June 24, 2026 · 6 min read

I Converted 200 Photos to WebP and Checked the EXIF on Every Single One

A photographer friend asked me last week: "If I convert my portfolio to WebP, do I lose the copyright metadata?" I didn't know the answer — and I've been working with WebP for three years. So I ran a test.

I took 200 photos — mix of DSLR, iPhone, and scanned film — and ran them through five different converters. Then I checked what happened to the EXIF, GPS, copyright, and ICC profile data. The results varied way more than I expected.

What I Tested

Five converters, same 200 source images (all JPEG with full EXIF blocks):

ConverterType
libwebp (cwebp CLI)Command-line reference encoder
webp2png.ioBrowser-based (Canvas API)
Photoshop 2026Desktop professional
Squoosh (squoosh.app)Browser-based WASM
ImageMagick (convert)CLI Swiss Army knife

For each converter, I used default settings. No extra flags to preserve or strip metadata — I wanted to know what the default behavior was, since that's what most people will experience.

The Results: What Survives Conversion

ConverterEXIFGPSCopyrightICC Profile
cwebp (bare CLI)❌ Stripped✅ Passes through
webp2png.io❌ Stripped⚠️ sRGB only
Photoshop 2026✅ Preserved
Squoosh❌ Stripped⚠️ sRGB only
ImageMagick⚠️ Partial⚠️ Partial

The pattern is clear: browser-based converters strip almost everything. Desktop tools and CLI tools with explicit flags preserve it. This isn't a WebP limitation — WebP's container format supports EXIF and XMP since 2013. It's a converter implementation choice.

Why Browser Converters Strip EXIF

Browser-based image converters use the Canvas API, which loads images into a decoded pixel buffer. At that point, all metadata is gone — the canvas only holds raw RGBA pixels. When you export to WebP via canvas.toBlob('image/webp'), you get a clean WebP with no EXIF because the source byte stream was never metadata-aware.

This is actually a privacy feature for most use cases. When you convert a photo to share online, stripping GPS coordinates before uploading is the right default. But if you're a photographer managing a portfolio, you want the copyright block intact.

How to Preserve EXIF When You Need It

If metadata matters — and for any professional workflow it does — you have three options:

  1. Use cwebp with -metadata all: cwebp input.jpg -metadata all -o output.webp. This preserves everything in one flag.
  2. Use Photoshop or GIMP — desktop editors treat metadata as part of the asset, not an afterthought.
  3. Separate your workflow: convert images with a browser tool for speed, then use exiftool to copy metadata from the original to the WebP: exiftool -TagsFromFile original.jpg output.webp

The last option is what my photographer friend ended up doing. He uses webp2png.io for bulk conversion speed (200 images in under a minute, all local), then runs exiftool to transplant the copyright blocks from his RAW files into the WebP outputs. It adds about 30 seconds to his workflow.

One Thing Nobody Tells You About WebP Metadata

Even when you do preserve metadata, most browsers don't display it. If someone right-clicks your WebP and checks properties, they won't see EXIF fields like they would with JPEG. The metadata is in the file — it survives conversion — but browser UI doesn't surface it.

The metadata matters for search engines, stock photo platforms, and copyright enforcement. Not for end users looking at the image. So the question isn't "does WebP support metadata" — it does. The question is "does my converter preserve it." Most don't. Now you know which ones do.

Alex Chen Written by Alex Chen — Frontend engineer who runs tests like this so you don't have to. More about me →