What can image metadata reveal about your photo?

A photo file can quietly carry the camera that took it, the minute it was taken, and sometimes the exact place.

Some of those hidden notes are private, and a couple keep your photo the right way up and in the right colors. Here's how to tell them apart, and how to take out the private ones before you share.

FluffyBench · 7 min read · Measured October 15, 2026 with FluffyBench Remove Image Metadata

What is image metadata?

Think of a digital photo as a picture with a small label tucked inside its file. You never see the label when you look at the photo, but it travels with the file wherever you send it. Those notes are image metadata, and they come in two kinds: notes about the photo, and instructions for showing it.

Here's one photo file pulled apart: the picture on one side, its hidden label on the other.

One file: the picture and its hidden label

Most of the label describes the photo. A couple of lines tell apps how to show it. How we measured and reviewed

That's why you can't treat every note as private, or every note as safe to delete.

What can photo metadata reveal?

The best-known set of camera notes is called Exif. It can list the camera and lens, the date and time, the exposure settings, who took the photo, and a GPS location (the CIPA standards listing has the full specification). Location is optional: a camera or phone can leave it out.

Here's a real file: a National Park Service photo of Stanton Mountain. It names the camera, a Nikon D3200, and the minute it was taken, January 20, 2014, at 4:16 p.m., but no location.

The Stanton Mountain photo, ss-stanton.jpg, records the camera (Nikon D3200), when it was taken (January 20, 2014, at 4:16 p.m.) and its exposure settings, with no location. Other notes a photo can carry: location, author and copyright, title and keywords, and editing history, sometimes in a separate .xmp file. — This file has camera and time notes but no location. Other photos can carry more. · Photo: GlacierNPS / National Park Service, Public domain in the United States: National Park Service employee work; Public Domain Mark, via Wikimedia Commons.

Several of the other notes belong to a second set, XMP, which holds things like a title, keywords, copyright and editing history (see the Adobe XMP specifications). XMP can live inside the photo or in a separate .xmp file saved beside it.

Which image metadata helps your photo display correctly?

Two notes are instructions instead of descriptions. An orientation note tells apps which way is up, so a photo taken with the phone turned sideways still shows upright. A color profile is like a small recipe for the photo's colors; without it, an app has to guess. Our guide to image color profiles explains when colors shift.

These two instructions change what you see on screen, so they belong with the picture.

Two notes that change how a photo looks

An illustration: the Stanton Mountain photo has neither note.

What changes when you remove metadata?

Careful cleanup takes out the descriptive notes and keeps the instructions. FluffyBench Remove Image Metadata removes the camera, date, location and comment notes it recognizes in JPG, PNG and WebP files. It keeps the format, the color profile, an orientation note the picture needs, and the animation in animated PNG and WebP files.

Here's the Stanton Mountain photo before and after FluffyBench took its notes out.

Before: ss-stanton.jpg lists the Nikon D3200 camera, the time it was taken and its exposure settings, with no location. After Sanitize all: ss-stanton-sanitized.jpg shows the same photo with those notes removed. 0 of 2,449,920 pixels changed. — The camera notes are gone, and every pixel of the picture is the same. · Photo: GlacierNPS / National Park Service, Public domain in the United States: National Park Service employee work; Public Domain Mark, via Wikimedia Commons.

The picture itself isn't saved again. When we cleaned the Stanton Mountain photo, its camera notes disappeared and all 2.4 million pixels came out exactly the same.

Removing image metadata in FluffyBench

Remove Image Metadata runs in your browser, so your photos never leave your device. You can add one photo or several at once.

Here's the whole job in FluffyBench Remove Image Metadata, using the Stanton Mountain photo.

