ZONN.ai Forensic Report

Case · C7F8B2C9 · IMAGE

Analyzed byAnonymous
ZONN Analysis
0

Probably real

More signals lean toward real than AI, but some give weaker readings. Worth a second look on close inspection.

Signal ConfidenceWeak · 26/100

Analysed Specimen

Original analysed image
Forensic suspicion heatmap
OriginalHeatmap
POS55/100
No flagged regions

Heads up — 3 things to know

Why this analysis might be off

We highlight every disagreement and unusual signal we found so you can judge for yourself. Stronger warnings come first; informational notes are at the bottom.

Multiple detectors unreachable

Note

2 ML detectors did not respond (Ml Python Siglip, Ml Commfor). The verdict was computed with reduced evidence; reliability is lower than usual.

Upsampling artifacts in the frequency domain

AI evidence

FFT analysis found strong upsampling patterns — a fingerprint of diffusion-model VAE decoders (latent → pixel-space upscale).

Weak overall confidence

Note

Aggregate verdict confidence is 26/100. Several detectors returned uncertain answers or were offline. Read the verdict as a guideline, not as a final answer.

Origin Check

Trace this image elsewhere

Cross-reference the source against major reverse-image services. Each link opens in a new tab with the image URL preloaded — ZONN.ai does not re-upload the image.

Why this verdict

  • Error Level Analysisread real · 23/100

    Error Level Analysis. Re-saves and diffs to expose uneven compression regions.

  • Frequency Analysisflagged AI · 75/100

    Fast-Fourier inspection. Diffusion outputs leave periodic spectral peaks.

Model Agreement

100%

Variance across 3 ML detectors. Higher agreement means the models converged on the same reading; lower agreement means treat the verdict with care.

Evidence — 13 detectors reviewed

What each detector saw

Each detector independently gave this imagea score from 0 (definitely real) to 100 (definitely AI). The score above is their weighted consensus — detectors with higher confidence count more. No single detector decides; you read the spread.

ML Models3 detectors · mean 50
▸ expand
Manipulation Map (IML-ViT)
50
SigLIP AI Detector
50
CommFor (4803 Generators)
50
Pixel & Frequency Forensics7 detectors · mean 45
▸ expand
Error Level Analysis
23
Frequency Analysis
75
Pixel Analysis
35
Color Distribution
37
Edge Consistency
47
Noise Pattern
52
Compression Quality
48
Provenance & Metadata3 detectors · mean 54
▸ expand
ICC Profile
62
Metadata
50
C2PA Provenance
50

Image Quality

Dimensions
1132 × 1390 px
Aspect
0.814
File size
234.6 KB
Bytes / pixel
0.153

Frequency Analysis

Radial1.000
DCT0.841
Upsampling1.000
Cross-channel0.141
Power-law β
-3.38
Grid energy
0.238

Edge Consistency

CV 0.587
Cell 1: 9.7870Cell 2: 16.3430Cell 3: 14.1837Cell 4: 9.8499Cell 5: 12.7165Cell 6: 35.5160Cell 7: 23.2964Cell 8: 11.9061Cell 9: 15.1834Cell 10: 43.8180Cell 11: 13.0137Cell 12: 9.3448Cell 13: 8.0825Cell 14: 23.8849Cell 15: 14.6113Cell 16: 7.4992

Per-region edge density (4 × 4 grid). Uneven distribution may indicate localized editing or splicing; uniform fields are typical of fully synthetic outputs.

Range: 7.499243.8180

Noise Fingerprint

Variance
129.59
Std deviation
11.38
Mean
-0.0
Spatial corr.
3.864
Mean Δ
2.29
σ
2.65
CV
1.158
Uniformity
-0.158

Provenance

Source Dossier

PlatformDirect upload
Author
Content Typeimage
Analyzed OnMay 12, 2026, 8:38 PM
Analyzed byAnonymous