Understand what changes in your image.
Practical guides and original tests on metadata, SynthID, AI detectors and image quality. Start with the question you want to answer.
SynthID
Does converting PNG, JPG or WebP remove SynthID?
A format change can alter the container, encoding and pixels. Learn which changes matter and why the extension alone cannot predict watermark detection.
Does cropping or resizing an image remove SynthID?
Understand what cropping and resizing change, how to compare them fairly, and why a smaller image is not a verified watermark-free image.
Does JPEG compression remove SynthID?
Learn why a smaller JPEG is not proof of watermark removal, how to compare exports, and which quality changes matter before accepting a result.
Does removing SynthID bypass AI image detectors?
An image can pass a watermark check and still be flagged as AI. See controlled examples and learn which result matters for your image.
Does taking a screenshot remove SynthID?
A screenshot changes the capture and often the resolution. Learn what that does to metadata and why it does not establish watermark removal.
How to check whether an image contains SynthID
A practical verification workflow: keep the exact file, choose the right check, record its wording and separate missing results from negative results.
Visible watermark removal vs. SynthID: what is the difference?
A removed corner logo does not establish that an invisible watermark is gone. Learn how to distinguish the two and evaluate a remover’s claim.
What is SynthID, and where is it stored in an image?
Understand the difference between an invisible watermark, a visible logo and file metadata, and choose the right check for each.
AI detectors
Can a real photograph be flagged as AI-generated?
Yes. Learn how false positives happen, what evidence helps resolve a disputed result, and why repeatedly editing a real photo can make the investigation harder.
Can an image pass two AI detectors and still fail a third?
Yes. Different detectors can give different answers to the same file. Learn why a low score is useful only when you know how it was obtained.
Does adding camera grain help an AI image pass detection?
Camera-like texture and detector results are different things. Tests of grain, noise and camera-style processing show why appearance alone is not enough.
How to test whether an AI image remover actually works
A practical test design with positive controls, unchanged exports, exact files, quality checks and honest reporting of failures.
What does a 99% AI image detector score actually mean?
A detector score is not the percentage of an image made by AI. Understand thresholds, calibration and why scores from different vendors cannot be averaged.
Why does an AI image remover work on one picture and fail on another?
A successful example does not predict every image. See the same processing chain succeed on a selected portrait and fail on two other scenes.
Image quality
Can you remove SynthID without losing image quality?
A lower detector score can come with changed faces, text or colors. Compare exact archived images and learn what “no quality loss” should mean.
Does stronger AI image processing produce better detector results?
More visible change does not always produce a lower AI-detector score. Compare stronger redraws and learn when to stop processing an image.
Lossless vs. lossy image compression: what actually changes?
Understand file bytes, decoded pixels and visible quality with simple examples. Learn which kind of preservation matters for your export.
PNG, JPEG or WebP: which format should you use?
Choose an image format based on transparency, sharp detail, compatibility and size. Compare real exports instead of relying on universal format rankings.
Metadata
Does removing Content Credentials remove SynthID?
Understand C2PA records, invisible watermarks and credential recovery. An absent embedded manifest is not proof that every provenance signal is gone.
Does stripping metadata stop AI image detection?
Metadata cleanup removes information stored with an image. It does not necessarily change an AI-detector verdict or remove a pixel watermark.
How to remove photo GPS data without changing the picture
Separate location privacy from image editing. Learn what to inspect, how to preserve a master, and why orientation and color records need care.
Why did my image get bigger after removing metadata?
Fewer metadata fields can coexist with a larger file. Learn how re-encoding, format changes and compression settings affect the final byte count.
More guides
A careful checklist before you share a photo
Check the image, its attached details, and any credentials before making a sharing copy. Metadata cleanup has limits.
Photo metadata is not the same as an AI watermark
EXIF and XMP describe a file. A pixel-level watermark is embedded in the image. Learn why removing one does not remove the other.
What SynthID can tell you—and what metadata cleanup cannot
How Google DeepMind's SynthID differs from file metadata, what a watermark check can show, and what Deprint handles today.