32

2.1.2 Video 2 - MH17 crash: leaked tape proven FAKE by audio analysis. Анализ перехвата разговоров ополчения ДНР.

Figure 2.1.2.1: Video Source 2: MH17 crash: leaked tape proven FAKE by audio analysis. Анализ перехвата разговоров ополчения ДНР.
URL: https://www.youtube.com/watch?v=T34AB6CImTE

34

Figure 2.1.2.3: Video Source 2: MH17 crash: leaked tape proven FAKE by audio analysis. Анализ перехвата разговоров ополчения ДНР. (Web
Inspector view) - Last Modified 4th Aug 2014

36

2.2 Video Acquisition and Verification
YouTube has copyright of all YouTube videos and the permission to download the video file is
restricted (nodownload) as shown in Figure 2.1.1.2 and Figure 2.1.2.2. The acquisition was done using a
third-party website https://y2mate.com/ to acquire the two subject videos. This website maintains highest
integrity of YouTube downloaded video files. Refer to Table 2.2.1 showing the acquired file.
Acquisition was done using two tools,
i. FAW Project – FAW Professional
ii. Magnet – WPS

38

Figure 2.2.2 MH17 crash: leaked tape proven FAKE by audio analysis. Анализ перехвата разговоров ополчения ДНР..mp4

40
i. SSU, radio interception of conversations between terrorists, Boeing-777 plane crash.mp4

Figure 2.1.1: Video 1 - SSU, radio interception of conversations between terrorists, Boeing-777 plane crash.mp4 (acquisition using Magnet WPS)

42

ii. MH17 crash leaked tape proven FAKE by audio analysis. Анализ перехвата разговоров ополчения ДНР..mp4

Figure 2.1.2: Video 2 - MH17 crash leaked tape proven FAKE by audio analysis.Анализ перехвата разговоров ополчения ДНР..mp4 (acquisition using
Magnet WPS)

44

3.0 Video File Analysis and Audio Findings Statements
This section covers both Video files and Audio file finding statements.
3.1 Video File Analysis and Statements
i. Two video files were acquired (downloaded) from YouTube video medial channel. Both Video 1
and Video 2 are referred as below (Refer to Table 2.2.1).
a. Video 1 - SSU, radio interception of conversations between terrorists, "Boeing-777" plane crash
- URL: https://www.youtube.com/watch?v=BbyZYgSXdyw
- Original Format: WeBM - audiovisual media file format
- Option to Download only with Resolution of 360 pixel at size of 3.3 MB. Refer to Figure 2.1.1.2
- Downloaded file for analysis: SSU,radio interception of conversations between terrorists,
"Boeing-777" plane crash.mp4 (Refer to Table 2.2.1)
MD5: 43DC84D53A31B73E0D2188E9300E051F
b. Video 2 - MH17 crash: leaked tape proven FAKE by audio analysis. Анализ перехвата
разговоров ополчения ДНР.
- URL: https://www.youtube.com/watch?v=T34AB6CImTE
- Format: MP4 - MPEG-4 Part 14
- Option to Download with multiple choice resolution of 360 pixel, 720 pixel and 1080 pixel. This
is due to the quality of video uploaded into YouTube channel. Refer to Figure 2.1.1.2
- Downloaded file for analysis: MH17 crash: leaked tape proven FAKE by audio analysis. Анализ
перехвата разговоров ополчения ДНР..mp4 (Refer to Table 2.2.1)
MD5: 725F08830E3D7782DF9248CE5F0DE5A4
ii. In YouTube channel media, both video file (Video 1 and Video 2) was set to “no download”. Refer
to Figure 2.1.1.2 and Figure 2.1.2.2 showing the setting in YouTube media channel. However, the
acquisition of the video was done using a third-party tool https://y2mate.com/, FAW Project –
FAW Professional and Magnet – WPS. Refer to Figure 2.2 showing the video Acquisition and
Verification.
iii. Analysis on both videos date published and date when exactly created showing inconsistence of
dates. Published date is much earlier than last modified date, both videos had be modified.
a. Video 1
- Showing Published date on 17th Jul 2014

