A photograph is more than pixels. Sitting alongside the image data in almost every file your phone or camera produces is a block of structured information called EXIF (Exchangeable Image File Format) describing how, when and where the picture was made.
Most people have never looked at it. It travels with the file anyway, and it survives being emailed, uploaded, downloaded and shared, right up until something deliberately removes it.
What is actually in there
The exact set depends on the device, but a modern phone photo typically carries:
Camera and device: make and model, often precise enough to identify a specific phone model and its generation. Also lens information, firmware version and, on some devices, the software that last edited the file.
Capture settings: aperture, shutter speed, ISO, focal length, whether the flash fired, white balance, metering mode and exposure compensation.
Timestamps: when the photo was taken, usually to the second. There are often three separate timestamps (original, digitised and modified), and on newer devices a UTC offset that reveals the time zone you were in.
Orientation: which way up the camera was held. This one is functional and carries no extra information about you. It is why a photo sometimes appears rotated in one program and correct in another.
Location: on phones with location services enabled for the camera, GPS latitude and longitude, frequently to five or six decimal places, a precision of roughly a metre. Often altitude, GPS timestamp and the compass direction the camera was pointing.
Thumbnails: a small embedded preview. Notoriously, this preview is not always regenerated when the main image is edited, so a cropped photo can carry a thumbnail of the uncropped original.
Alongside EXIF, files may carry IPTC blocks (captions, keywords, copyright, creator name) and XMP blocks (editing history from software like Lightroom, sometimes including the path of the folder on your computer).
Why this matters more than it sounds
The location field causes the real problems, because it is precise and because nothing in the normal flow of sharing a photo warns you about it.
A photo taken in your living room and posted to a forum contains your home address to within a few metres. A photo of an item you are selling, taken in your driveway, tells any buyer exactly where the item is. A photo taken at a friend’s house discloses their address, without their knowledge or consent.
The timestamp is quietly revealing too. A set of photos with EXIF intact establishes a timeline of where a person was and when, and because the data is structured and machine-readable, extracting it from a folder of images is trivial.
Some context worth knowing:
- Major social platforms strip EXIF on upload. Facebook, Instagram, X and most others remove metadata when they process an image, partly for privacy and partly to save space. They also often re-encode aggressively, which degrades the picture.
- Almost nothing else does. Email attachments, direct file transfers, cloud storage links, forums, marketplace listings, personal websites, Discord and most messaging apps that send “as a file” instead of “as a photo” all preserve the metadata exactly.
- Reading it requires no expertise. It is visible in the file properties dialogue on Windows and in the Info panel on macOS, and every image viewer and a hundred free websites will lay it out neatly.
The reasonable posture is deliberateness, with no need for alarm. Metadata is genuinely useful when you want it: sorting a photo library by date, remembering what lens you used, proving authorship. It just should not travel with a file by accident.
How converting removes it
When you convert an image in your browser, the file is decoded into raw pixels and then re-encoded from those pixels into the target format. The EXIF block is metadata attached to the original container and is no part of the pixel data, so it is simply not present in the newly built file.
Nobody had to implement this as a feature. It follows from how re-encoding works: the new file is constructed from the image and never copied from the old file. Nothing carries over except the pixels themselves.
What survives and what does not:
| Data | After conversion |
|---|---|
| GPS coordinates | Gone |
| Camera make, model, lens | Gone |
| Exposure settings | Gone |
| Timestamps | Gone |
| Embedded thumbnail | Gone |
| IPTC captions, copyright, creator | Gone |
| XMP editing history | Gone |
| Colour profile | Rendered into standard RGB |
| Orientation | Applied to the pixels, then discarded |
| The image itself | Kept |
The orientation row is worth a note. Because the rotation is applied to the pixels during decoding and is not carried as a tag, the converted file is correctly oriented everywhere. That incidentally fixes the “photo is sideways in one app and upright in another” problem that trips people up when a viewer ignores the EXIF orientation flag.
The colour profile row matters only if you are working in a wide-gamut workflow. Converting in a browser renders into standard RGB, which is right for the web and imprecise for professional print.
Doing this without uploading the photo
There is an obvious tension in using a website to remove location data from a sensitive photo. On most converter sites, stripping the GPS coordinates requires first sending the file, GPS coordinates and all, to a server belonging to someone you have never met.
That is why this site does the work in your browser instead. Your image is read from your own disk by the page, decoded in memory on your machine, re-encoded on your machine, and handed back to your downloads. No server receives it, because the site has no server-side component at all. You can verify this by opening your browser’s network tools during a conversion, or by disconnecting from the internet after the page loads and converting anyway. It still works.
Practical recipes
To strip metadata and keep the same format, losslessly, use the lint trap. It shows you what the file is carrying, then rewrites the JPEG without its metadata segments while copying the compressed picture data across untouched. The image is byte-for-byte identical to the original, so unlike a conversion it costs no quality at all.
To strip metadata with no quality loss in a new format, convert to PNG. It is a lossless container, so no further degradation occurs beyond what the original JPG already had. The file will be larger, which is the expected trade.
To strip metadata for a web upload, convert to WebP. The file is smaller, has no metadata, and shows no visible difference.
To see what is in there first, use the lint trap. It prints the camera, the timestamps and the coordinates before you decide anything, and links the location to a map so the implication is concrete.
To keep the metadata, keep the original file. Both stripping and conversion are one-way, and this site never had a copy to restore from.
To bulk-clean a folder, add up to a hundred files at once. Each is decoded, re-encoded and downloaded individually, and nothing is uploaded however many you add.
Summary
If you are about to share a photo anywhere other than a major social network, assume it is carrying the camera you used, the second you pressed the shutter, and the coordinates where you were standing. Converting the format removes all three, and doing that conversion locally means the coordinates were never sent anywhere in order to be deleted.