BLOG

Smart Phone Video Enhancement

March 4th, 2016

snapshot-1-1419005We have hit the tipping point for smart phone video. Many people today own smart phones and use them for various day to day activities. Included with these tasks is the recording of video, which can come in handy in the event that an accident has taken place or a crime is being committed.

Many of our recent video forensic cases include video recordings that were created on a smart phone. Many of you reading this post may have an interest in learning more about smart phone video enhancement. In the following paragraphs, we hope to communicate some information that will guide you on how to handle recording an accident or illegal event taking place and what to do should you need enhancement of a smart phone video.

Since most smart phone video evidence is recorded freehand on a basic smart-phone camera, you can’t guarantee that the video evidence will be of legitimate quality. The video quality isn’t always the best, people may shake while recording or the placement of your subject can lead to lackluster video evidence.

When recording video on your smart phone, hold the phone horizontally when recording and do not zoom in. One of the most prevalent problems with the footage is that the shots are not stable. The chaotic motion of a cell phone video can make it hard to make out exactly what’s happening.

The two clips below will help you understand what can and cannot be done with video enhancement, including stabilization of shaky video. The first clip is of an iPhone video as it was recorded. No stabilization has been applied to this video.

In the next enhanced sample we apply warp stabilization. Notice how the frame follows the motion of the video, creating a more stable picture for analysis.

As you can see, the subject in the video was not close enough to the smart phone that was recording the video, making it difficult to make out the events as they occurred. Clip One is not very helpful in this condition for use in court. The subjects are very far away, making it difficult to make out what is happening.

Any digital video editing software will allow you to zoom in on a given shot. This will lower your overall video quality, and as CSI may lead you to believe, fixing it isn’t nearly as simple as pushing the “enhance” button on your computer.

The reason that the quality of video is reduced when you zoom in has to do with something referred to as “pixels.”

Simply put, pixels are the small boxes of color code that combine to make up a given image. Each image is composed of a series of pixels that display the image you see in front of you.

Expanding an image means also expanding the size of each pixel. Because each pixel is representative of one “frame” of a bigger picture, such as photos making up a collage, each individual square pixel will expand along with the image. The increase of pixel size can cause distortion to the image, therefore making it even harder to decipher the contents of an image.

This is where a video forensic expert comes in. We have training in video enhancement and experience using the hardware and software tools capable of clarifying and enhancing the events that occurred when recording your video.

As you saw in the second video sample above, the image has been expanded so that we can see the events that occurred up close. The quality of the video hasn’t been compromised. This is due to experience.

At the risk of being redundant, the biggest problem we see with smart phone video evidence is the orientation of the cell phone while recording. When we receive smart phone video evidence, many of the video recordings are vertically oriented. If your footage becomes evidence; not having a horizontal (or landscape) recorded video to work with really constricts your range of sight of a given event.

Utilizing the footage above as another example, notice that around the :25 second mark, the camera focuses on the Jeep pulling up, as another officer runs towards the scene. This takes our focus away from the scene, and we miss a substantial amount of the incident.

Had this evidence been shot horizontally, we may have been able to see what happened when that officer arrived to the scene, along with seeing the white Jeep parking in the background.

With the expansion of smart phone video technology, your smart phone can be of assistance recording crime and criminal activity. It is important to understand smart phone video enhancement from the forensic expert’s perspective. The highest quality video recording will reveal more in court. However, despite the quality, it is crucial to employ a video forensic expert to make the most out of your smart phone video.

Forensic Video Enhancement

December 17th, 2015

forensic Video enhancementOne of the activities we perform daily at Primeau Forensics is Forensic Video Enhancement. Video enhancement is one of our most requested forensic services.

Forensic video enhancement can help litigators understand events that have been recorded on video but are difficult to see because of movement or shaky cameras, the subjects distance from the camera or video is too dark. During the Boston bombings for example, CCTV cameras outside retail stores helped the FBI capture the terrorists who were responsible.

Some of the reasons video needs forensic enhancement is because too often cameras are not properly maintained. This is a huge problem that is not noticed until after the crime has been committed. Or, worse yet, the cameras are not installed or positioned properly.

In the following blog post I will provide you with a few particular tips when performing forensic video enhancement.

• Original Video- Always begin with the original video recording. If you have to export an AVI file to provide to police, keep a backup on thumb drive for insurance. If you have is a copy on DVD and not the original, then load the DVD copy of the video directly into your computer for forensic video enhancement.