Step 1. Add the photo you want to share. JPG, PNG and WebP files work.
1. Choose images opens your file picker.
Step 2. The photo gets its own tile. Read what FluffyBench found before you clean it.
1. EXIF found: this photo carries camera notes.
2. Ready to sanitize: nothing has changed yet.
Step 3. One press cleans every photo in the list.
1. Press Sanitize all to take the notes out.
Step 4. Save the cleaned copy. Your original file stays as it was.
1. Private metadata removed: the camera notes are gone.
2. Download saves ss-stanton-sanitized.jpg.

Want to take the camera notes out of your own photo?

Remove metadata

Does converting an image remove its metadata?

Often, but that's a side effect, not the point. Converting unpacks the picture and saves it again as a new file. FluffyBench's converters generally don't copy the old notes across, so Exif and XMP usually disappear too.

Here's how the two jobs compare.

Removing metadata versus converting

Removing metadata leaves the picture alone. Converting makes a new picture file, and the picture, its colors or its size can change.

If all you want is to drop the notes, use Remove Image Metadata. If you need a different format anyway, our guide to image quality during conversion explains what can change.

What should you check before sharing a cleaned photo?

Removing metadata only touches the notes inside the file. A street sign, a house number, a face or a well-known landmark in the picture still shows where and who. A file name like "our-new-house.jpg" can say just as much, and a separate .xmp file saved beside the photo isn't touched, so leave it out when you share.

FluffyBench Image Viewer shows you the cleaned copy the way someone else will see it.

Step 1. Open the cleaned copy and look at the picture itself, the way the person you send it to will see it.
1. Look for places, signs and faces a stranger could recognize.
Step 2. Then open the file's details.
1. Image info opens the file details.
Step 3. The cleaned copy keeps your file name, with “sanitized” added at the end.
1. Check that the name gives nothing away, such as a street or a person.
Step 4. Image info shows the basics. It doesn't list camera notes or GPS.
1. Same dimensions as the original photo.

Image Viewer shows the picture and the file basics, not the notes. If one detail really matters, such as a location, open the cleaned copy in an app that lists Exif and GPS fields before you send it.

How we measured

Date: October 15, 2026. Browser: Google Chrome 155.0.8059.40. Device: Windows 11 Pro; AMD Ryzen 9 5950X 16 processor cores; 32 logical processors; 137.3 GB RAM.

Remove Image Metadata; Chrome image decode to sRGB canvas; compare exact decoded RGBA channels before and after cleanup.

An archived public-domain NPS JPEG with camera EXIF, plus our two 64 × 48 PNG and WebP test files with synthetic fields.

Decimal bytes and decoded RGBA pixels.

The NPS JPEG is the exact archived source revision recorded in our test-set list; its camera EXIF was not edited before the test. It has camera, capture-time and exposure fields but no GPS directory, orientation tag or ICC profile.

NPS JPEG: 678,202 bytes before and 677,986 bytes after cleanup, about 0.03% smaller, because the camera notes were a small part of the file. 0 of 2,449,920 decoded pixels changed.

Synthetic PNG with Exif, text and GPS: 1,410 bytes before and 747 after; descriptive fields and GPS removed, ICC profile kept, 0 of 3,072 pixels changed. Synthetic WebP with Exif, XMP and GPS: 1,438 bytes before and 902 after; descriptive fields and GPS removed, ICC profile kept, 0 of 3,072 pixels changed.

The PNG and WebP results describe those two test files. Animation and orientation handling come from the tool's format rules, not from these samples.

The before-and-after diagram is drawn from the measured input and output JPEGs, resized for display. The orientation and color-profile diagram is an illustration. The walkthroughs were captured from the current local build at phone width in light and dark, using the same two JPEGs; they weren't timed. Image Viewer reports file size in binary units, so its 662 KiB is the same 677,986-byte file.

Image credits

Stanton Mountain reflected in Lake McDonald — GlacierNPS / National Park Service. Public domain in the United States: National Park Service employee work; Public Domain Mark. Resized to 1,280 × 851 and converted to WebP for display; the archived 2019 JPEG source revision retained its original camera EXIF for the local metadata-removal check.