Progressive JPEG Checker

Workflow & Usage

1. Add JPEGs. Drag & drop or click to select one or more .jpg/.jpeg files.
2. Instant analysis. The checker reads headers client-side - no uploads, no WASM, no workers.
3. Review results. Each card shows:
   • Progressive or Baseline (SOF2 vs SOF0)
   • Dimensions & precision (sample depth)
   • Subsampling (e.g., 4:2:0, 4:4:4, or grayscale)
   • DQT/DHT table counts
   • Scan list with Ss–Se and Ah/Al per SOS
   • JFIF presence & basic EXIF orientation (if found)

What This Checker Verifies

• Progressive vs Baseline
  Detects SOF2 (progressive) and SOF0 (baseline) markers so you know how the image will stream.

• Scan Structure (SOS)
  Lists every scan’s component IDs, spectral selection (Ss–Se), and successive approximation (Ah/Al) to reveal the pass strategy.

• Chroma Subsampling
  Infers 4:4:4 / 4:2:2 / 4:2:0 / 4:1:1 / 4:4:0 / GRAY from component sampling factors in the SOF segment.

• Tables & Metadata
  Counts DQT (quantization) and DHT (Huffman) tables; notes JFIF and basic EXIF orientation when present.

Tip: Use this checker to audit image pipelines, verify email-safe assets, or confirm CMS/CDN behavior after transforms.

Interpreting Results

• Progressive (SOF2): Renders in multiple passes. Good for perceived speed where streaming matters.
• Baseline (SOF0): Decodes top-to-bottom in a single pass. Fine for local fast loads; less pleasant on slow links.
• Scans: More scans can mean smoother progression but not always smaller files - it depends on the encoder strategy.
• Subsampling:
  • 4:2:0: great default for photos (smaller file size).
  • 4:4:4: use when color edges matter (logos/UI/illustrations).
  • GRAY: for true grayscale assets; smallest when appropriate.

Limitations

• No partial-pass preview: Browsers don’t expose progressive pass rendering to canvas.
• Header-only: We don’t decode pixels; corrupted files may be partially readable or fail early.
• Metadata variety: Some producers write non-standard segments; we aim to be polite but strict enough to avoid false positives.

How It Works (Under the Hood)

• Marker walk: Reads the byte stream, confirming SOI/EOI, then iterates segments by marker (0xFFxx) and length.
• SOF/SOS parsing: Extracts width/height/precision/components, then parses each SOS for components, Ss/Se, and Ah/Al.
• Subsampling inference: Uses component sampling factors (H×V) to map to human-readable subsampling (e.g., 4:2:0).
• Tables & metadata: Increments DQT/DHT counts; notes APP0 JFIF; lightly parses APP1/Exif for orientation.
• Privacy by design: Everything happens locally; you can export a JSON report for reproducible audits.

Use Cases

• CDN/CMS QA: Verify whether transformations keep progressive encoding and intended subsampling.
• Email & Newsletters: Ensure images remain progressive for better perceived load across clients.
• Performance Audits: Inspect scan strategies and table counts for assets that feel slow or heavy.

Tips

• Measure before changing. If a file is baseline but loads fast, switching to progressive might not be noticeable in all contexts.
• Mind subsampling for UI art. Logos and crisp graphics often look better as PNG/WebP or at least with 4:4:4.
• Pair with the Converter. Need progressive output? Re-encode with the Progressive JPEG Converter and compare size/quality.