• Computer Software- Use professional forensic software program to enhance like Adobe Premiere Pro and Clear ID. Primeau Forensics uses these as well as Adobe Photo Shop to enhance images exported from the video for forensic video enhancement. Remember, you can also export frames as still images to identify activity and other video components.

• Color Correction- Perform a color correction process first before performing any other forensic enhancement process. This is especially important if your video recording is dark and or was recorded at night. Be careful not to add too much brightness.

• Enlargement- If you need to enlarge a portion of the recorded video viewing area, apply after you review the footage on a large video monitor. At Primeau Forensics we use minimum 27” professional video monitors. Remember, the larger the playback monitor, the better you can see events in the video and the less you need to enlarge your video as an enhancement step. This is even more important in the courtroom.

• Equipment- If the recording system that made the recording you wish to enhance, before contacting a video forensic expert, learn the type of equipment that made the video you are enhancing. Often times a poor CCTV video export may be to blame for the poor video quality.

Forensic video enhancement is an art as well as a science. It is important to understand that you should use different filtering to get different results. Always begin with the largest file size and structure as close to original digital video recording as possible.

Before buying forensic video enhancement software, download a trial version and determine if you are comfortable using the software or would rather contact a video forensic expert to perform your forensic video enhancement.

Laquan McDonald Police Video With Audio is Fake

December 4th, 2015

On November 24, video of the Laquan McDonald shooting was released by the Chicago police department – video that did not have audio. Subsequently, a video of the same incident, allegedly with audio, was posted on YouTube. I was asked to take a look at the video and offer my professional opinion as to the authenticity of the audio portion.

During my investigation of the YouTube video titled “Is this the audio? Chicago Police dashcam video of Laquan McDonald shooting”, I discovered several anomalies and inconsistencies using Time Domain Analysis, Frequency Domain Analysis, and Critical Listening Skills. I have outlined these anomalies and inconsistencies.

Frequency Analysis:

Frequency_Analysis_Laquan

  • In the image above, the Spectrogram shows the cutoff frequency of the gunshots well above the cutoff frequency of the noise floor, or background noise (radio chatter & siren). A closer look at the difference in frequency content between the gunshots and background noise is displayed in the image below:

Frequency_Analysis_Laquan_2

  • The cutoff frequency of the audio content from the YouTube video titled “Is this the audio? Chicago Police dashcam video of Laquan McDonald shooting” is around 16 kHz. This is displayed in the image below:

Frequency_Analysis_Laquan_3

Based on my experience, audio recorded evidence produced from law enforcement vehicles contain a cutoff frequency of 4kHz. I have examined the frequency analysis of the audio recorded in the original video evidence with lack of radio communication & officer dialogue. The cutoff frequency analysis of the original video evidence is displayed below:

Frequency_Analysis_Laquan_4

Based on my testing and analysis, I can confidently say that the audio portion of this video has been manufactured and added to the video. For what purpose? Only the ‘creator’ of the video can answer that.

Frequency_Analysis_Laquan_5

  • The original video released by Chicago Police contains recorded audio content of crosstalk, and alternator or engine noise (see above image). The noises that are audible within the original video recording are low in amplitude but can be heard with a significant increase in volume. Because this digital recorder recorded an audio track, it is my opinion that it was functional and had the ability to record sound. Because of the lack of officer dialogue, radio chatter, we believe the lack of these sounds was due to the following reasons:
  1. The on-person lavalier microphones within the vehicle were muted
  2. The on-person lavalier microphones within the vehicle were disconnected
  3. The on-person lavalier microphones within the vehicle were deactivated
  • The gunshots, and radio chatter heard throughout the YouTube video titled “Is this the audio? Chicago Police dashcam video of Laquan McDonald shooting” are duplicated, equalized and are not genuine or authentic. Previously in this blog I discussed the inconsistency between the cutoff frequency of the gunshots and cutoff frequency of the background noise within the audio content. In addition, the audible fingerprint of the gunshots within the Spectrogram has a distinct shape, size, and intensity that are consistent with duplication or repetition. The frequency decay of the gunshot, timbre or sound of the gunshot, as well as duration of the sound are almost identical. The gunshots are displayed in the image below:

Frequency_Analysis_Laquan_6

The radio chatter sample at timecode 0:04.387 (MM:SS:MS) is an exact duplicate of the radio chatter sample 0:03.000 (MM:SS:MS). The conversation being spoken is identical. The difference between the two is that the duplicate has been processed using equalization to deceive the listener into believing it is additional radio conversation.

