Forensic video: 10 things every investigator should know

Forensic video evidence has changed the way police investigations happen around the world.

Twenty years ago, forensic video was limited to analog video devices and had fewer technical challenges. Today, however, video evidence is involved in 85% of all criminal investigations and 94% of users experience challenges caused by the digitization of data.

This incredible growth of video has led to a democratization of video evidence – no longer is it feasible for a forensic video lab to exclusively handle all video evidence. While forensic video expertise may be critical in many investigations, video is also frequently reviewed by detectives, investigators, and others with minimal training. In fact, 70% of sworn officers interact with video evidence and 47% have minimal to no training.

Recent survey data shows that the more training an investigator receives, the more concerned they become about misinterpreting video evidence.

Why is that? Video evidence can be misleading and what you see may not always be an accurate representation of what actually occurred. In this article, we will share 10 important things every investigator should know about video evidence, which will help you know when it’s safe to review video evidence yourself and when to seek help from an expert.

1) Infrared cameras change the colors in ways you might not expect 

Infrared cameras are great for recording evidence in dark places. It’s important to know, however, that the black and white image of an infrared camera changes the color of things differently than you might think.

The image above showed the gloved hand of a perpetrator involved in multiple home invasions and sexual assault. Investigators seized gloves from a potential suspect and wanted to compare to the glove captured on video.

Despite the surprisingly different pattern, this glove with the yellow stripes is the same glove that was captured on video. As additional proof that it was the same glove, investigators wore it front of the same camera and demonstrated that it was indeed a match.

The glove appears so different in the video image because infrared alters the appearance of objects based on the material being illuminated, not the color of the object. The glove was comprised of multiple different fabrics, each of which resulted in a different degree of illumination.

Investigators must understand and take into account the effect infrared cameras can have on the appearance of clothing and other objects and take that into account when conducting investigations.

2) Always work with the original file 

Investigators frequently work with converted video files. For instance, a CCTV camera might record to a proprietary .dav file format that frustratingly cannot play on the investigator’s computer.

Officers who encounter a file they cannot play spend an average of 83 minutes searching for a converter or player online. Unfortunately, these methods are not reliably accurate. Proprietary DVR players and online converters can alter the video evidence by changing colors and shapes, inverting the image, dropping frames, and more.

Investigators in this Use of Force case leveraged a screen recording of a proprietary player. While it may appear like the officers slam the suspect’s head into the ground, the capture process actually dropped frames and accelerated the appearance of the use of force in the video. The original video file told a very different story. Always work with the original file if you can!

The court eventually threw out the altered clip and instead reviewed the original file that provided a more accurate depiction of the event.

This is one of the reasons we created Axon Investigate. There are thousands of different file types, and we know that Investigators need a reliable way to review evidence accurately.

3) Perspective matters

Many lawyers have argued, “the camera doesn’t lie.”

If only it were so simple! When investigating video evidence, forensic video experts know that perspective matters. The placement of the camera and the suspect can influence our perception of events, which is why it is so important to always look for multiple angles and to understand how the perspective changes the image.

For instance, check out this image of Prince William.

If this is all you saw, you would automatically think he was making a rude gesture. Fortunately, we have multiple camera angles that show he was actually holding up three fingers.

One camera does not necessarily tell the full story. Investigators should always look for multiple sources to sync in time so that they can view events from multiple angles.

4) Video evidence does not speak for itself

Another popular misconception is that “video evidence is a silent witness that speaks for itself.”

This idea was given a lot of credibility because of a Supreme Court ruling in Canada, and the ruling has been adopted in many places around the world. It was also determined during the days of analog video, when there was a lot more consistency in video evidence quality and playback. 

Today, however, things have changed. Grant Fredericks, a certified Forensic Video Analyst, had this to say during an interview with Eugene Liscio,

“The concept that video speaks for itself is probably the most egregious introduction of an argument that I have heard, in court as it is testified today. This is because video does not speak for itself. Depending on how the video has been captured, colors may not present correctly, shapes may not present correctly. Oftentimes the video is not acquired accurately. Images will be dropped…So interpretation is critical. Video, like any other witness, can be misleading and must be interrogated.”

5) The timestamp on video evidence may not be correct

Many cameras will display the date and time on the video. This can be especially helpful when searching through hours of video to find a crime that only took a couple minutes and when syncing video from multiple sources.

Unfortunately, the times displayed on cameras cannot generally be trusted. User error, power outages, glitches, daylight savings, and many other factors can cause the timecodes to be wildly incorrect. In fact, it is not unusual to have a case where each device is off from real time by minutes, hours, or more.

Visit our MyAxon page to learn more about timecodes.

Investigators should always measure the offset to real time so that the evidence can be validated and aligned with multiple sources of evidence. This can save significant time and help move cases forward more efficiently.

