Multimedia
1. Two Types of Image — Bitmap & Vector
Bitmap: picture built from tiny boxes called pixels. Each pixel has its own colour; combine all the pixels ⇒ the image. Think: bitmap = pixels.
Vector: built from drawing objects / geometric shapes (lines, curves, rectangles…) + instructions about them. No pixels inside. e.g. the "T" = 2 rectangles; logos & text are vector too.
"Instructions" ≠ step-by-step moves. They are stored data: which shape, its length, width, colour, and where it sits (e.g. green rectangle inside red one).
2. Bitmap — Terms to Memorize
① Pixel
A pixel is basically a dot with one colour; nothing smaller makes up a picture. Link it to the grid picture above (one grey box, one green box…).
② Image Resolution
- Measured in pixels per inch; this value decides the amount of detail an image has.
- Higher resolution ⇒ more pixels per inch ⇒ more pixel information ⇒ high-quality image.
- Mental picture: keep the 1-inch bar + the letters PPI in mind.
- Screens usually show images at about 72 PPI; push it to e.g. 200 PPI for a much higher-quality image.
Same 2.54 cm (1 inch) width: 20 PPI circle is smoother than 10 PPI — more pixels in the same space ⇒ more detail. (Both are still bitmaps.)
③ Screen Resolution
- Width px × height px = total pixels on screen: 1920 × 1080 = 2,073,600.
- Every screen differs (laptop vs monitor vs TV).
- Look closely at a TV: each pixel is made of red, green, blue dots; all colours come from mixing these three. Black = R,G,B all 0; white = all 1.
- More pixels ⇒ better quality.
④ Colour Depth / Bit Depth
- A bit = smallest memory unit (0 or 1); bits join to make a byte (8 bits).
- Allow 1 bit per pixel ⇒ bit depth 1; allow 2 bits ⇒ bit depth 2; then 3; then 8 (needs 8 bits).
- Left box has place value 128, …, right box 1 — the 8 boxes of a byte.
- 1 bit: only 2 colours — 0 = black, 1 = white (no mixing).
- 2 bits: 00 black, 01 dark grey, 10 light grey, 11 white ⇒ 2² = 4. Now you can pick light grey instead of white.
- 3 bits: 2³ = 8 shades. 000 stays black, 111 stays white; new shades in between.
- 8 bits: 2⁸ = 256 shades.
Higher bit depth ⇒ more combinations ⇒ more accurate / better-quality image — but bigger file. Most commonly 8 bits are used to store each colour.
RGB colour (1 bit each)
RGB = Red, Green, Blue. With 1 bit each every colour is either off (0) or on (1):
- 000 → black (all off) 111 → white (all on)
- only R on → pure red; only G → pure green; only B → pure blue
- R + G on → yellow; G + B → cyan; R + B → magenta
- Raise the bit depth ⇒ you get shades of red, shades of green, shades of blue (mixed ⇒ many more colours). Don't memorise combos — understand them.
Past paper gives number of colours + options for bit depth; pick the smallest n with 2ⁿ ≥ colours.
Write powers: 2¹=2, 2²=4, 2³=8, 2⁶=64, 2⁷=128, 2⁸=256, 2⁹=512, then match.
• 2 colours → 1 bit • 256 colours → 8 bits • 249 / 250 colours → still 8 bits • 68 colours → 7 bits (2⁶ = 64 is too small, 2⁷ = 128 ✓)
Check both: number of colours and the bit options given — marks depend on it.
⑤ Metadata & File Header
Header = the "container"; metadata = what is stored in it. Open any picture's About / Properties and you see it. Not very important in detail — just know the definitions.
3. How Images Are Encoded Into Digital Format
- The images are stored as bitmap images.
- Each image is made up of pixels.
- Each pixel is of a single colour.
- Each colour has a unique binary number.
- The sequence of binary numbers is stored for the image (in the same pixel order).
Best order: enlarge the vector first, then rasterize ⇒ stays sharp. A rasterized image enlarged afterwards blurs.
4. Vector Graphics in Detail
Core idea: vector graphics store a set of instructions about how to draw the shape — not fixed pixels. When you enlarge it, it is re-drawn / recalculated, not stretched.
- Can be enlarged without pixelation (no blur, no visible squares).
- Images are re-drawn with each adjustment.
- Smaller file size — commands are stored, not individual pixels.
- Can be transferred and downloaded quicker (small file).
- Individual elements can be grouped to build an object.
- Suitable for geometric shapes: logos, text, simple drawings.
Vector terms
| Term | Meaning |
|---|---|
| Drawing object | An object made from different shapes (e.g. the "T" from two rectangles) |
| Property | Data about the shapes: length, width, colour, position… |
| Drawing list | The list of shapes used (e.g. the 2 rectangles) with their properties |
Even though a vector file has no pixels, it still shows on screen as pixels (after rasterization).
5. Vector vs Bitmap — Differences (Past-paper Q)
Describe the differences between vector graphics and a bitmap image.
| Point | Bitmap | Vector |
|---|---|---|
| made of | pixels | stores a set of instructions on how to draw the image |
| enlarged | pixels get bigger ⇒ pixelates | recalculated, does not pixelate |
| file size | usually bigger — must store data about each pixel | smaller — only instructions to create the shape |
| compression | can be compressed with significant reduction in file size | doesn't compress well — little redundant data |
6. Features of Graphic Editing Software
| Feature | What it does |
|---|---|
| Resize | increase / decrease the size of the image |
| Crop | remove a particular part of the image |
| Blur | reduce the focus |
| Red-eye reduction | reduce the red reflection of flash light in eyes (modern cameras often do it automatically) |
7. Calculating Bitmap File Size
- width × height = total pixels in the image (this is the image, not the screen).
