Digital photographs rarely arrive as “just pixels.” Inside the same file that looks innocent in a gallery app, cameras and phones routinely embed Exchangeable Image File Format (EXIF) records, vendor notes, timestamps, software strings, and—when location services were enabled—latitude and longitude that can pinpoint where the shutter fired. PNG and WebP containers add their own optional text chunks, ICC color profiles, modification-time records, and sometimes Extended XML Platform (XMP) packets that duplicate or extend EXIF. TIFF and derived professional workflows may stack even richer tag sets across pages.
That metadata is useful while you are editing, auditing gear, or reconstructing a shoot. It becomes a liability the moment you redistribute the file outside a trusted circle. A single JPEG attached to a support ticket, uploaded to a classified marketplace, dropped into a public bug tracker, or forwarded to a source can leak home addresses, workplace perimeters, device serial context, and exact capture times. Social networks often strip some fields on upload, but the behavior is inconsistent across platforms, compression passes, and APIs; relying on “the app will clean it for me” is not a privacy policy.
SynthQuery’s Remove EXIF/Metadata utility is built for the opposite assumption: your machine is the only computer that should see the bytes until you decide otherwise. You load JPG, PNG, WebP, or TIFF files directly in the browser, review a concise summary of what was discovered, choose whether to remove everything, only GPS, only personal identifiers, or a custom combination of categories, and export cleaned copies with an explicit before-and-after byte comparison. Batch processing and ZIP packaging make folder-scale hygiene practical before you hand assets to PR, legal, HR, or the open internet. When you still need AI-assisted review of captions or documents that travel with those images, pair this page with the AI Detector or Humanizer workflows elsewhere on SynthQuery without ever mixing image bytes into our servers here.
Why metadata is easy to overlook
Operating-system file dialogs and lightweight viewers often hide EXIF behind an extra inspector pane. Mobile sharesheets compress or re-encode in ways that obscure whether tags survived. That visibility gap trains teams to treat “export as JPEG” as synonymous with “safe to post,” even when the encoder preserved every field from the camera.
Client-side processing as a policy control
Security questionnaires increasingly ask where media is transformed. Answering “in the analyst’s browser on a controlled workstation” is materially different from “uploaded to a third-party stripping service.” This tool keeps the sensitive operation local while still offering structured choices instead of opaque command-line utilities.
What this tool does
The interface centers on a queue: add one file or dozens via drag-and-drop or the file picker, each validated for type and size before any work begins. While summaries load, you see a short natural-language line describing format, whether GPS tags were detected, camera make and model when present, primary timestamps, artist or copyright strings, counts of PNG text chunks, and whether a WebP file carried EXIF or XMP at the RIFF level. That preview step answers the question “what am I about to publish?” without forcing you to interpret hexadecimal IFD listings.
Removal presets map to real-world privacy stories. “Remove all metadata” deletes EXIF APP1 segments from JPEG, drops ancillary PNG chunks such as textual metadata, embedded EXIF blobs, and modification-time records where configured, removes EXIF and XMP chunks from WebP containers, and rebuilds TIFFs through a lossless RGBA decode path so legacy IFDs do not survive. “GPS / location only” targets coordinate-bearing structures while attempting to leave benign camera fields intact when the container allows precise edits. “Personal data only” focuses on creator-facing strings—artist, copyright, host computer, common Windows XP style user fields, comments, and similar tags—plus personal-leaning PNG text keywords. “Custom categories” exposes five toggles—GPS, date and time, camera capture parameters, personal identifiers, and software strings—so security teams can align exports with data-minimization policies without recompiling a desktop tool.
Every processed row shows a byte-size before-and-after readout so you can explain file deltas to compliance reviewers. Individual downloads append a clear “-clean” suffix before the extension, and a single action zips the entire batch for handoff to DAM systems or email. JPEG, PNG, and WebP outputs prefer structural edits that avoid recompressing image data; TIFF exports note when a lossless re-encode occurred because baseline TIFF rewriting in the browser is the reliable way to shed nested tags across vendors.
Selective stripping versus blunt export
Stock agencies sometimes want color-friendly metadata while forbidding GPS; newsroom counsel may insist on removing authorship comments but preserving capture time for verification. Presets and custom toggles exist so you do not have to choose between “everything” and “nothing” on every batch.
Batch throughput and ZIP packaging
Journalists processing witness imagery, IT admins sanitizing ticket attachments, and marketplace sellers normalizing supplier photos all benefit from parallel queueing. ZIP export mirrors how those teams already move folders between tools.
Quality preservation philosophy
Where the format permits, SynthQuery edits container metadata without touching compressed scanlines or VP8 bitstreams. TIFF is the exception: stripping arbitrary vendor IFDs reliably requires rebuilding a clean baseline image, which the UI labels explicitly.