This video claiming to have genuine audio is indeed a fake.

The Video Recordings of the Shooting of Laquan McDonald

December 3rd, 2015

Laquan McDonaldOn November 24, 2015, police video that captured the shooting of Laquan McDonald in Chicago, Illinois was released to the public, almost 13 months after the incident took place. There has been a public outcry regarding not only the death of McDonald, but also the videos themselves.

The following blog post clarifies why there could be an absence of audio during the multiple videos released by the city of Chicago from the Laquan McDonald shooting.

Dash Camera Surveillance Systems Record Video and Audio

Video surveillance systems are closed circuit video recording systems that keep a pre-determined area under security by recording video and audio. Surveillance video cameras aid in deterring and documenting crime and other activity.

Some video surveillance systems record audio, others do not. Video surveillance systems in many department stores do not record audio. Video surveillance systems in gas stations, convenience stores and banks often do record audio, as well as video.

Police car dash cam surveillance systems record both audio and video for two reasons:

  1. To protect the police officer
  2. To deter profiling

Video Surveillance Systems like the ones used by the Chicago Police include Audio Recording

The question remains, why do the videos released by Chicago authorities of the Laquan McDonald shooting lack audio? What are the sounds heard on the video recordings that sound like sirens or whining?

If you watched the videos that have been released so far, electronic noise can be heard. This is defined as electronic cross talk. The digital video recorder in the squad car was recording audio, that is why we hear that whining sound. The problem is that no audio signal was being sent to the recorder. Was it a technical glitch or did the officers elect not to activate their body microphones? Perhaps they were all muted? Was there a problem downloading the digital recordings from each police car? Who has the chain of custody logs for the handling of this evidence?

In an interview for the Associated Press, our lead Audio Video Forensic Expert, Ed Primeau, comments on the lack of audio from multiple police cars being a red flag.

In a second interview for the Associated Press, published by ABC News, fellow Audio Video Forensic Expert Gregg Stuchman comments that ‘It’s plausible for a single squad car to have a glitch preventing sound recording.’ How could multiple cars not have recorded audio?

“I’ve never heard of it before,” Stutchman said. “It raises a red flag. The more likely explanation is that audio was intentionally switched off.”

The reality of this situation is that a full forensic investigation needs to be performed by a neutral, independent Audio Video Forensic Expert in order to determine the reason for the absence of police dialogue and radio communication from the audio portion of the digital video recordings released in the Laquan McDonald shooting.

We have examined thousands of police dash cam systems here at Primeau Forensics. Some include audio, others do not. There are three methods of activating dash cam video and audio.

  1. The officer activates the recording manually
  2. The squad car reaches a certain, predetermined speed for recording to begin
  3. The video is engaged in record when the squad car flashers are activated

One important note; the officers have the ability to mute or not activate their body microphone transmitter should they decide. However, the in car microphone almost always remains on.

YouTube Fake Squad Car Video

The Associated Press asked Primeau Forensics to examine a YouTube video of the Laquan McDonald shooting that has audio. After careful forensic examination of the  video, Ed Primeau concluded beyond a reasonable degree of scientific certainty that the audio track on the video is fake. Careful analysis of the audio spectrum contained on the video indicates many signs of post event alterations.

We hope this blog post about dash cam video and audio surveillance recordings will help clear up any misconception about the absence of audio on the Chicago Police Dash Camera Videos.

Audio in Video Evidence

April 29th, 2015

Audio EvidenceThough there are many ways to detect an edit within video evidence, sometimes a critical ear can be just as important to a video forensic expert. Though they say “A picture is worth a thousand words,” an audio file can be worth just the same. You just need to know what to look (or listen) for.

Many CCTV systems now have the capability to capture audio, and this audio could be crucial to the legitimacy of the evidence. Audio is one of the easiest ways to detect an edit when watching a piece of footage, but to do so there’s a bit you need to learn about audio, in itself.

Audio is comprised of “sound pressure waves,” which are waveforms representative of the change in air pressure in a recording. One characteristic of sound pressure waves is that they are always smooth and continuous, no matter what.

Let’s say, for example, you’re recording in an open, quiet room. While you’re recording, a rebellious teenager comes in the room and blows off his air-horn. Even though that loud sound completely changed the overall sound in the room, the wave that represents the pressure change will always be smooth and continuous.

