approximate because the continuous wave is reduced to a series of discrete points.

Doubling the sample rate doubles the file size in bits (everything else equal).

Higher sample rates AND higher bit depths both produce better-sounding audio, at the cost of a bigger file.

Increasing sample rate captures sound more accurately but makes the file bigger.

A higher sample rate gives a more accurate digital copy of the original sound, but a bigger file.

Human hearing ranges from roughly 20 Hz to 20,000 Hz, though the upper limit declines with age.

The standard unit for sample rate is the Hertz (Hz): samples per second.

To accurately capture a sound's highest frequency, the sample rate must be at least double that frequency (the Nyquist theorem).

Sample rate and bit depth are not the same thing: rate is "how often"; depth is "how detailed each measurement is."

Sampling turns continuous sound into discrete digital data.

Sampling does not make the sound louder: that's volume / amplitude, a separate property.

Measuring a sound at regular intervals is called sampling.

Sound file size (in bits) = sample rate × bit depth × duration (× channels for stereo).

Stereo sound stores two independent channels, so a stereo file is twice the size of the same recording in mono (the file-size formula multiplies by the number of channels).

Sound must be sampled because analog waves must become discrete numbers before a computer can store them.

A 2-second clip at 1,000 Hz sample rate and 8-bit depth = 1000 × 8 × 2 = 16,000 bits.

1 second at 8000 Hz sample rate and 16-bit depth = 8000 × 16 × 1 = 128,000 bits.

Lowercase 'a'

ASCII 97.

ASCII

a character encoding standard that maps characters to numbers.

ASCII

a standard 7-bit code that maps characters to numbers.

Unicode

created to cover characters from every world language.

Unicode

not limited to English: it covers Chinese, Arabic, Japanese, Hindi, Hebrew, Cyrillic, and many more.

Unicode

preferred for international websites because it can handle every common script and symbol.

Unicode

created to support characters from all the world's writing systems, not just basic Latin.

UTF-8

the most common Unicode encoding on the web (around 98% of web pages).

Standard 7-bit ASCII can represent 128 different characters (2⁷): letters, digits, common punctuation, and a few control codes.

Standard ASCII uses 7 bits per character, with 128 possible code values (0–127).

In ASCII, capital 'A' is encoded as 65.

Standard ASCII canNOT encode Chinese, Arabic, or Japanese: it only covers the basic Latin alphabet.

ASCII letter codes run in sequence: if A = 65, then B = 66, C = 67, and so on.

Capital and lowercase letters have different ASCII codes (A = 65, a = 97).

The first 128 Unicode characters are identical to ASCII, so plain ASCII text is also valid Unicode.

Unicode uses more bits per character on average because there are far more characters to distinguish.

Unicode covers most of the world's languages and scripts.

Unicode extends ASCII to cover most of the world's writing systems, plus emoji and mathematical symbols: modern systems use Unicode rather than plain ASCII.

Unicode includes emoji as well as letters.

Modern Unicode can represent over a million possible characters (codepoints).

Unicode supports far more characters than ASCII (Unicode has over 100,000 characters; ASCII has only 128).

Computers need character codes because internally they store everything as numbers (binary).