Edit bitmap images (selection, correction, resizing)

Graphics Creation – Editing Bitmap Images (Cambridge AS/A‑Level IT 9626)

Learning Objectives

By the end of this lesson you will be able to:

• Select specific areas of a bitmap image using the appropriate selection tools, masks and layers.
• Apply correction techniques – colour balance, levels, contrast, brightness and retouching – both globally and locally.
• Resize bitmap images while preserving visual quality and understand the effect on resolution, file size and colour‑space.
• Contrast bitmap and vector graphics, and outline the basics of frame‑by‑frame animation.
• Choose the most suitable file format, explain compression and colour‑profile issues, and recognise ethical considerations when using graphics.

1. Bitmap Fundamentals

A bitmap (raster) image is a rectangular grid of pixels, each pixel storing colour information.

• Resolution – measured in pixels per inch (ppi) or dots per inch (dpi). Higher resolution = more detail.
• Colour depth – bits per pixel (bpp). Common depths: 8‑bit (256 colours), 24‑bit (true colour), 32‑bit (true colour + alpha channel).
• Colour models
  • RGB – additive model for screen display.
  • CMYK – subtractive model for printing; conversion can shift colours.
  • HSL / HSV – hue, saturation, lightness/value; useful for selective colour adjustments.
  • CIELAB – perceptual model; often used when colour‑balance must be device‑independent.

Worked Example – Uncompressed file size

For a 1920 × 1080 pixel image at 24‑bit colour depth:

Size (bytes) = width × height × (bits per pixel ÷ 8)
               = 1920 × 1080 × (24 ÷ 8)
               = 1920 × 1080 × 3
               = 6 220 800 bytes ≈ 5.93 MB

2. Quick‑Reference Cheat‑Sheet (Exam Command‑Words)

3. Selection Tools, Masks & Layers

3.1 Selection Tools

3.2 Masks

• A mask stores the selection as a greyscale channel (white = visible, black = hidden, greys = partial transparency).
• Masks are **non‑destructive** – you can edit the mask at any time without altering pixel data.
• Typical workflow: Select → Add Layer Mask → Refine (Feather, Refine Edge) → Paint on mask with black/white.

3.3 Layers & Adjustment Layers

• Layers stack independent image elements; they can be reordered, hidden, or blended.
• Adjustment layers (Colour Balance, Levels, Curves, Brightness/Contrast) apply changes only to the layers beneath them, preserving the original data.
• Group related layers to keep the workspace tidy and to apply a collective mask or blending mode.

4. Correction Techniques

4.1 Colour Balance & Levels

• Colour Balance shifts the contribution of each RGB channel in Shadows, Mid‑tones and Highlights.
• Levels stretch or compress the histogram. Use the three sliders:
  • Shadows – left‑most slider.
  • Mid‑tones – centre slider.
  • Highlights – right‑most slider.

Linear stretch formula (input range a–b to output 0–255):

output = ((pixel – a) / (b – a)) × 255

4.2 Contrast & Brightness

Contrast expands the distance between light and dark values; brightness adds a constant offset.

newPixel = α × pixel + β

• α (alpha) – contrast factor (α > 1 = higher contrast).
• β (beta) – brightness offset (positive = lighter).

4.3 Retouching (Healing & Clone Stamp)

• Healing Brush – samples texture from a source area and blends it to match surrounding tones automatically.
• Clone Stamp – copies pixels exactly; useful for precise duplication but may need manual blending.

4.4 Exam‑Style Task

Task: Using the supplied photograph, increase the contrast of the sky by +30, brighten the foreground by +15, and remove a stray telephone pole using the Healing Brush. Save the final image as a JPEG with 80 % quality.

1. Make a Lasso selection around the sky.
2. Create an Adjustment Layer → Brightness/Contrast and set Contrast = +30.
3. Make a Rectangular Marquee selection of the foreground, add another Adjustment Layer → Brightness/Contrast with Brightness = +15.
4. Choose the Healing Brush, Alt‑click a clean area of the ground, then paint over the telephone pole.
5. Toggle the adjustment‑layer masks off to review the whole image.
6. Export → Save for Web (JPEG, Quality = 80 %).