The only time that a wave is not smooth and continuous is when an edit is made. Keeping this in mind will give you more of an idea of what you’re listening for.

When an edit is made to a recording, this disturbs the waveform. This makes it temporarily rigid and inconsistent. Now, remember; all sound pressure waves should be the opposite of that. So, the second that you hear a sound that is outside of that smooth, uninterrupted audio file, you know you have an edit.

How is that disturbance represented? Well, it will usually manifest itself in the form of a ‘pop’. In the context of video, it usually will only last a frame, but it will be there. If you hear anything out of the already established waveform, you know that the evidence has been edited.

Adobe (as always in the world of forensics) has other software that can allow you to more accurately detect these edits. For example, Adobe’s “Audition” uses what’s known as a “spectrogram,” that detects something known as the “noise floor”. As with audio waves, your noise floor will always be level and uninterrupted. In regards to a spectrogram, the spectrum recorded for a noise floor will always be consistent in visual characteristics. When you see a disturbance in that consistency, just as if you hear one in the audio file, you can tell that the audio, and sub-sequentially the video, has been altered.

Obviously there are many ways to detect edits visually. However, being a forensic expert means heightening your senses to a critical level, and though sight is the most prominent sense for a video forensic expert, having a critical sense of audio is just as important. The two work hand in hand to present the evidence, so having some knowledge of both will only excel your skills as an expert.

Increase in Body Worn Cameras

April 15th, 2015

vievuIn the last year, Primeau Productions has seen a huge increase in surveillance video cases. Most of this began with CCTV (closed circuit television) systems, which back in 2013 helped identify the Boston Marathon Bombers. More recently, body worn cameras have been adopted by many police stations across the US ever since President Obama authorized funding for police stations to purchase these systems. Currently, there are numerous police stations around the country that are testing out different systems and learning how to properly integrate them into their procedures. Many of the stations have been open with the public about their testing and have begun to approve further funding to outfit more officers with cameras.

The Grand Rapids, Michigan Police Department was recently testing two different kinds of body-worn cameras among eight different police officers, who presented their feedback on the camera systems online. The city has now approved the funding for two hundred officers to be outfitted with cameras. The Seattle, Washington Police Department has also been very open about their body camera testing, even releasing some of the footage online for the public to see. To maintain privacy, they blurred the video and removed the audio so no individuals could be identified. A large amount of the public has been pushing for police worn body cameras ever since the Michael Brown shooting in Ferguson, Missouri last August.

The benefit of body worn cameras is that they protect both the public and police officers. Many police stations are in full support of the cameras because there will no longer be a question of what happened during an altercation. If a complaint is made against an officer, they will be able to check the body camera video and see whether the officer was acting inappropriately or if the complaint is accurate. Many stations believe this will be very helpful with training officers and improving the relationship between the public and police.

The biggest issue arising from the increase of body worn cameras is the huge amount of data being created. Not only does this require a very large amount of storage space, but it must be stored securely so that the video evidence cannot be tampered with. Thankfully, many of the companies providing these body worn cameras, such as VIEVU, also include software that ensures the evidence cannot be meddled with between the camera and the system. VIEVU’s system in particular makes sure that any video being downloaded from a camera has not been tampered with prior to the download. Access to the video will be limited to authorized personnel only to maintain the authenticity and safety of the video evidence.

As a Video Forensic Expert, I see many benefits to this increase in body worn cameras. I have worked on numerous cases in which evidence from a body worn camera greatly helped the investigation and proceeding trial. Police dash cameras have often been used as video evidence for investigations but they often do not show the entire altercation because of their stationary view. Police body worn cameras add a second perspective to be used along with the dash cam which can be invaluable to an investigation. Having the two angles of the event in question allows anyone involved in the case to get a better picture of what happened.

 

Video Evidence – South Carolina Officer Shooting Unarmed Black Man

April 8th, 2015

South Carolina Officer Shooting Unarmed Black ManA South Carolina police officer was arrested yesterday for the murder of an unarmed black man. This is all because of a video that surfaced of the North Charleston officer, Michael Slager, firing eight times at the unarmed man as the man fled in an open field.

According to police reports, the victim in question, 50-year-old Walter L. Scott, continued to flee after being hit by the officer’s stun gun. Police report that Mr. Scott had taken Officer Slager’s stun gun, which lead to a reasonable pursuit.