- × colour depth (bits per pixel) ⇒ answer is always in BITS.
- bits → bytes: ÷ 8 (1 byte = 8 bits).
- bytes → KB → MB: ÷ 1000 each (or ÷ 1024 if told).
A digital camera takes a bitmap image 2000 px wide × 1000 px high, colour depth 24 bits. Calculate the file size in MB.
24-bit depth ⇒ 24 bits are used to store one pixel. Total pixels first, × depth, then convert bits → bytes → KB → MB.
Answer: 6 MB (mark scheme also accepts 5.7 MB)
8. Sound — Digital vs Analogue Data
Analogue examples: temperature, sound amplitude, length. Computers can't use these directly ⇒ convert to binary = digital. A sound wave is analogue; we must encode it digitally to store it on a PC.
Recording → storing → playing
- Microphone gives analogue data.
- ADC measures the wave at tiny regular time gaps and converts each amplitude to binary; the sequence is stored.
- To play back, binary goes to the DAC → speaker rebuilds the sound wave.
9. Sampling, Rate & Resolution
(Sound: taking measurements of sound at regular time intervals.)
General idea: filling a pool and measuring the water depth every 5 minutes — each depth reading is a sample.
Interval must be regular. In real recordings the gaps are milliseconds (very tiny), so it feels continuous — but it is still not perfectly accurate.
Low rate ⇒ wide gaps ⇒ many chances of error. High rate ⇒ bars follow the line closely ⇒ higher accuracy. (Blue line = original sound; bars = samples.)
How much memory each sample gets. More bits per sample ⇒ more accurate value stored… but a bigger file. (Same idea as bit depth for images.)
How is sampling used to record a sound clip? (same answer for "How is sound recorded digitally?")
- The amplitude of the sound wave is determined (measured)
- at set time intervals
- to get an approximation of the sound wave
- encoded as a sequence of binary numbers
- Increasing the sampling rate will improve the accuracy of the recording
Write all the points — it is worth 3–4 marks.
10. Quantization Error
Definition: the difference between the analogue signal and the closest digital value at each sampling instant.
- Blue = original analogue; red steps = what the digital recording stores.
- With a high sampling rate the gaps shrink ⇒ quantization error becomes much smaller; data is more accurate (never perfect).
11. Impacts of Changing Sampling Rate & Resolution
The greater the sampling rate, the greater the accuracy, but file size increases. Smaller gaps between samples ⇒ lesser quantization error.
Effects of increasing the sampling resolution
- More bits per sample
- Larger range of values ⇒ larger file size
- More accurate representation of sound
- Reduces quantization error
- Greater range of values can be stored
"What happens if sampling rate goes 40 → 60 Hz (or 60 → 40)? What if resolution goes 8 → 16 bits?" Answer in three things:
Mention accuracy, quantization error and file size.
12. Features of Sound Editing Software
- Edit start time / stop time
- Delete part of a clip
- Frequency of sound can be altered
- Fade in / fade out of a clip
- Mix multiple sound tracks
- Use of filters (enhance sound — e.g. auto-tune, voice changer)
- Conversion between different audio file formats
Fade in: sound starts soft, then increases. Fade out: loud ⇒ slowly gets softer (how background music ends in videos).
13. Calculating Sound File Size
- Rate in Hz = samples per second; × time ⇒ total samples; × bits per sample ⇒ total bits.
- Answer comes in bits ⇒ ÷ 8 for bytes (same conversions as images).
Leonardo records his voice twice — same length and sampling resolution. 1st: 44 100 Hz; 2nd: 21 000 Hz. Describe how the different sampling rates affect the recording and the sound file.
2 marks ⇒ two things asked: effect on the sound file + effect on the recording. Write both, and compare with 21 000 Hz.
Answer: 44 100 Hz → larger file + more accurate recording
Quick Recap — Whole Chapter
- Bitmap = array of pixels; vector = drawing objects + properties (instructions).
- Pixel = smallest picture element. Image res. = pixels per inch (PPI). Screen res. = width × height px (1920 × 1080 = 2 073 600).
- Colour depth = bits per pixel; colours = 2depth (8 bits = 256). Min bits: smallest n with 2ⁿ ≥ colours.
- Metadata = data about data; file header holds size, pixels, compression, colour depth, location.
- Encoding: bitmap → pixels → unique binary per colour → sequence stored. Vector → rasterized first.
- Bitmap: pixelates when enlarged, bigger file, compresses well. Vector: recalculated, no pixelation, smaller file, compresses poorly.
- Bitmap size (bits) = W × H × depth → ÷8 bytes → ÷1000 KB → ÷1000 MB (÷1024 for kibi/mebi).
- Editing: resize, crop, blur, red-eye reduction.
- Digital = binary 0/1. Analogue = measurement of physical property, continuous range.
- Sound path: mic → ADC → binary → DAC → speaker.
- Sampling = measuring at regular intervals. Rate = samples/s (Hz; CD 44 100). Resolution = bits per sample.
- Quantization error = analogue − closest digital value at each sample.
- ↑ rate or ↑ resolution ⇒ ↑ accuracy, ↓ quantization error, ↑ file size (more storage, slower transfer, more processing).
- Higher resolution ⇒ large dynamic range (soft ↔ loud).
- Sound file size (bits) = rate × resolution × time.
- Editing: start/stop time, delete part, change frequency, fade in/out, mix tracks, filters, convert formats.