A big conspiracy:
Channel 4 video an absolute fake - Experts
It is said that proving a baseless allegation or a lie is more
difficult than creating a lie. In this hi-tech era creating a false
allegation through digitally improved versions needs lots of efforts.
The most controversial ‘lie’ of the UK based TV channel - the Channel 4
- may have to face the same scenario after airing the alleged video
footage that tarnished the image of the Sri Lanka government and its
Army which waged a professional war to defeat the deadly LTTE. With the
recent announcement of the US based LTTE lawyer, Vishuanadan
Rudrakumaran that the LTTE will continue as the LTTE’s proposed
‘Provisional Transnational Government of Tamil Eelam (PTGTE)’, it is
more apparent that though the LTTE militants were ‘swept’ militarily,
the ‘ghosts’ of the terror outfit are still haunting, sometimes taking
the bankrupt tv stations like Channel 4 for a ride.
While the government is denies all the allegations as baseless, the
experts in the field have technically proved that the video footage was
a fake and modified. The Sunday Observer interviewed the experts in the
field who strongly claimed that the Channel 4 video clip was fake and
made in favour of the LTTE.
In an exclusive interview with the ‘Sunday Observer’,the world
renowned specialist in video cording Siri Hewawitharana said that the
video footage was an absolutely fake and warned the ‘liars’ who intended
to tarnish the image of the country to be careful when producing and
airing fake documentary as the experts have access to modern tools to
Siri Hewawitharana, one of the world’s leading digital video systems
experts and the former head of Cisco’s global broadcast and digital
video practice, is presently the Executive Director at IPTV Systems. He
was the head of systems engineering for Star TV Hong Kong, Head of
Visual Communication for OTC and Director of Engineering for WIN TV. He
was also the chief architect of Optus’ Network Systems Design Broadcast
and Satellite TV operation, and was responsible for creating and
operating Optus’ $47 million Pay TV Video Operations Centre.
Excerpts of the interview:
It is done by an amateur and a poor effort
- Siri Hewawitharana
Q: What is your view about the video clip?
A: It is done by an amateur and a poor effort taken to create video
to show that it is real. The video is dubbed.
Q: If it is a fake video how do you prove it?
A: The total video clip is 1:02.781 (min). At the end Video stopped
at 01:02.312. Can see the video from .312 to .781 some how video editing
stopped at .312
The interesting factor is that audio editing stopped at 01:02.125,
indicating original video is edited since original layer stopped at
1:02.781 and video editing stopped at 01:02.312 and audio dub stopped at
1:02.152. If it is original, audio should have been all the way to
1:02.781 and should not have 2 video layers indicating original and
edited version. The original VIDEO 1:02.781 and edited video stopped at
1:02.312 and edited Audio stopped at 1:02.125 Why? Because it is edited.
Q: According to your findings, was this clip produced by a mobile
phone or separately and then included into the mobile?
A: As a specialist on video coding I decided to look into this video
footage and verify technical features of the video against the claim
that it came from a mobile phone.
Looking at the footage first thing I found to be strange was that
high quality of the video and lack of cascading effects and motion blur
associated with mobile video coding.
I got hold of original video that was on QuickTime format and other
was on AVI format and decided to put through various analyzers to see
the origin of the video from the mobile source.
The original video is of high quality (avi and QT format) and never
came from a mobile phone. Later they try to create this as a mobile
phone video format which is called 3GPP. Mobile phones can’t make this
quality motion vectors. We can analyze any video in this domain with
powerful analyzers for verification of motion vectors, video coding or
any stream and these tools are NOT available on public domain. In Video
coding or broadcasting if you want to create a final video for any
occasion it is the practice to start with high quality video first then
editing later for final product. Mpeg4 is a compressed format.
If we take channel 4 video, it says it came from a mobile phone video
source. There are only two formats in mobile video formats. One is old
3GPP format and new one is Mpeg4, H-264 part 10 which is MP4 format
which is highly processor intensive encoding and due to this issue,
mobile phones in today’s market do not have high quality since top end
of the mobile phones like iPhone or smartphones processors are not that
powerful to take good quality video that Channel 4 claim since Channel 4
video is much high quality than existing smartphone can create today.
Within H264 coding we also have extra component called Motion Vectors
(VMC) which are used to predict motion on temporal and spatial domain.