However, the video seems to show a different story. The stun gun is dropped, and after Mr. Scott is gunned down and the officer is seen dropping something next to the unarmed man. It is not clear as to what was dropped; however, some fear that this was planted on the man after his shooting, as police report that the officer’s taser was taken.

Whatever the case, the innocent bystander who took this footage deserves praise for his actions. Not only did he take the responsibility to record the events, but he also utilized landscape mode on his cell phone to shoot the evidence.

As you see in the first few seconds of the video, his shot in “portrait” orientation would not have accurately captured the events as they occurred. Mr. Scott would have run off screen, and we never would have seen this happen. However, because he shot the video in landscape mode, we’re able to see the shooting clearly.

Situations like this help reinforce the importance of the little devices in our pockets. Smartphone video can make or break a case like this, and we need to understand why it’s so crucial to utilize the tools we have when an unlawful event occurs. If it weren’t for this Good Samaritan, this story may have gone unseen and unnoticed. It really shows how important video can be.

You can watch the video below via the NY Times:

 

Creating Video Work Products as an Audio/Video Forensic Expert

March 9th, 2015

Video Forensic LabVideo work product is a way to document forensic investigations, like evidence recovery, for reference at a later date. Processes and procedures are documented using a video camera during a forensic investigation for future use. I have referred back to my video work product many times when I have questions later during the evolution of the case.

As an Audio & Video Forensic Expert, I have examined hundreds of audio and video recordings to determine authenticity, as well as enhance characteristics of the digital evidence to clarify the events as they occurred. This video recording is referred to as ‘video work product’ and comes in handy.

There are a few different digital video recording platforms that I use when creating ‘video work product’. I often use VIEVU body worn cameras and HDSLR
photography based cameras. Each one of these types of systems serves a certain purpose in assisting with a forensic investigation, as well as the investigative process.

Over the last few years, I have seen firsthand the significance and overall efficiency that body worn cameras and their recorded video can bring to the public, law enforcement and legal proceedings. I personally use the VIEVU LE2 and LE3 body worn cameras. The LE3 records in 720p HD resolution and utilizes a 68 degree field of view. Other competitor cameras tend to use a 130 degree field of view, which captures a wider field of view but captures less detail. Detail is often more important when it comes to video evidence, which is one of the reasons I prefer to use VIEVU cameras.

These body worn cameras also contain digital audio recorders, which record MP3 format audio at a 44.1kHz sampling rate and a 64kbps bit rate. This high sampling rate captures the full range of human hearing, making any audio that is recorded on the camera more audible. In some cases, the client lawyer or law enforcement agencies that I work for require that no audio be recorded while video is being taken. The LE3 audio recorder can be switched off separately from the video, which gives me flexibility in such a situation. The LE3 records to either MP4 or AVI video format for easy playability across various platforms utilizing the H.264 codec. These formats also allow easy integration into my forensic programs, such as the Adobe Production Premium Suite. The 16 GB flash style storage system allows for either 12 hours of SD video or 6 hours of HD video and quick data transfer rates. The battery will last 5 hours during SD recording and 3 hours during HD recordings. The unit is also compatible with an external battery pack for extended battery life.

My main use for the LE3 body camera in my investigations is recording my forensic process in the field. This includes retrieving evidence from different systems so I can review the video later and include it in my report to support the authenticity of my work product and any evidence used in the case. Often times a forensic expert will be challenged by a client or opposing lawyer to verify the chain of custody of the materials produced during an investigation. Even minor details about how the investigation was conducted can have a large bearing on the authenticity of the evidence. Having a digital video recorder on my person during my forensic investigation allows me to capture both video of my process and my dialogue explaining the process. Including this work product to my forensic reports verifies the chain of custody and protects me as a forensic expert.

Another type of digital video camera that I use to produce video work product is an HDSLR photography camera. This type of camera equipment has become popular among the scientific community, as well as production companies, for its portability, versatility, quality and functionality. An HDSLR photography camera can use different size lenses to capture both images, as well as video, in different ways depending on the investigation requirements. HDSLR cameras record in 720p, 1080p, anamorphic and even 4k resolution. These cameras typically record at 30 minute intervals and have a battery life of approximately 2 hours of recording time, depending on the preferred quality and the available storage space. When connected to an external power source, these cameras can record for longer intervals of time. HDSLR cameras are great for recording a locked down alternative perspective to body cameras of an investigation or retrieval process. The flexibility of being able to produce individual still images as well as video throughout an investigation is also helpful with my forensic process.