46

Table 3.1.1: Video File Analysis – Metadata from Video Files

Remarks
i. Date Published (17/07/2014) does
NOT match with date modified/date
encoded – 11/01/2018. Tampered.
ii. The encoded and Tagged date
showing UTC 2018-11-01 02:44:30
which is after ~ 4 years 3 months
incident of MH17 fatal incident.
iii. Total of 3 audio segment with 5
audio tracks added/encoded into
Video 1 audio stream and were
uploaded into YouTube channel.
iv. The audio tracks in the video are
not the first/original source of
original audio, hence these audio
tracks cannot be accepted for
analysis as they are not genuine and
not from original recorded form
regardless if this video is genuine.
v. Uploaded in YouTube in WeBM
format which had very high
compression and low quality of
audio (Lossy and exported file size
3.26 Mb).
vi. The author of the video did not
speak in the video only show text of
transcript in English. No other
speaker voice in the video except
for the speakers in the 5 audio
tracks.
vii. Refer to Table 3.2.1 for audio
stream analysis vii. Refer to Section 3.3 showing the
metadata of the Video 1.

48

3.2 Audio Track Analysis and Finding Statements
i. Video 1 " SSU, radio interception of conversations between terrorists, Boeing-777 plane
crash.mp4” file (source from uploaded YouTube videos) analysis was focused more on the audio
streams as the audio aired in the video was the intercepted recorded audio conversations.
ii. There was no speaker in the video speaking before and after the aired audio conversation. It was
intended to show the audio track with details who was speaking and the transcripts. Refer to
Section 4.0.
iii. The first main requirements and the dependencies for all audio tracks in the video file is the source
and authenticity of the audio files. It is very important to know when the audio tracks were
recorded, where it was recorded, how it was recorded/stored and how it was recorded. Since this
information is not available and undetermined, none of the audio tracks in the video can’t be
accepted as a valid, genuine and authentic source of audio recording. Refer to statement iv. and
v. Also refer to Table 3.1.1 and Table 3.2.1.
iv. Analysis in Video 1 metadata or content does not show any original source of the audio tracks,
none of the audio tracks are genuine. Refer to Table 3.1.1, Table 3.2.1 and Section 4.0.
v. YouTube only shows the only details of Video files uploaded as listed in Table 2.1.1 and Table 3.1.1,
it does not show the content of the audio track aired in the video container. There is basically no
way to get the details of each audio tracks in the audio stream due to it was compress and added
into single Video 1 container and uploaded into You Tube social media channel. Video 1 container
showing as single audio of Stream# 0.1 (und) below. Refer to Section 3.3 showing Video 1
Metadata detail.
Stream #0.0(und): Video: h264, yuv420p, 480x360, 25 tbr, 25 tbn, 50 tbc
Stream #0.1(und): Audio: aac, 44100 Hz, stereo, s16
vi. In a way, the audio stream (Stream #0.1) in Video 1 still holds the audio tracks (which was aired by
SSU) which can be used analyze each audio tracks and verify the content of the audio. This analysis
is based on the audio wave, the spectral of the speaker voice and the background noise floor which
still can be analyzed of each audio tracks. Refer to Section 4.0 showing the audio stream analysis.
vii. Analysis on Video 1 audio stream (Stream #0.1(und): Audio:) showing there are 5 intercepted
audio tracks conversation which were added by SSU in the video file. Refer to Table 3.2.1 showing
the details of each audio tracks. Refer to Section 4.0 showing the audio track analysis.

