Privacy, Authenticity & Forensic Analysis Score & Summary

Results are indicative only. All checks — including grades — are produced by automated heuristic algorithms and may contain false positives or false negatives. Results must not be used as evidence in legal proceedings, insurance claims, or professional forensic investigations, and do not constitute a professional opinion on image authenticity. See our Terms of Service for full details.

Original Photo taken from Camera Concerns

This does not look like it is an original photo taken from a camera. If you have already shared or sent the original photo then the personal & private information contained may already be extracted and removed. Otherwise the photo has been edited or resized and have some metadata lost in the process. It is always best to use the original photo taken from the camera as this is generally what is selected when sharing online. Use the original source to see what was revealed to the other party (i.e. the social platform, online service, or recipient).

C2PA Content Credentials

No C2PA Content Credentials were found in this file. C2PA (Coalition for Content Provenance and Authenticity) is an open standard that embeds verifiable provenance metadata — including the creator tool, signing identity, and edit history — directly into the file. The absence of C2PA data does not mean the content is inauthentic; most files do not carry it.

Clone / Copy-Move Detection

Copy-move regions detected. 38 matching block pairs found at spatially distant locations. This is consistent with content cloning, object duplication, or background filling.About this check: Copy-move forgery is a common image manipulation technique where a region of the image is copied and pasted elsewhere — to clone an object, cover up content, or duplicate a texture. This check divides the image into overlapping 16×16 px blocks and computes a compact block-DCT descriptor for each. Blocks whose descriptors match closely (within a Euclidean distance threshold) but are spatially far apart are flagged as candidate copy-move pairs. Low-variance blocks (uniform textures such as sky or walls) are excluded to suppress false positives. False alarms can occur on highly repetitive patterns such as fabric, grass, or tiled floors.

Resampling / Interpolation Detection

Strong resampling artifacts detected. Periodic patterns detected in the horizontal and vertical prediction residual — strongly consistent with the image (or a region) having been resized and re-saved.About this check: When an image is resized using bilinear, bicubic, or Lanczos interpolation, the resampling kernel introduces periodic correlations between adjacent pixels. This check computes a prediction residual (each pixel minus the mean of its left and right neighbours), averages the residual across rows and columns to produce 1D signals, then looks for a dominant spectral peak (FFT peak-to-mean ratio) that would indicate a regular, repeating pattern. A strong periodic peak suggests the image was resized; the axis (horizontal/vertical or both) indicates the direction of resampling. Nearest-neighbour scaling and lossless crops do not produce this pattern. Heavy JPEG compression can partially mask it.

ELA / Compression Artefact Inconsistency Analysis

No significant compression artefact inconsistencies detected. ELA residuals are consistent with a uniformly compressed image.

About ELA: Error Level Analysis re-compresses the image at a known JPEG quality level and measures the per-pixel difference between the original and re-compressed versions. A uniformly compressed image produces a uniform residual map. Composited or locally edited regions — having a different compression history — produce elevated residuals. Results are less conclusive on images that have already been re-encoded multiple times.

Histogram / Re-encoding Gap Analysis

No significant histogram gaps detected. The luminance histogram is continuous with only 0 zero bins in the usable range — consistent with an image that has not undergone repeated lossy re-encoding or destructive levels adjustment.

About this check: Every lossy compression cycle discards certain tone values from the luminance histogram, leaving zero-count bins ("gaps") that cannot be recovered by subsequent saves. An image saved once from a camera has a near-continuous histogram; an image that has been re-encoded, resaved, or edited repeatedly develops progressively more gaps. A comb-like alternating zero/non-zero pattern specifically indicates a levels or curves adjustment applied to an already-compressed image. Analysis excludes the extreme shadows (0–9) and highlights (246–255) to avoid false positives from clipping.

Metadata / Cross-field Consistency Analysis

No cross-field metadata inconsistencies detected. All checked metadata fields are internally consistent.

Double JPEG Compression Detection

No double JPEG compression detected. The re-compression residual curve is monotonically decreasing — consistent with a single-pass JPEG encode. No quantisation ghost was found across the probed quality range.About this check: When a JPEG is compressed twice (camera original at quality Q1, then re-saved at quality Q2), the quantisation tables from the first pass leave a characteristic fingerprint. Re-compressing the doubly-encoded image at a quality near Q1 produces a "ghost" — a local error minimum below the expected monotone trend. This probe tests five quality levels (60, 70, 80, 90, 95) and checks whether the mean absolute residual (vs the original) decreases consistently as quality increases. A rise in the residual at any quality level is the ghost signature. Note: this is a statistical heuristic; some innocent operations (thumbnail extraction, embedded preview generation) can also produce minor non-monotone artifacts.

JPEG Ghost Detection

JPEG Ghost detected. The probe found a MAD minimum at quality 95 (dip depth: 21.9%). This indicates the image was originally saved as a JPEG at approximately quality 95 before any subsequent processing. A ghost at a quality different from the current save settings can indicate re-compression — for example, uploading to a social media platform that re-encodes at a lower quality, or editing and re-saving the image in a different application. This does not by itself imply manipulation.About this check: The JPEG Ghost technique re-saves the image at every quality level from 5 to 95 and measures the mean absolute pixel difference vs the original. For a singly-saved JPEG the error decreases smoothly. If the image was previously saved at quality Q, re-saving at Q produces a local minimum — the "ghost" — because the quantisation tables approximately match. This probe uses a coarse sweep (every 5th quality) followed by a fine sweep around the minimum (±4 qualities) for precision. A ghost does not confirm tampering — it confirms that the image passed through a JPEG encoder at approximately quality Q at some point in its history.

Orientation Consistency

Orientation consistency check was not applicable for this file type.

Flash Consistency

Flash consistency check was not applicable for this file type.

Thumbnail / Embedded Preview Mismatch Analysis

Thumbnail mismatch analysis is only available for JPEG images with an embedded EXIF thumbnail. This file either is not a JPEG or contains no EXIF block.

Digital Watermark / Steganography Analysis

Anomalies detected — statistical or frequency-domain irregularities were found. This may indicate steganography or an invisible watermark, but is not conclusive.

Pixel-Level Statistical Analysis — These tests examine the raw pixel values of the image for patterns that are statistically unlikely in natural photos and may indicate hidden data embedded by steganography tools.Legend:

1. Chi-square and RS results on JPEG files may include false positives because JPEG re-compression alters pixel values in ways that can resemble steganography patterns.

Frequency-Domain & AI Analysis — These tests look at the mathematical structure of the image rather than raw pixel values, detecting invisible watermarks and anomalies that survive JPEG compression.

Add or Modify Copyrights / Authoring Details

Add or modify Copyrights / Authoring to your own content. Protect your works and declare ownership to your photo / images and metadata. Reduce photo theft by signing your works.

Manage Copyright in Vaultify