Technical details
JPEG stores EXIF primarily inside APP1 markers that begin with the “Exif\0\0” preamble followed by a TIFF-style IFD structure. Removing the marker without recompressing the entropy-coded scan leaves pixels untouched, which is why SynthQuery uses binary-aware paths for full and partial JPEG edits whenever piexifjs can rebuild a valid APP1 or omit it entirely. Orientation tags live in that same IFD; deleting all EXIF therefore deletes orientation hints, and viewers that ignore automated rotation may display the raw sensor framing.
PNG represents metadata through length-prefixed chunks. Ancillary chunk types such as tEXt, iTXt, zTXt, eXIf, and tIME are dropped or rewritten when policies require, while critical chunks like IHDR and IDAT remain intact so hashes of pixel data stay stable aside from deliberate edits. WebP wraps VP8 or VP8L bitstreams in a RIFF container; optional EXIF and XMP chunks sit alongside VP8X headers and animation frames. The tool walks top-level chunks, removes or rewrites EXIF payloads, and omits XMP when policies target location or personal bundles that commonly appear together.
TIFF is decoded through UTIF.js into RGBA, then encoded into a fresh baseline TIFF with minimal tags—enough for viewers to interpret width, height, and samples. That approach guarantees vendor-specific IFDs and private tags disappear, at the cost of potential file-size shifts versus heavily compressed source TIFFs. Animated or multi-page TIFFs only export the first raster page today; the summary notes when more pages exist so you can split work in desktop software if needed.
Why “no re-encode” is format-dependent
Raster recompression always introduces generation loss for lossy codecs. Metadata-only edits avoid that penalty for JPEG and many PNG/WebP cases. TIFF’s flexibility makes safe partial surgery rare without a full rebuild.
Thumbnail and MakerNote caveats
Some cameras embed large preview JPEGs inside EXIF. Removing EXIF removes those previews, which can shrink files dramatically and is usually desirable for privacy, but worth noting when comparing byte counts.
Use cases
Privacy-conscious creators scrub vacation photos before posting to neighborhood groups where EXIF can reveal front-door viewpoints. Humanitarian field workers remove GPS from documentation stills while retaining exposure metadata inside a separate secure spreadsheet when partners require both safety and technical traceability. Corporate communications teams sanitize executive headshots that were shot on personal phones so leaked attachments do not disclose home Wi-Fi naming patterns embedded in vendor-specific tags.
E-commerce operators preparing supplier JPEGs for open marketplaces strip location and serial-adjacent fields that sometimes appear in modern smartphone EXIF without realizing it. Legal teams responding to GDPR data-subject requests use stripping as one step in redacting imagery that will be returned to individuals or opposing counsel. Whistleblowers and journalists with sensitive sources should combine metadata removal with separate operational security; this utility removes file-level clues but cannot anonymize reflections, landmarks, or unique sensor dust.
Stock contributors often need copyright text in a sidecar contract rather than baked into EXIF personal fields; selective personal removal keeps color science tags while shedding author strings. Developers attaching screenshots to public GitHub issues benefit from PNG text-chunk removal when macOS or Linux tools embed unexpected descriptors. Whenever resized derivatives are also required, chain this step with the Image Resizer or WebP Converter so compression decisions happen after you know which metadata survived.
Social and marketplace safety
Even when a platform strips EXIF on upload, intermediate copies—email attachments, Slack uploads, cloud sync folders—may retain tags. Local stripping closes that gap at the source file.
Regulatory and contractual framing
Metadata can qualify as personal data under GDPR when it identifies individuals or precise locations. Document your stripping policy alongside retention schedules; SynthQuery surfaces measurable byte deltas to aid those records.
Creative and technical pipelines
Colorists may preserve camera profiles while dumping GPS via custom mode. Archivists might keep everything internally but ship stripped derivatives to vendors; this page targets the export slice of that story.
How SynthQuery compares
Desktop file property dialogs and basic “Save for Web” exports sometimes show a subset of tags yet offer little guidance on batch folders, GPS-only workflows, or PNG chunk hygiene. SynthQuery focuses on repeatable, policy-shaped exports with explicit size accounting.
Aspect
SynthQuery
Typical alternatives
Batch folders
Queue many files, process all, ZIP in one click.
Per-file property dialogs rarely scale.
Selective removal
GPS-only, personal-only, or custom category toggles.
All-or-nothing export settings vary by app.
Quality preservation
JPEG/PNG/WebP structural edits avoid recompression when possible.
Re-saving through editors may re-quantize DCT coefficients.
Privacy posture
Runs locally in your browser tab; no SynthQuery upload.
Online converters may upload to remote workers.
Cost
Free while you use the page; no account wall for stripping.
Some suites paywall batch metadata tools.
