Image editing Study Guide

📖 Core Concepts Raster vs. Vector – Raster images = grid of pixels; each pixel stores color/brightness. Vector images = mathematical descriptions (lines, Bézier curves). Layers – Transparent sheets stacked; each holds separate image data, enabling non‑destructive edits and blending. Selection – Isolates part of an image so edits affect only that region. Automatic Enhancements – One‑click tools (hue, brightness, red‑eye, sharpness) that apply global adjustments. Resolution & PPI – Pixels‑per‑inch determines printed size/quality; changing PPI without altering pixel count rescales the physical output. Histogram – Graph of pixel count vs. brightness; guides exposure, contrast, and brightness tweaks. --- 📌 Must Remember Rasterizing a vector → convert shapes to pixels; always possible. Resampling creates new pixel values when scaling images. Blend Modes: Screen = lighter additive; Soft Light = darker, semi‑transparent. Gamma Correction: $I{\text{out}} = I{\text{in}}^{\gamma}$ (non‑linear luminance stretch). Color Depth: 2 → 4 → 16 → 256 → 65 536 → 16.7 M colors (JPEG/PNG max 16.7 M). HDR Blending = merge multiple exposures for a single‑instant look. Noise Reduction reduces random variation but can blur fine detail if over‑applied. --- 🔄 Key Processes Converting Vector → Raster Open vector file → choose “Rasterize” → set pixel dimensions → export as bitmap. Layer‑Based Editing Workflow Create new layer → apply edit (adjustment, filter, paint) → mask if needed → blend mode → repeat. Making a Precise Selection (Pen Tool) Click to place anchor points → drag to shape Bézier curves → close path → convert to selection → refine with edge detection or feathering. Histogram‑Driven Exposure Adjustment Open histogram → move shadows slider right to lift blacks; move highlights left to recover blown highlights; fine‑tune mid‑tones with curves. Chroma Key (Green‑Screen) Replacement Select uniform background → apply “Color Range” or “Key” → delete/replace with new background layer → feather edges. --- 🔍 Key Comparisons Raster vs. Vector – Fixed pixels vs. scalable math shapes. Screen vs. Soft Light Blend – Lighter, additive vs. darker, translucent. Crop vs. Rescale – Crop cuts out a region, preserving its original resolution; rescale changes pixel dimensions (may lose detail). Unsharp Mask vs. Edge Enhancement – Unsharp mask boosts edge contrast via a blurred mask; edge enhancement directly raises local contrast around edges. --- ⚠️ Common Misunderstandings “Cropping reduces resolution.” – Cropping only discards pixels; the cropped area keeps its original pixel density. “Higher PPI always means sharper prints.” – Sharpness also depends on the printer’s DPI; exceeding printer capability yields no benefit. “Noise reduction can’t affect detail.” – Over‑aggressive reduction smooths fine textures, making the image look plasticky. --- 🧠 Mental Models / Intuition Layers as Transparent Sticky Notes: Each note can be moved, hidden, or recolored without affecting the others. Histogram as a “Brightness Balance Scale”: Left side = dark pixels, right side = light pixels; a well‑balanced scale (no big gaps) indicates good exposure. Vector as “Math‑Drawn Shapes”: Imagine a shape defined by equations; scaling the equations never pixelates the shape. --- 🚩 Exceptions & Edge Cases Resampling Artifacts: Down‑sampling with a simple nearest‑neighbor algorithm can produce jagged edges; use bicubic or Lanczos for smoother results. Color Depth Limits: Some file formats (e.g., GIF) cap at 256 colors; converting a high‑color image to such a format will dither or lose fidelity. Lens Correction Limits: Extreme barrel/pincushion correction can introduce stretching at image edges. --- 📍 When to Use Which Selection Tool Choice: Marquee – regular rectangles/ellipses. Lasso – freehand irregular shapes. Magic Wand – regions of uniform color/luminance. Pen – precise, editable paths for complex objects. Blend Mode Choice: Use Screen for brightening multiple exposures. Use Soft Light for subtle contrast without harsh highlights. Resizing vs. Cropping: Crop when you need better composition without losing pixel density. Resize when you must fit a specific pixel dimension (e.g., web upload). --- 👀 Patterns to Recognize “Banding” in gradients → need higher color depth or dithering. Histogram clumped at edges → over‑exposed or under‑exposed image. Halo around edges after sharpening → over‑applied unsharp mask. Red‑eye in portrait → automatic red‑eye removal will target bright red pixels in the eye region. --- 🗂️ Exam Traps Distractor: “Increasing PPI always improves print quality.” – Wrong; printer DPI limits matter. Distractor: “Cropping changes pixel count per inch.” – Wrong; cropping only removes pixels, resolution stays the same. Distractor: “Vector images cannot be edited pixel‑by‑pixel.” – Partially true; after rasterizing they can, but vector editors also allow pixel‑level edits on raster layers. Distractor: “Noise reduction never affects sharpness.” – Wrong; excessive reduction blurs details. ---