Video Evidence: Handle with Care

Some of the new simplicities afforded to us can, unfortunately, cloud ones’ judgement when dealing with images and video for legal use.

Technology has, in most parts, made things incredibly easy. Take the example of photographs and video.  We all now love to quickly snap a memory or record some footage of an event. We can adjust the colour or light, crop out unwanted parts, or trim the end of a video. It’s then a simple click on the share button to immediately have that sent to friends or family via a messaging app or social media.

Not long ago, we had to carefully remove a film reel and have someone ‘develop’ the images for us. To have your own photographic dark room was a little extreme. The same goes for video where, having the ability to record, edit and then output a new ‘film’ was preserved for the select few with the knowledge, and equipment to do so.

Some of the new simplicities afforded to us can, unfortunately, cloud ones’ judgement when dealing with images and video for legal use. Why do it one way, when it’s so much easier to do it another, perhaps quicker way?

Even the relatively simple task of reviewing video evidence can be problematic.

In late 2016, at the conclusion of a trial in Nottingham Crown Court, for four men involved in the murder of Aqib Mazhar, Judge Rafferty stated, “there must never be another case in this country where those analysing CCTV don’t have the best equipment.”

The quote stems from the fact that it wasn’t until the trial had started that video material was properly reviewed and that significantly changed the weight of the evidence.

Whether it is CCTV evidence, mobile phone video or a sequence of images, the software used to review that evidence can alter the viewer’s interpretation. It could be that the player drops or misses frames. The player could present the video too dark, or too light. The player could change the shape and size of the image or video, resulting in objects appearing smaller or larger.  Many surveillance system players alter the image to make it look better, even though that is not what was originally recorded – scary, but true.

Again, you may initially think I am exaggerating here.

In 2015, a conviction of Indecent Assault was overturned at the Court of Appeal. Mr Mohammed Islam was earlier convicted at Flintshire Magistrates Court, where a CCTV image of a vehicle, alleged to be his, was used as evidence. After analysis and enhancement, it was proved not to be his vehicle and his conviction quashed. Mr Islam’s lawyer, Adam Antoszkiw, later stated the crucial evidence was not properly examined by police because of financial constraints.

Multimedia evidence, especially CCTV or low-quality mobile phone footage must be handled with care.

When someone asks, “send me a shot of the car, from the CCTV”, you need to trust that the CCTV evidence was acquired correctly to start with, that the processes involved in managing that digital evidence have not altered it in any way, and then, that the image was extracted from the video in a manner that can be relied upon.

Image integrity is key, which means that corners must stop being cut for the sake of ease or speed.

I remember a time when a colleague of higher rank approached me, and shared his frustrations with CCTV acquisition. Let me explain.

Of all video evidence, the most common is that from CCTV. It also causes the most technical difficulty. The reason behind this is that the industry is not regulated to ensure that the material captured is suitable for the purpose. I know – quite unbelievable really. As a result, we have systems that capture and retain information in a way that may not be understandable. The company does not even have to tell us. They can store the visual data, the timing data and, becoming increasingly popular, the audio data, using technologies that are secret, closed or proprietary.

It’s a bit like having DNA evidence, but no one being able to understand it because the method to interpret the data has been kept secret.

The proprietary and unstandardised method to capture, store and export CCTV has meant that many forces started to invest in dedicated staff to acquire CCTV, just like a Crime Scene Investigator. This was great, but then the cuts started to kick in, and guess what? That investment stopped.

So, back to the senior officer, I was being asked if there was a black box to magically get the evidence from CCTV systems. The explanation took a few cups of coffee, but suffice to say that although many people were pushing and developing products – each one had its limitations.

It came back, again, to the answer: if you want the actual, best evidence, you can’t cut corners. Send someone with the right knowledge and the correct equipment.

To understand the extent of this problem, look how many images end up in the legal system that are purely a mobile phone snap of a CCTV monitor. They have not known how to extract the native evidence, they have not been able to call upon the services of a CCTV recovery or retrieval team. They cut a corner and simply record the monitor with their iPhone. Not exactly best evidence! Does the native evidence ever actually get recovered – who knows?

There is one caveat here: major incident and fast, real-time suspect recognition. Grabbing that first shot and getting it out to the press have enabled several terrorists to be recognised and quickly located. However, in each case, the native evidence has been obtained afterwards to ensure integrity.

Multimedia can be copied, pasted, cropped, resaved and changed a billion different ways. So, when the final video arrives, can you clearly understand how it was created? I don’t mean with what software, I’m talking about its history.

Let’s work backwards from a video compilation that you may have received. The compilation shows the accused / defendant / suspect from three locations over a period of time. It’s a standard video file and you can play it in Windows Media Player. (Hopefully you are not still getting Video DVD’s, as there is another world of pain associated with those, but let’s not go there now!). That compilation has been created. It’s a presentation format. It is not the original evidence.

Each premise would likely have had a different system. There would, probably, have been other cameras. It’s also likely that there would have been more footage obtained.

When I receive evidence to be analysed, enhanced or compared, over 50% of initial submissions are not the original evidence. This is vitally important for integrity, authenticity, and for any subsequent work required to restore or enhance footage.

Working on a pre-processed image or video is like reading a statement with some of the words in a different language. During the translate stage, some things get added and some things get lost. As a result, the meaning can change.

Everything from a weapon, a licence plate, a mark, a scar: if they are being used as part of the evidence, I must ensure that I can rely on how that image was created.

I may need to look at other videos or images not related to the case but assist immensely in the integrity stage of comparing and contrasting data to ensure that I understand the generation of the file.

For many years I used several different pieces of software throughout the analysis chain. I knew the limitations of each and understood when to stop and reconsider my workflow. If that is not done, if people continue to cut corners and do not handle the multimedia with care, evidence will either be lost, misinterpreted, or used incorrectly.

I never enjoyed using many applications. I found the writing up of my workflow between each one too time consuming and, in many cases, quite complicated. Some software would deal with time, some would deal with total frames. Ensuring the frame counts matched the timing, and vice versa, could often be a little tricky.

Shortly before the end of the Forensic Video Unit (there are those cuts again), I started to take an interest in a product by Amped Software. Amped FIVE, Forensic Image and Video Enhancement, wrapped up many of the challenges I faced into the single application I had been looking for. What was even better, was that they wanted to implement many of the ideas and workflows that I had developed over the years directly into the software.

You can understand why then, when I had to leave the Police Service after the closure of the unit, I jumped at the chance to work with the Amped Software team. The software aligns with my own beliefs on video and image analysis – using science to ensure what can be done, is capable of being done.

At the end of the day though, my part in an investigation is just a small cog in a large wheel. I must rely on many other people, and trust that they have done everything correctly and the video is what they tell me. If it’s not, and I pause to ask more questions – it just slows that wheel down. Although I must create further ‘presentation’ material, there is a clear path back to the original exhibit.

The Criminal Justice System in the UK is at breaking point. The disclosure of digital evidence, which includes CCTV and digital multimedia, is daily headline news. If the material is handled with care, using diligence and competency, then there shouldn’t be a problem. If corners are cut then, well, that’s when it can hit the fan.

By David Spreadborough, Amped Software

Share this post