6) Compression can remove important details

In movies and sports, reviewed video evidence is often incredibly high-definition. In the real world, however, investigators must frequently work with low quality footage.

The higher the quality of the recording, the more space it takes on a camera and computer. Storage space can be costly, so many recording devices and converters get around this through the process of compression. Compressions removes high-frequency detail so that viewers can get a reasonable picture of events, even if finer (often key!) details are lost.

The image below is from an armed robbery case. Defense counsel argued that his client couldn’t be the same person as the suspect on video because his client has hand tattoos. As you can see in the video evidence, the suspect does not appear to have any tattoos.

In this case, the tattooed defendant was proven to be the person captured on video. The DVR had problematically removed a significant amount of high-frequency detail in the image. This effect can be measured by using the macroblock analysis tool within Axon Investigate and then presented in court to articulate to a jury why images can’t be enhanced to the extent often shown by Hollywood.

This is a frequent challenge with license plate footage as well. Compression can blur the letters on the license plate beyond recognition or make them disappear entirely. When that data is not recorded due to compression, you cannot enhance it back. It simply does not exist in the recording.

Although some compression occurs in the original recording, additional compression is typically added each time a file is converted. This is another reason why it is important for investigators to work with the original file whenever possible.

7) MP4 and WMV Files may not be the original

When forensic video specialists are given an .mp4 file or .wmv file from a CCTV system to review, they immediately ask, “is this the original file?” Both .wmv and .mp4 are popular video file types and can often play on many devices. Far from being an advantage, it often means the file has been converted and is not the original format.

In a slip and fall case, the attorney reported that “no one could play the original file”, so they converted it to an .mp4 to make it playable. However, the conversion dropped 59 out of every 60 frames, inverted the image, and distorted the aspect ratio. When playing the original file in Axon Investigate, all of the frames were available for playback. The actual events became much more evident, and the lawsuit was quickly dropped.

Investigators must be aware that .mp4 files and .wmv files are rarely the original file type. While some cameras do record video in an .mp4 format, most CCTV cameras do not. Always look for the original file.

8) Check the aspect ratio 

Aspect Ratio is “the ratio of the width to the height of an image or screen.” Older “tube” televisions had a 4×3 aspect ratio with the width of the screen being 4/3 of the height. Today, most television screens and monitors are a wider format, like 16×9.

While many surveillance systems record files that are intended to display at 4×3 or 16×9, they often record at alternative sample aspect ratios. For examples, see below image from this webinar on Aspect Ratios and Resizing.

Notice the common 2CIF resolution is 704×240. This produces files that are recorded at “half resolution,” which stretches the appearance of objects. The image below, for instance, is only half as high as it should be. This makes the vehicle appear to be shorter and wider than it actually is.

Before the video should be played in court, the field should be converted with the proper tools that can correct these issues. 

And of course, we’ve said this multiple times, but investigators need to ensure they start with the original file. Many free converters and some DVR players will alter the aspect ratio of an image, distorting the shapes and making it more difficult to understand or determine what is happening. 

9) DVRs that record proprietary video may still play the video incorrectly 

Many CCTV camera systems record proprietary video files that only play within the special video player that is made by the same manufacturer. Investigators often assume that if you want to play the video correctly, you should use the corresponding proprietary player. Unfortunately, time and experience has consistently demonstrated that the proprietary players often misread their own video files!

Axon Investigate was designed to help solve this challenge, and provides an intuitive, efficient and forensically-sound solution for investigators to view, investigate and safely share thousands of different video evidence files and formats. Our team includes certified forensic video analysts and former law enforcement officials, and constantly progressing our software to meet the evolving needs of public safety around the world.

In addition, the Third-party video playback feature leverages the same video engine from Axon Investigate to accurately play countless video file types inside of Axon Evidence. This means agencies can store, manage, play and share all of their video evidence in one, secure location.

10) Wide-angle lenses can distort the appearance of distance 

Body worn cameras are among the most common sources of video evidence.

Body worn cameras typically record with a wide-angle lens so that more of the scene is captured, but this lens also distorts the appearance of distances. It is important for investigators to understand that “Objects in the mirror may be closer than they appear.”

Perceived distance comes up regularly in Use of Force cases. In the image above, while the suspect may seem to be 15-20 feet away, but he is less than 10 feet from the officer and may pose a greater threat than the lay person may assume.

Want to learn more? 

If you want to learn more about video evidence, we strongly recommend checking out our video investigation courses, available both online and in-person at events like Axon Week.

Investing in these training courses can help ensure public safety professionals — from detectives, investigators, forensic video specialists, patrol officers, field officers, attorneys and command staff — can build cases and testify about video evidence with accuracy and confidence.

Learn more and register for an upcoming class here