The Channel 4 video have quite high quality VMC and this VMC came from a
video camera and not from a mobile phone source. Also Mobile source also
tend to be blocked in nature when it come to motion.
Since original video is from AVI and Quicktime format, the whole
video scenario indicates that original video is of high quality that
came from a video camera source since mobile formats do not use AVI or
Quicktime (QT) which are high quality video formats. If they change
mobile format to avi or QT then video is supposed to be very bad
quality. BUT in this case avi and QT formats are of very high quality.
This indicates that someone transferred the camcorder video to a
computer for editing and sound was dubbed later and I can see that gun
shot was not in sync with the video and normally audio is always way
ahead of the video since video processing takes time and in this case
audio is very late indicating very amateurish video and audio editing.
Finally, I can see that, right at the end there are two video layers
and just before the last layer finishes, audio stopped.
Q: What is your final analysis on the issue?
A: This is a fake done by an idiot or idiots that does not understand
faking. Looking at the results, I can say this video never came from a
mobile phone since original video is of a quite a high standard and
motion vectors were of high quality (that never come from mobile phone)
and I also found that Tamilnet try to put this video as a 3GPP format
associated with mobile phones.
This also gives some clues since mobile phones or older got 3GPP
format and all the new ones are of H-264 which is mpeg4 part 10.
Original video is from a good quality video camera and later someone
tried to transfer this as 3GPP QT format where we can see some cascading
Q: What is you view about Channel 4?
A: The Channel 4 is near bankrupt and facing severe financial
Therefore, they try to keep their heads up by creating sensational
issues to attract their viewers with gutter journalism and make revenue
with fake stories. And there is also a rumour going around London that
New Labour politicians are behind this smear campaign against Sri Lanka.
It shows the extent to which the bosses of Channel 4 go in combining
malice and ignorance in equal measure.
Q: Why did you think that other countries did not pick up this issue
and is it because they know it is fake?
A: At least USA did not believe this since the original video was
mailed to me from US defence source for Verification. We do have friends
except old colonials. Issue is are they (Colonials) serious about Sri
Lanka and to engage with us in a honest matter?
Q: What are your recommendations to prevent such forgery?
A: There are so many modern tools that you can use for forgery and
also there are experts in this field who can go after those for
verification. Advanced Video Engineering knowledge is with very few
people globally and forgers can try and we will respond in kind.
Q: You are talking about high tech equipment that we can use to prove
this forgery and do we have access for such to prove these fake videos?
A: Motion Vector (VMC) verification and cascading effect verification
can be done with proper tools.
Forgers cannot get their hands on these unless they are in advanced
Conclusive evidence the original source of the video was not a mobile
phone - Dr. Chathura Ranjan de Silva
Dr. Chathura Ranjan de Silva, a Senior Lecturer, Department of
Computer Science ad Engineering of the University of Moratuwa and also
the Director, Centre for International Technology Unit of the same
university also confirmed that this video footage was a fake.
“First there were several evidences to conclude that the original
source of the video was not a mobile phone but a video camera such as a
digital camcorder”, he said.
According to Dr. de Silva, when a video is captured and encoded for
stage in digital format, it is processed by the encoding device to
estimate the level of information change between subsequent image
“In a video clip these differences are represented using set
parameters known as motion-vectors. The accuracy and smoothness of these
motion vectors depend on the capability of the encoder.
When encoding is done using a device such as a mobile phone, the
resultant motion vectors exhibits a coarse granularity whereas an
imaging device such as a digital camcorder would result very smooth
motion vectors”, he said.
He said that motion vectors were available in the said video footage
displaying a high level of smoothness that was not available on a mobile
“The optical features (response to contrast, back-lighting and field
depth) in the said video footage too showed evidence of using a complex
optical lenses and aperture control system, which again available only
in video cameras and not on mobile phones”, he added.
Dr. de Silva said that the audio pick-up mechanism (i.e. microphone
system) of a mobile phone was designed to capture human voice while
rejecting other types of background noise.
“However, this video footage has background noise outside the
response range of a typical mobile phone. This again is evidence to show
that the audio track had been captured using a microphone with wider
frequency response (i.e. like in a video camera) and not through a
mobile phone”, he pointed out.