50

viii. The Audio 1 Track-1 spectral and noise analysis showing clear edition of the audio track (mono).
a. In Audio 1, Track-1 - I. Bezler (“Bes”) and V. Geranin were talking in the conversation as per the video.
b. Bezler (“Bes”) voice was recorded in right channel and V. Geranin was recorded in left channel. Original audio is actually a mono
recording.
c. It is clearly seen in the spectral analysis in audio Track 1 that, V.Geranin voice and Bezler voice speaking at the same time. This
making the real conversation being overlapped with one another at 0.18.5, 0:25.75 & 0:31:5. The noise level are seen different for
V.Geranin. It is possible phrases were taken from a different conversation and merged. Refer Figure 4.4.1.1.
d. Frequency level in spectrogram appears in all three phrases are changing, first phrase appears very low, second phrase appears
to be in high frequency, but middle spectrum was unseen, and third phrase shows medium frequency level. This shows phrases
are from different conversations and a gap which appears at the end of third phrase uttered by V. Gernanin in the Audio 1 Track-
1 is clear indication that his voices were overlaid and merged into the conversation. Refer Section 4.4.1.
e. A gap which appears at the end of third phrase ~0:32.6 uttered by V. Gernanin in the audio Track-1 (Left Channel) is a definite
possible indication that Geranin voices are merged into the conversation. Audio spectral at 0:31.75 – 0:31.85 (Left Channel)
showing added cut, paste and merger of other sound in this audio. Refer Figure 4.4.1.1.
f. Seen merged of audio at at 0:22.5, 0:23.5 and 0:30:5 (Refer to Figure 4.4.1.3). Different audio conversation starts at 0:23.30 (Refer
to Figure 4.4.1.3).

52

h. In Audio 2, Track-4, failed to overlay the merging of other audio between time frame ~1.24.2
and 1.24.3, editing is very clear at this part. Another merging between ~1:24.50 and 1:24.85.
obvious editing failure can be seen. Refer Figure 4.4.4.3
i. In Audio 2, Track-4, a new speaker (militant) voice appeared instead of Grek and Major voice
speaking between ~1:09.5 and 1:49.0. However, SBU stated that the conversation was
between Major and Grek. Refer Figure 4.4.4.4.
j. Voice samples were taken for details voice analysis but none of the voices are meeting the
requirement for voice biometric analysis as it has very short speech below 15 second and the
audio quality is low. Refer Table 4.5.1 and Table 4.5.2 showing the audio samples details.
a. Audio 2, Track 2 (Refer to Section 4.5.2)
i. Grek’s Audio recording
Frequency Response: ~2692 Hz
SNR 13 dB
Voice: 0.51 s
ii. Major’s Audio recording
Frequency Response: ~2810 Hz
SNR 24 dB
Voice: 4.13 s
b. Audio 2, Track 3 (Refer to Section 4.5.3)
i. Grek’s Audio recording
Frequency Response: ~1744 Hz
SNR 16 dB
Voice: 0.54 s
ii. Major’s Audio recording
Frequency Response: ~2627 Hz
SNR 34 dB
Voice: 5.11 s
c. Audio 2 - Track 4 (Refer to Section 4.5.4)
i. Grek’s Audio recording
Frequency Response: ~2110 Hz
SNR 38 dB
Voice: 3.99 s

54

iii. Militant 2 Audio recording
Frequency Response: ~2885 Hz
SNR 37 dB
Voice: 6.64 s
xi. Analysis on Video 2, MH17 crash leaked tape proven FAKE by audio analysis.Анализ перехвата
разговоров ополчения ДНР..mp4 were done at very high level.
i. There were many pauses/mutes in the audio stream of Video 2.
ii. Pauses/Mute were reasonable in the audio as the speaker was explaining the audio released
by the SBU.
iii. Original audio seems to be a mono recording.
iv. There is only one speaker explaining about the audio released by SBU.
v. Voices from the audio released by SBU was played in the video to explain about the
tampering.
vi. The speaker in Video 2 didn’t cover Audio 3, Track-5. The focus of this case is to analyze audio
tracks in Video 1. Video 2 is just a reference. Refer to Section 5.0

56

Figure 3.3.1a: MediaInfo results SSU, radio interception of conversations between terrorists, Boeing-777 plane
crash.mp4

58

Figure 3.3.1c: MediaInfo results of audio stream of SSU, radio interception of conversations between terrorists,
Boeing-777 plane crash.mp4 is produced by Google Inc.