EXIF & Image Metadata Explained

A photo is not just pixels. Almost every JPEG your phone or camera saves also carries a hidden block of text called EXIF — Exchangeable Image File Format — that records how, when, and often where the shot was taken. Most of it is harmless and genuinely useful. One field is not: the GPS geotag that pins the picture to a spot on Earth, often accurate to a few metres. This guide explains what that metadata actually contains, where it physically sits inside the file, the concrete privacy risk of sharing a geotagged photo, how stripping it really works under the hood, and when you should keep it rather than remove it.

What EXIF actually records

EXIF is a standard set of tagged fields written by the device at the moment of capture. It is descriptive data about the image, stored right next to the image, and it falls into a few groups. There is camera identity — make and model (e.g. Apple iPhone 14 Pro), the lens, and sometimes the firmware or editing software. There are the exposure settings the photographer (or the phone's auto mode) used: aperture as an f-number, shutter speed, ISO sensitivity, focal length, and whether the flash fired. There is an orientation flag, a small integer from 1 to 8 that tells a viewer how to rotate the sensor data so a photo shot sideways still displays upright. There is a capture timestamp — usually DateTimeOriginal, down to the second. And there is the field that matters most for privacy: GPS geotags.

The GPS geotag: a real, quiet privacy risk

When location services are on, a smartphone writes the latitude and longitude of the camera into the GPS sub-block of EXIF for every photo, typically to five or six decimal places — which corresponds to a resolution of roughly a metre or two. It may also store altitude and even a compass heading. On its own a geotag on a beach photo means nothing. The problem appears when the coordinates point at a place you did not mean to publish: your home, a child's school, a friend's flat, the gym you visit every morning. A single geotagged photo of a living room posted to a marketplace listing or a forum quietly hands a precise home address to anyone who downloads the file — and combining several photos reveals a routine: where you sleep, where you work, when you are out.

The reason this is easy to overlook is that the coordinates are invisible in normal viewing. You see the picture; you do not see the header. And reading it back requires no skill — any free EXIF viewer, or the tool this guide pairs with, will print the location straight out of the file. Some large platforms strip metadata automatically when you upload, which lulls people into assuming it always happens. It does not: many smaller sites, and nearly every file sent by email, chat, or a direct transfer like AirDrop, keeps the EXIF block fully intact.

Where the metadata physically lives in the file

To understand stripping, it helps to see where the block sits. A JPEG file is a sequence of segments, each introduced by a two-byte marker. The file starts with FFD8 (Start of Image) and ends with FFD9 (End of Image); the actual compressed pixels live in the Start of Scan segment near the end. EXIF rides in an application segment called APP1, marked by the bytes FFE1, which usually appears right after the file header. Inside APP1 the data begins with the ASCII string Exif\0\0, followed by a small TIFF structure — a byte-order mark, then Image File Directories (IFDs) that list each tag as a numeric ID, a type, and either a value or a pointer to one. The GPS data lives in its own IFD referenced from the main one.

The key insight is structural: the metadata and the pixels are separate segments. The APP1 block can be removed without touching the compressed image data at all — the picture does not know or care that its EXIF neighbour is gone. HEIC/HEIF, the newer format on recent iPhones, stores metadata differently — it wraps EXIF inside an ISO Base Media (MP4-style) box container alongside the HEVC-compressed image — but the same principle holds: the descriptive block is a distinct, removable part of the file.

How stripping actually works

There are two ways to remove metadata. The precise way is to parse the file, locate the APP1/EXIF segment, and rewrite the file without it — leaving the compressed pixels byte-for-byte identical. That is ideal but fiddly, and it must handle every format's quirks. The simpler, more robust way — and the one browser tools use — is to re-encode the pixels into a fresh file. The image is decoded to raw pixels and drawn onto an HTML canvas; a canvas is nothing but a grid of colour values, with no slot for an EXIF header. When the browser exports the canvas back to a JPEG or PNG, it writes a brand-new file containing only those pixels. There is simply nowhere for the old EXIF, GPS, IPTC, or XMP data to go, so it is gone by construction.

That bluntness is a feature: you do not have to trust that a parser found every hidden field, because the output was built from pixels alone. The honest trade-off is that JPEG re-encoding is lossy — the new file is re-compressed and is not a byte-identical twin of the original, so quality drops very slightly on each pass. PNG re-encoding is lossless, so PNG pixels survive exactly. A well-built stripper also reads the orientation flag first and bakes the correct rotation into the pixels before it discards the tag, so a sideways-shot photo does not flip when its orientation metadata disappears.

When to keep metadata, and when to strip it

Removing EXIF is not always the right move — the data exists because it is useful. If you are a photographer, the exposure settings are how you learn from a shot and how you cull and sort a library; the timestamp and camera fields drive catalogue software; and copyright and creator tags baked into IPTC/XMP metadata are how you assert ownership of your work. For archival and legal or forensic uses, the original capture time and the untouched file matter. In all of those cases you want to keep the metadata.

The time to strip is when you publish or share a photo beyond people you trust — a public social post, a marketplace listing, a forum avatar, a photo attached to a support ticket, a screenshot bug report. There the exposure trivia is worthless to the viewer and the GPS geotag is a liability. A good rule of thumb: keep metadata in your private, backed-up originals, and strip a copy at the moment you make it public. That way you lose nothing you might want later while never leaking your location to strangers.

The honest note on doing this privately

One caveat is worth stating plainly, because it is the whole point of a metadata tool. The safest place to strip location data from a private photo is a tool that never sends the photo anywhere. The paired EXIF Remover does exactly that: it reads the EXIF block and re-encodes the cleaned copy entirely in your browser, using the FileReader and Canvas APIs, with no upload and no server round-trip. You can even disconnect from the internet and it still works. Uploading a sensitive geotagged image to a remote "metadata cleaner" to protect your privacy is self-defeating — you have just handed the file, geotag and all, to a third party. Client-side is the only version of this that is coherent.

Open the EXIF Remover → to inspect and strip a photo's metadata locally, or browse all Polyatic tools.