ZONN.ai Forensic Report

Case · 0549001B · 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 · 31/100

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 31/100. 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 · 3/100

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

  • Color Distributionread real · 6/100

    Global color distribution shape vs natural-photo baselines.

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
SigLIP AI Detector
50
CommFor (4803 Generators)
50
Manipulation Map (IML-ViT)
50
Pixel & Frequency Forensics7 detectors · mean 47
▸ expand
Error Level Analysis
3
Color Distribution
6
Noise Pattern
91
Pixel Analysis
75
Frequency Analysis
72
Edge Consistency
35
Compression Quality
50
Provenance & Metadata3 detectors · mean 45
▸ expand
ICC Profile
35
Metadata
50
C2PA Provenance
50

Image Quality

Dimensions
400 × 400 px
Aspect
1.000
File size
20.7 KB
Bytes / pixel
0.133

Frequency Analysis

Radial1.000
DCT0.831
Upsampling1.000
Cross-channel0.033
Power-law β
-3.37
Grid energy
0.253

Edge Consistency

CV 0.874
Cell 1: 0.1921Cell 2: 3.5262Cell 3: 1.5566Cell 4: 0.2770Cell 5: 0.2445Cell 6: 6.0783Cell 7: 2.7313Cell 8: 1.7592Cell 9: 0.6874Cell 10: 4.9362Cell 11: 7.6463Cell 12: 13.5954Cell 13: 2.7881Cell 14: 4.3712Cell 15: 7.2943Cell 16: 8.3642

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

Range: 0.192113.5954

Noise Fingerprint

Variance
21.14
Std deviation
4.60
Mean
-0.0
Spatial corr.
0.852
Mean Δ
0.88
σ
1.28
CV
1.449
Uniformity
-0.449

Provenance

Source Dossier

PlatformDirect upload
Author
Content Typeimage
Analyzed OnMay 12, 2026, 8:37 PM
Analyzed byAnonymous