A bit (ビット) is the smallest unit of information, and it represents only one of two values: 0 or 1.
Bits
Patterns that can be told apart
1
2
2
4
3
8
n
2 to the n
Each extra bit lets every existing arrangement be followed by either 0 or 1, so the number of patterns that can be told apart doubles. That is why n bits can represent 2 to the n combinations. When characters and the like are written as bit strings, the arrangement of 0s and 1s assigned to one character is called its code.
AnalogyA single light switch in a room is either on or off, but put three switches side by side and the on-off combinations grow to eight.
How to rememberBit is short for binary digit, so reading it as "one binary digit" reminds you that it has only two possible values.
Exam points
Examn bits (ビット) can tell apart 2 to the n patterns. Every time the number of bits goes up by one, the number of kinds you can represent doubles.
ExamIf codes (符号) are assigned so that no code matches the opening part of any other code, a bit string (ビット列) made of codes of different lengths can still be split from the start in only one way and turned back into the original characters.
A byte (バイト) is a unit of information that handles 8 bits together as one group.
Unit
Relation to the next unit down (SI prefixes)
1 byte
8 bits
1 kbyte
1,000 bytes
1 Mbyte
1,000 kbytes
1 Gbyte
1,000 Mbytes
1 Tbyte
1,000 Gbytes
One byte is 8 bits, so it can tell apart 2 to the 8th, or 256, patterns. The capacity of files and storage devices is written by putting a prefix such as k, M, G or T in front of the byte.
AnalogyIt is like counting eggs by the eight-egg carton instead of one by one: say how many cartons there are and you know at once how many eggs are inside.
How to rememberByte is said to be a coined word playing on bite, so you can remember it as one "mouthful" of data.
Exam points
ExamTo find how many patterns a number of bytes (バイト) can represent, multiply by 8 to convert to bits (ビット) and then take 2 to that power.
ExamSI prefixes (SI接頭語) grow by a factor of 1,000 (10 to the 3rd) at each step, from k to M, from M to G and from G to T.
Mix-upIn unit symbols, a lowercase b stands for bits (ビット) and an uppercase B for bytes (バイト), so the same number means amounts that differ by a factor of 8.
The amount of information (情報量) is how much information a piece of data holds, expressed as a number in units such as the number of bits.
The more states you need to tell apart, the more bits you need. To tell apart 100 cases, for example, 6 bits (64 patterns) fall short, while 7 bits (128 patterns) are enough.
AnalogyThink of how many yes-or-no questions you would need to guess the number someone has in mind; that count corresponds to the number of bits needed.
Exam points
ExamThe minimum number of bits (ビット) needed to represent N kinds is the smallest n for which 2 to the n is N or more.
ExamThe size of uncompressed text-only data is the number of characters times the bytes (バイト) per character. How many items fit in a given capacity is found by dividing the capacity, in matching units, by the size of one item.
Digitization (デジタル化) is replacing analog information that changes continuously with separate, discrete numbers.
Analog signal
Sampling
Quantization
Encoding
Digital data
The resulting bit string becomes data a computer can handle.
4 / 4
Digital information is recorded as a sequence of numbers, so it hardly degrades even when copied or sent again and again, and a computer can calculate with it and edit it. Sound and similar information goes through sampling and quantization, and finally encoding, which writes each level number as a sequence of 0s and 1s, to become digital data.
AnalogyIt is like rebuilding a gentle slope as a staircase: once you record the position and height of each step as numbers, you can build the same shape again as many times as you like.
How to rememberDigit originally meant a finger, so picture counting on your fingers one by one: digitization turns information into values you can count.
Exam points
ExamA music CD records a sequence of values made by turning sound into numbers at fixed intervals, so it counts as a digital (デジタル) recording medium.
Sampling (標本化) is the step of taking, once every fixed time interval, the value that a signal changing over time has at that moment.
The time between readings is called the sampling period, and the number of readings taken per second is called the sampling frequency. A shorter period means correspondingly more values to record.
AnalogyIt is like taking burst photos of a runner at regular intervals: the shorter the gap between shots, the more of the motion in between you keep.
How to rememberThe kanji 標本 means a specimen taken from a larger whole, which links to picking sample points out of a continuous wave.
Exam points
ExamThe shorter the sampling period (標本化周期), the more closely spaced the points taken, and the closer the restored waveform comes to the original.
Quantization (量子化) is the step that fits each value taken by sampling to the closest of a set of levels decided in advance.
The number of levels depends on how many bits are used to represent one sample, and this number of bits is called the quantization bit depth (also called the encoding bit count). The gap from the original value that arises when rounding to a level is called the quantization error, and it gets smaller as the levels get finer.
AnalogyIt is like the difference between a bathroom scale that shows 0.1 kg steps and one that shows 1 kg steps: the finer the steps, the closer the reading is to the true value.
How to rememberThe kanji 量子 is the physics word for a quantum, the smallest step-like amount, which links to placing values onto separate marks on a scale.
Exam points
ExamUsing more bits (ビット) per sample increases the number of levels, so the size of each value can be recorded more finely. The more levels there are, the closer the restored waveform comes to the original.
Mix-upSampling (標本化) divides things along the time axis, and quantization (量子化) divides them along the axis of value size; quantization comes after sampling.
A/D conversion (A/D変換) is the processing that turns an analog signal into a digital signal.
Sound picked up by a microphone (analog)
A/D conversion
Processed by a computer (digital)
D/A conversion
Sound from a speaker (analog)
In A/D conversion, the input signal is sampled and quantized, and the resulting values are encoded and output as a bit string. The circuit or component that does this is called an A/D converter.
AnalogyIt is like reading a changing air temperature on a thermometer once a minute and writing down the nearest mark on the scale in a notebook.
How to rememberA stands for Analog and D for Digital, so read the slash from left to right as "from A to D" to get the direction of the conversion.
Exam points
ExamThe bit string produced by A/D conversion (A/D変換) is made of the level numbers found by quantization (量子化) at each sampling (標本化) moment, encoded (符号化) and lined up in order.
Mix-upTurning a digital signal back into an analog signal is D/A conversion (D/A変換), which runs in the opposite direction to A/D conversion (A/D変換).