How to use this tool effectively
Start from the highest-fidelity originals you are allowed to process; re-saving a heavily compressed derivative before stripping only preserves whatever tags that derivative already carried. Open Remove EXIF/Metadata and choose a preset aligned with your policy—full removal for public blog heroes, GPS-only for real estate interiors that still need camera model proof, personal-only when you must retain exposure diagnostics but drop names, or custom when legal has enumerated allowed tag classes.
Use the dashed upload region or the browse control to add files. Each row begins reading metadata asynchronously; wait until the “Found:” line stabilizes so you know whether GPS or PNG text chunks were present. If a file fails validation, the toast explains whether the type or size limit was the cause so you can re-export from your editor. When the queue looks right, press “Process all” or run files individually if you want to spot-check one asset before batching the remainder.
After processing, compare the before and after byte counts. Large drops usually indicate removal of chunky MakerNote blobs or embedded thumbnails; small drops may reflect only a short EXIF header. Download single files for immediate upload, or bundle everything with “Download ZIP” when your workflow expects an archive. Keep an untouched master offline if you might need forensic metadata later—stripping is intentionally irreversible by design. For downstream text that accompanies the sanitized stills, consider running captions through the Grammar checker or Paraphraser when tone must match brand guidelines.
Upload and validation
Accepted types are JPG, PNG, WebP, and TIFF up to fifty megabytes each, capped at thirty simultaneous queue entries to protect lower-memory laptops. Unsupported extensions never enter the queue, which avoids confusing half-processed states.
Preset and custom mapping
Switching presets does not clear the queue; you can re-run a different policy on the same files after resetting results with “Clear.” Custom mode requires at least one category so you never accidentally export with zero effective changes.
Download and archival hygiene
Renaming cleaned files is still your responsibility if a CMS auto-versioning scheme depends on exact filenames. The tool never writes back to disk automatically; every save is an explicit browser download event.
Soften robotic captions or alt text drafts that describe imagery after metadata no longer supplies automatic context.
Frequently asked questions
Common fields include camera make and model, lens information, exposure settings, capture timestamps, software used to edit the file, author or copyright strings, and—when location services were on—GPS latitude and longitude, sometimes paired with altitude or heading. PNG text chunks and WebP XMP packets can repeat or extend those facts. Even without GPS, unique combinations of serial-adjacent tags and timestamps can fingerprint a device or workflow.
Many large platforms remove or rewrite at least some EXIF on published derivatives, but intermediates you control—email, messengers, raw CDN uploads, or internal wikis—may not. APIs and “original quality” modes change over time. Treat platform behavior as an unreliable safety net and strip locally when the file leaves your trust boundary.
For JPEG, PNG, and WebP, SynthQuery prefers deleting sidecar bytes or rebuilding EXIF segments without touching compressed image data, so pixels stay identical aside from orientation viewers infer differently when orientation tags disappear. TIFF cleaning uses a lossless RGBA round-trip that may change file size but not per-channel sample values.
The queue accepts up to thirty files, each up to fifty megabytes, to remain responsive on typical laptops. For larger archives, run multiple batches or continue using dedicated DAM tools with scripting APIs.
No. Stripping overwrites downloadable outputs and cannot recover deleted IFDs from SynthQuery because bytes never leave your device. Keep archival masters offline if you might need provenance later.
Many phones rely on EXIF orientation rather than rotating pixels in the file. Removing all metadata deletes that hint, so some viewers show the raw sensor orientation. Re-rotate in an editor if you need upright pixels without tags.
XMP bundles are difficult to partially sanitize in the browser without a full XML-aware editor. When GPS or personal categories are targeted, SynthQuery may drop entire XMP chunks to avoid leaving location or rights fragments behind. If you must retain specific XMP fields, use desktop DAM software with granular XMP editing.
The tool decodes and exports the first image page when multiple IFDs exist, and the summary warns you. Multi-page archival TIFFs should be split or processed in specialized software if every page must be sanitized.
No. Metadata removal hides file-level data only. Visual identifying features in the pixels remain unless you blur or crop them separately, potentially using other SynthQuery image utilities alongside manual review.
Processing uses JavaScript, Web APIs, and in-tab libraries such as piexifjs, UTIF, and JSZip. Files never POST to our infrastructure for this feature; network activity is limited to loading the page assets themselves.
Upload images to preview embedded metadata, choose what to strip, and download clean copies. Nothing is uploaded: processing runs in your browser.
Removal preset
JPEG, PNG, and WebP are updated without re-encoding image data when possible (structure or EXIF edits). TIFF is decoded and written as a new uncompressed RGBA TIFF, so size may change. Removing all JPEG EXIF also removes orientation tags—some viewers may show the raw sensor orientation.
Upload images
Drag & drop images here, or JPG, PNG, WebP, TIFF · Up to 30 files · 50 MB each