In some investigations, a single video recorded perspective may not be sufficient to display the forensic process or document the events. Having another high quality camera with flexibility of perspectives and interchangeable lenses can capture aspects of my investigation that body worn cameras cannot. This lockdown feature of a point and shoot camera can also allow an investigator or client attorney to view the process as if they are sitting there watching in real time. Another use for HDSLR cameras as a forensic expert is recording accident reconstruction videos. An accident reconstruction video is a recreation of an event or series of events in the same environment that they occurred so they can be shown to a client investigator, client attorney and/or law enforcement. An accident reconstruction video is most effective to show the real life series of events as opposed to a 3D animation or a written statement of the events.

Video recorded by Closed Circuit television (CCTV) surveillance systems has been the dominant source of video evidence that I have investigated during my 30 years of being an audio video forensic expert.

Video evidence produced by CCTV systems can help solve crime, as well as reproduce accidents and disasters as they occurred for play back in many different settings. One significant use a video forensic expert has when recording video from a CCTV system is to create an exemplar. An exemplar recording is a recording made in the most similar way possible to the original piece of evidence using the same equipment, settings, environment and conditions of the original evidence. This recording is used as a comparison file to the original evidence to help determine the authenticity of the original evidence. Both the quality of the video and the metadata included in the files will be compared when conducting a forensic investigation.

It is a best practice of ours at Primeau Forensics to video record many forensic investigations, such as our exemplar creation process or evidence recovery, so if our client has any questions during the life of their case, this video work product can be referenced.

Body-Worn Cameras: Life or Death

January 27th, 2015

vievu2The more video becomes available the more real life situations enter the courtroom. The quality of body worn cameras as well as surveillance CCTV video is improving to the point that video forensic enhancement is not necessary. In a recent fatal shooting in Flagstaff, Arizona the murder of a police officer was recorded on his very own, department issued body cam. Had his murderer not committed suicide (with the officer’s weapon) he would have, when captured, gone on trial to prove his innocence, possibly with a claim of ‘self defense’ or that the officer attacked him and the gun went off accidentally. If this video was not created hundreds of thousands of dollars would have been spent investigating the circumstances of the officer’s death. None of that expense was incurred, all because of a body camera video.

Primeau Forensics has worked with body camera manufacturer, VIEVU, based in Seattle, Washington, to help their research and development of their body worn cameras, which are primarily sold to more than 4,000 agencies in 16 countries. Primeau Forensics helped them test the field of view (FOV) for their cameras. We researched the various FOV options to learn what degree of a lens was most optimal in body worn camera situations.

President Obama recently proposed a bill to provide funding to police departments for the purchase of body worn cameras. The proposal includes $75 million to help pay for 50,000 of the lapel-mounted cameras, with state and local governments paying half the cost. This proposal occurred shortly after the Michael Brown shooting in Ferguson, Missouri. This is an important and crucial development in forensic video. As a video forensic expert I am thrilled to see body worn camera video help litigators learn more about a series of events that occurred, first hand, saving our judicial system a lot of money investigating cases and, often, ensuring a correct decision in a person’s guilt or innocence.

Tips for Mobile Video/Cell Phone Enhancement

January 20th, 2015

cell phone enhancementIn past blog posts, we have discussed techniques that we use as video forensic experts that get the best results when conducting CCTV video enhancement. What we have yet to discuss is how we get the best results when enhancing mobile video recordings.

Since most mobile video evidence is shot freehand on a smart-phone camera, you can’t guarantee that the video evidence will be of legitimate quality. People are often nervous when video recording something like a fight, accident or other public situation and naturally shake while recording. Smart- phone video quality is often good at best. The placement of your subject and method of recording video can lead to poor video evidence that needs forensic enhancement.

As a video forensic expert, we know firsthand about all of the tools available in software programs and out board digital signal processors that can help your video provide valuable information to the authorities.

At Primeau Forensics, we have a team of video experts that have reviewed and enhanced hundreds of videos, some of which are from smart-phone cameras. One of the most prevalent problems that we have noticed through experience is unstable smart-phone video recordings. The chaotic motion of a cell phone video can make it hard to tell what’s happening.

