# Benchmark sample sets (ImgifyTools)

Drop future original test assets here so comparison posts and outreach pitches can cite a **fixed, versioned** sample set.

## Folder layout (create as needed)

```
public/assets/benchmarks/
  README.md                 ← this file
  samples/                  ← lossless or high-quality masters (preferred)
    A-portrait.tif
    B-product-white.tif
    C-ui-screenshot.png
    D-logo-alpha.png
    E-landscape.tif
  exports/                  ← dated encoder outputs (optional)
    2026-MM-DD/
      A-portrait.jpg
      ...
  manifests/
    2026-MM-DD-format-test.json   ← settings, versions, notes
```

## Rules

1. **Do not invent numbers in blog posts** until files in `samples/` exist and a dated run is logged.
2. Keep masters **unchanged** across re-tests; only add new dated folders under `exports/`.
3. Record encoder name, version, quality/CRF, and whether metadata was stripped in the manifest.
4. Prefer open-license or original photos we own; never commit confidential client images.
5. Large binaries: if Git size is a concern, store masters externally and keep only manifests + small previews here—link the external bucket in the manifest.

## Linked content

- Blog methodology: `/blog/png-vs-jpg-vs-webp-vs-avif-file-size-quality-test`
- Compare pages: `/compare/remove-bg-alternative`, `/compare/tinypng-alternative` (methodology sections; results pending)

## Status

| Set | Purpose | Status |
|-----|---------|--------|
| Format size/quality (PNG/JPG/WebP/AVIF) | Authority format post | Samples pending |
| Background removal challenge (hair/glass/product) | Remove.bg comparison refresh | Samples pending |
| Compression target-KB | 20KB / TinyPNG outreach | Samples pending |