He also pointed out that the reverb pattern and frequency analysis of
the audio track showed clear evidence that the gun-shot sounds were
picked up at a distance from the microphone.
“This does not match with the visual details of the video footage
where the weapon is fired at a close distance to the camera. The sound
track shows evidence of the presence of strong breeze (wind-noise) which
does not have any related evidence on the video track. This too suggests
that the audio with the gun-shot sounds and the dialogue in Sinhala on
the background has been dubbed at a later stage”, he explained.
Sound was dubbed later on - Brig. Prasad Samarasinghe
Brig. Prasad Samarasinghe, Head of the Sri Lanka Army Signal Corps,
sounding the same expert version, said that the entire video footage
could be separated to 584 images and 30 of them are blank and they had
also carefully analyzed the rest of 554 images.
The Army expert team had also conducted a test firing to monitor the
manner a mobile phone camera captures sound and motion pictures in a
similar environment shown in Channel 4 video footage.”
In this test firing we have observed that there was no visible delay
in the sound and the picture. In the Channel 4 video footage there is a
delay of 296 milliseconds between the motion picture and the sound. This
clearly shows that this sound was dubbed later on”, Brig. Samarasinghe
said adding that even if the movie type is changed to AVI or quickTime,
there could not be any such imbalance between the sound and the moving
images. Maj. P. A. Bandara, explaining the fake motions in the video
clip said that during the first shooting of the ‘soldier’, victim was
moving his left leg and the blood stains in the muddy surface are not
spread in the water which clearly shows that the tinkering of the blood
stains were done subsequently.
“One victim was wearing a pure white shirt. It is highly suspicious
how he did not have any mud stains when he had fallen on a muddy
surface. The other factor is that the soldier is wearing a white shirt
underneath the uniform but the soldiers of the Sri Lankan Army are not
permitted to wear such colours”, he said.
Maj. Bandara also said that the victims were having stiff muscles
when falling down after receiving the gun shots. “This is highly
un-natural”, he said.
Journalists for Democracy in Sri Lanka (JDS)
JDS is the supposed source of the so-called execution video - have
given their address in several releases as C/o INSD, Kremmener Strasse
2, 10435 Berlin, Germany and given the email address
email@example.com. A reverse address lookup online did
not reveal a match for anything called “INSD” at this address. Most
probably this is a shell/fictitious organisation. An acronym search
revealed nothing relevant for “INSD”. Our Embassy in Berlin may be asked
to check up on this address and who lives there.
A web search revealed that this organisation (JSD) does not seem to
have issued any statements prior to July 22, 2009 and it may be
reasonably inferred that it did not exist prior to that.
When the first release was issued on 22 July, it was given a great
deal of prominence on some media freedom cum human rights oriented
websites. It was titled: Sri LankA:’ Thirty-four journalists and media
workers killed during present government rule. Commentary on the piece
includes: “Rights group Journalists for Democracy in Sri Lanka (JDS)
said on July 22 that, under the Government of President Mahinda
Rajapaksa, 34 media workers had been killed, including 30 Tamils.”
(See:http://www.greenleft.org.au/2009/805/41431 and also http:/asiapacific.ifj.org/assets/docs/227/085/6e499e3-5f85a55.pdf
Journalists for Democracy in Sri Lanka established its web presence
by launching a blog jdsrilanka.blogspot.com on or around August 1, 2009.
They have no material older than August 1 except for the July 22, 2009
statement which was posted in August.
On August 11, a statement was issued through the Asian Human Rights
Commission (HK) titled: “Sri Lankan Government bars media from covering
elections in North”. It also received wide coverage.
When the video of the so-called “execution” was released on or about
August 26, 2009 it was given wide coverage. CNN called JDS “an advocacy
group based in Berlin, Germany, that accuses the Sri Lankan Government
of Harassing and intimidating journalists”.
Conclusion: given that the organisation JDS probably did not exist
prior to July 22, how is it that the international media gave the video
so much credence and carried it as truth without verification? While
efforts to discredit the organisation may be dismissed as attempts to
“kill the messenger” which the GoSL is often accused of, there are
legitimate questions as to the provenance of the video coming as it did
through a mushroom organisation when there are much more credible
institutions that could have been sued.
Human Rights Division,
Ministry of Disaster Management and Human Rights