Did RBC manipulate audio file used for court evidence?
Audio Forensic Analysis Report
Date:10/27/2025
Prepared By: Mahmoud Khairy
Title: Audio Engineer
1. Purpose of the Examination
The goal of this analysis was to find out if a phone call recording between the complainant and a bank representative is genuine or if it has been changed in any way. The complainant believes the recording was edited to remove or change statements made by the bank employee.
This analysis looks for any signs of editing, tampering, or missing data that could affect the recording’s reliability.
File Name: 2025 08 02 at 12 35 45 Carene.wav Format: WAV (uncompressed, stereo, lossless format) Duration: 42 minutes 30.660 seconds
Source: Landline telephone call
3. Methodology
The recording was checked using standard professional audio examination methods. This included:
• Looking at the sound waves and visual sound patterns (waveform and spectrogram analysis)
• Reviewing the file’s technical information (metadata)
All checks were done on an untouched copy of the original recording to keep the evidence authentic.
1. Different Behaviour Between the Two Sides of the Call:
The recording has two separate audio channels, one for each person speaking. Silent gaps were found only on the complainant’s side, while the other side played normally.
2. Unusual Silence Pattern:
Only one person’s audio goes silent at certain times. This can’t be caused by a weak signal or phone interference, it strongly suggests that the recording has been altered
3. Signs of Manual Editing:
The silent parts start and stop very sharply, which usually happens when audio is deliberately cut or muted, not because of a technical problem.
4. Unnatural Silence:
During these gaps, there’s no background sound at all — total digital silence — which doesn’t naturally occur in phone recordings.
5. Landline Technical Impossibility:
In landline calls, both voices share the same connection, so losing only one person’s audio for a moment is not possible under normal conditions.
anomalies are summarized below:
Right when the call begins after being on hold and both parties say ?hello,? there’s a
0.747-second silent gap on the complainant’s side only.
This silence cannot be explained by a normal technical issue because:
1. The call was already active and two-way.
2. Only one side went silent.
3. There’s no natural reason for that pause.
4. The pattern matches other confirmed edits found later in the call.
6. Pattern of the Edits
The silences often appear right before key responses, for example, a small gap (from 00:10:19.745 to 00:10:20.048) comes just before the word ?yeah.?
This timing suggests the recording was intentionally edited rather than randomly interrupted.
5. Technical Explanation of Detected Edits
The spectrogram and waveform analyses of the two marked intervals reveal several features that make the edits unmistakable indicators of manual alteration rather than natural silence or line disturbance:
1. Abrupt Amplitude Drop: At both 00:03:05.580–00:03:06.327 and 00:10:19.745– 00:10:20.048, the waveform on the complainant’s channel shows an instantaneous vertical drop from active speech to complete digital silence (amplitude = 0). Natural pauses in speech or background quieting occur gradually, not as instantaneous straight- edged cuts.
2. Rectangular Spectral Profile:
On the spectrogram, the frequency energy terminates sharply across all bands at the same frame. This produces a vertical rectangular boundary, a visual signature of an edit or ?mute? command in digital processing software.
3. Noise Floor Discontinuity:
The normal noise floor before and after these points is around -48db. During the dead space, it drops to absolute silence (–? dB), something that cannot occur naturally on an analog landline recording, which always carries a faint line hiss.
4. Asymmetric Channel Behaviour:
Only the complainant’s channel is affected, while the bank representative’s channel remains acoustically continuous. True line disturbances (e.g., static, dropouts) would affect both channels equally. The one-sided pattern therefore reflects selective editing of one voice track.
Edited regions lack the ?click? or crossfade typical of authentic recording transitions. The waveform reappears instantly, demonstrating zero natural acoustic tail or decay a hallmark of hard digital splicing.
Contact us for the rest of the report.
Comments
There are 0 comments on this post
