I'm building a free image converter. One day I looked at my landing page for /tiff-to-webp and realized I had 4 sections of generic content — "TIFF is a lossless format, WebP is smaller, click convert." The kind of content that exists on 500 other sites.
So I spent a few hours actually researching TIFF. Here's what surprised me.
The file size math nobody explains
A 24-megapixel photo at 16-bit per channel is 144 MB. Not because TIFF is inefficient — because it stores every pixel at full depth:
Width × Height × Channels × Bytes per channel
6000 × 4000 × 3 × 2 = 144 MB
A 45MP Canon R5 TIFF hits ~260 MB. That's not a bug. That's the point. TIFF was designed for editing, not delivery.
Nine compression types — most people know two
Everyone says "TIFF is lossless." That's incomplete. TIFF supports:
- Uncompressed — raw pixels, maximum compatibility
- LZW — 1.5–2× compression, most common
- ZIP/Deflate — slightly better than LZW for 16-bit images
- JPEG-in-TIFF — lossy, 10–20× compression
- CCITT Group 4 — 15–20× compression, black-and-white only
That last one is interesting. CCITT Group 4 is a lossless codec designed for bilevel (1-bit) images — fax machines used it for 30 years. A scanned business document compressed with Group 4 goes from 5 MB to ~250 KB. That's why multi-page TIFF became the standard for document archiving, legal scanning, and insurance workflows.
Zero browser support is by design
Chrome, Firefox, Edge — none of them render TIFF inline. Not a bug, not an oversight. TIFF was never meant for web delivery.
The Library of Congress lists TIFF as their preferred format for permanent digital preservation. They hold over 3.5 petabytes of TIFF files. The specification hasn't changed since June 3, 1992 — TIFF Revision 6.0. Thirty-plus years of stability is a feature when you're archiving cultural heritage.
The industries that actually use TIFF
When I dug deeper, TIFF turns out to be everywhere in specialized fields:
Professional photography — Lightroom exports 16-bit TIFF for Photoshop roundtrips. 65,536 tonal values per channel vs JPEG's 256.
Medical imaging — whole-slide pathology scanners produce BigTIFF files (64-bit offsets, no 4 GB limit). A tissue sample at 0.25 µm/pixel = ~80,000 × 60,000 pixels.
GIS / satellite imagery — GeoTIFF embeds coordinate systems and projections directly in the file. NASA Earthdata recommends Cloud Optimized GeoTIFF for all their data distribution.
Document scanning — multi-page TIFF with Group 4 compression is the standard for batch document processing. Every enterprise scanner defaults to it.
Archival — FADGI (Federal Agencies Digital Guidelines Initiative) requires TIFF for all government digitization projects.
The conversion numbers
Once I understood what TIFF actually is, the conversion ratios made sense:
| Target | Reduction | Notes |
|---|---|---|
| WebP | ~96% | 36 MB → 1.6 MB at quality 80 |
| JPG | ~91% | 82 MB → 7.6 MB at quality 85 |
| PNG | ~40–60% | Lossless, same quality |
| HEIC | ~95–98% | 96 MB → 2 MB, Apple only |
PNG being "only" 40–60% smaller makes sense now — both are lossless, PNG just has better compression. HEIC being 98% smaller makes sense too — HEVC is an extremely efficient codec.
What this did for my content
Before research, my /tiff-to-webp page said: "TIFF is big, WebP is small, convert here."
After research, I could write about:
- Why photographers export 16-bit TIFF from Lightroom before Photoshop retouching
- Why NASA uses GeoTIFF for satellite imagery distribution
- Why CCITT Group 4 still powers document scanning workflows
- The difference between LZW and ZIP compression for 16-bit images
Same converter. Same tool. Completely different content that's actually useful to someone searching the topic.
That's the SEO approach I'm taking with Convertify — research every format deeply before writing a word. It takes longer. But generic content is invisible.
Week 5 of building in public. Previous posts in the series on my profile.