5. Resizing Bitmap Images

5.1 Interpolation Methods

5.2 File‑Size Impact of Interpolation

5.3 Calculating New Dimensions

If an image of width W and height H is resized by a scale factor s:

W' = s × W  H' = s × H

5.4 Worked Example – Upscaling with Bicubic

1. Original size: 1200 × 800 px.
2. Desired scale: 150 % (s = 1.5).
3. New size: W' = 1.5 × 1200 = 1800 px, H' = 1.5 × 800 = 1200 px.
4. Choose Bicubic interpolation.
5. Uncompressed size (24‑bit): 1800 × 1200 × 3 = 6 480 000 bytes ≈ 6.18 MB.

6. Vector Graphics – A Quick Contrast

• Definition: Images are described by mathematical paths (lines, curves, shapes) rather than a fixed pixel grid.
• Key tools: Pen tool, Shape tools, Bézier curve handles, Pathfinder operations.
• Advantages:
  • Unlimited scaling without quality loss.
  • Typically smaller file sizes for simple graphics (logos, icons, diagrams).
  • Easy colour‑profile changes – vectors are device‑independent.
• When to use vectors: Logos, icons, UI elements, technical diagrams, any graphic that must be resized frequently or printed at many sizes.
• Common vector formats: SVG (web), EPS (print), PDF (mixed).

7. Simple Animation Basics

• Frame‑by‑frame animation – each frame is a separate bitmap layer; the timeline plays them in sequence.
• Export options
  • Animated GIF – up to 256 colours, suitable for short web loops.
  • MP4 / WebM – video codecs, full colour, longer duration.
• Key‑frame concept – define the start and end states; the editor interpolates intermediate frames (available in advanced raster editors).

8. File Formats, Compression & Colour‑Profile Considerations

Compression & Quality Tip

When saving a JPEG, a quality setting of **80 %** usually offers a good balance between file size and visible artefacts. Always preview at 100 % zoom before finalising.

9. Responsible Use of Graphics

• Verify that you have the legal right to use an image (copyright, licence, Creative Commons).
• Provide appropriate attribution where required.
• Avoid misleading edits – e.g., altering a portrait in a way that could be defamatory.
• Consider accessibility: add concise alt‑text for web images.

10. Practical Workflow (Step‑by‑Step)

1. Open & duplicate: File → Open, then duplicate the background layer (Ctrl + J).
2. Select the area: Choose the most suitable selection tool; refine with Feather (1–5 px) or Refine Edge.
3. Mask (optional): Click “Add layer mask” to keep the edit non‑destructive.
4. Apply corrections:
   • Adjustment layer → Colour Balance / Levels.
   • Adjustment layer → Brightness/Contrast.
   • Healing Brush or Clone Stamp on the masked region.
5. Review: Toggle mask visibility or press Ctrl + D to deselect and see the whole image.
6. Resize:
   1. Image → Image Size (or Scale Image).
   2. Enter new dimensions or a percentage; ensure “Constrain Proportions” is checked.
   3. Select an interpolation method (Bicubic for photos, Nearest‑Neighbour for pixel art).
   4. Confirm.
7. Save & export:
   • For web: File → Export → Save for Web (PNG‑24 or JPEG 80 %).
   • For print: File → Save As → TIFF (CMYK, embed ICC profile).
   • Keep a master copy in the native format (PSD / XCF) with all layers intact.

11. Common Pitfalls & How to Avoid Them

• Distorting aspect ratio: Always lock the width‑height ratio when scaling.
• Quality loss on up‑scaling: Use Bicubic interpolation; if the image is a logo, recreate it as a vector.
• Hard edges after selection: Feather the selection or apply a slight Gaussian blur to the mask.
• Colour banding after heavy contrast changes: Work in 16‑bit mode or add a subtle noise layer.
• Saving JPEG at too low quality: Preview at 100 % zoom; increase quality if banding or blockiness appears.
• Forgetting colour‑profile conversion: Convert RGB → CMYK before sending to a printer to avoid unexpected colour shifts.

12. Self‑Assessment Questions

1. Explain the difference between a selection and a mask. When would you choose one over the other?
2. Calculate the uncompressed file size of a 3000 × 2000 pixel image saved at 32‑bit colour depth.
3. Given an image that must be reduced from 2500 × 1800 px to 1250 × 900 px, which interpolation method would you select and why?
4. List three situations where a vector format (e.g., SVG) is preferable to a bitmap format.
5. Describe the steps required to create a 5‑frame animated GIF from a single photograph where the subject moves across the frame.