One of the tools we use is made by Adobe; Adobe products have excellent tools for video enhancement assignments. One tool can be found in Adobe After Effects. It’s a plug-in entitled the “warp stabilizer.” This plugin creates a full frame reference for the shakiness of a smart phone video. It enhances the viewer’s ability to view the video better and determine the events that occurred more effectively.

For example, we have posted two clips below. The first clip is from an actual video enhancement case. It was recorded using an iPhone. No stabilization has been applied to this video. Look at how difficult it is to see the events as they occurred.

IMG 1363 from Primeau Forensics on Vimeo.

In the next video, we take the same source clip, but this time apply several filters including warp stabilization. Notice how it is easier to view the video as the frame follows the motion of the video, creating a more stable picture for analysis.

IMG 1363.MOV CLARIFIED VIDEO from Primeau Forensics on Vimeo.

In many cases, such as the video above, the subject at hand will not be close enough to the smart-phone capturing the video, making it nearly impossible to make out the events as they occurred. In the first video example above, the subjects are very far away, making it difficult to make out what is happening.

A trained video forensic expert knows how to apply tools that will enhance the video evidence. These tools allow you to zoom in on a video and clarify difficult-to-see events. This process may lower your overall video quality, and as Hollywood may lead you to believe, fixing it isn’t nearly as simple as pushing the “enhance” button.

The reason that the quality of this video is reduced when zoomed in has to do with the video “pixels.”Simply put, pixels are the small boxes of color code that combine to make up a given image. Each image is composed of a specific amount of pixels that make up the clear image you see in front of you.

To further explain, think of a collage. Have you ever seen a collage composed of much smaller pictures? Once the pictures are combined in a specific order (usually by color), it creates a much bigger image. Think of pixels as those smaller square pictures that make up the entire picture.

Expanding an image also means increasing the size of each pixel. Because each pixel is a representative of one “frame” of a bigger picture, such as the photos making up a collage, each individual square pixel will expand along with the image. The increase of pixel size can cause distortion to the image, therefore making it even harder to decipher the contents of an image.

Luckily, Photoshop and other Adobe programs provide solutions called “pixel interpolation.”

Generally, pixel interpolation provides two options for this video enhancement situation. You can utilize a plugin known as the “bi-cubic smoother,” which blends each cubic pixel together to create a more cohesive image.

The other plugin is known as the “pyramid shaper.” This allows for a similar process, however, this follows what’s known as pyramid coordinates, an alternate method of morphing the pixels.

Which one of these methods is best comes down to the circumstances surrounding your investigation.

As you see in the second video, the image has been enhanced so the viewer can better see the events that occurred. It is important to note that the quality of the video hasn’t been compromised. This is because steps were taken like applying the bi-cubic smoother allowing the pixels to cooperate making a image more clear.

The biggest problem we see at Primeau Forensics with mobile video evidence is the orientation of the cell phone while recording.

If you should find yourself recording a video using a smart-phone, first and most importantly be sure to turn the phone on it’s side or ‘landscape’. We receive a lot of video evidence where the phone was held vertically instead of horizontally. This is fine when your intent is to view the footage directly on the phone, but when that footage becomes evidence; not having a full frame horizontal recorded video eliminates a large portion of the scene being video recorded. The vertical position records a much more narrow view of the scene and instead records larger height dimensions. This keeps potentially valuable information off the videos and out of the court room. It dramatically constricts the viewer’s range of sight of a given event.

Regarding the footage shown above; notice that around the 0:25 second mark, the camera focuses on the Jeep pulling up as another officer runs toward the scene. This distracts our focus away from the scene, and we miss a substantial amount of the incident.

Had this evidence been shot horizontally, we would have been able to view more of the events including the activity that happened as the officer arrived at the scene. We would also have been able to better see the white Jeep in the background which was an important detail.

Using another smart-phone video as another example, the footage below is a recording from a smart-phone of the events that transpired after this year’s tragic Charlie Hebdo shooting in Paris. The two gunmen are seen driving away from the scene of the crime before encountering a police car. The suspects shoot at the police car, causing the police to retreat, and speed off again.

If you notice, you’ll see two black bars along the outside of the vertical video. This video was also recorded holding the smart phone in a vertical position. Had this video been recorded holding the phone horizontally we would have more information recorded that may have revealed additional useful information. This scene happened just before the gunmen got away and killed additional people.

With the expansion of mobile video technology, any cell phone could effectively record a crime scene that can go on to become crucial evidence. The highest quality video recording can help litigators more effectively.