Opinion: CCTV evidence and the need for a universal viewer

sira-newsdeskPrepared by Steve Lower & Robin Stevens, Sira Defence and Security Ltd, and Garry Parkins, 3G Associates Ltd

If you believe what you see on the TV, solving a crime is so easy.  A crime is committed in full view of a CCTV camera, then law enforcement staff can easily and simply retrieve the footage, and with no difficulty watch clear high quality images from every angle, and armed with the evidence, the suspect confesses, the crime is solved and the criminal dealt with by the courts.

That is the theory but the reality is that law enforcement staff, throughout the Criminal Justice System (CJS) face a myriad of problems, from retrieving the CCTV footage, to being able to replay it both for the police, and other parts of the CJS, before getting a conviction using CCTV evidence.

This problem was recently highlighted in a report by the HMIC called “Stop the Drift 2” The report states in section 4.37 the variation in formats used, the lack of available equipment to convert the CCTV evidence into a compatible format for use by the police, CPS and courts, the lack of suitable equipment to play the CCTV back, were consistently raised by forces; and the inability to send a CCTV clip across the CPS/police interface because the clip file size was too big.”

The holy grail of a universal viewer was developed by SiraView in consultation with law enforcement staff and Centre for Applied Science Technology (CAST), to address the problems of the multitude of formats and the need to provide a standard easy to operate user interface with the functionality required by the majority of law enforcement staff.  Many forces use it and recognise the benefits. Many other parts of the CJS, including the CPS and courts do not use a universal viewer which is surprising given it is a cost effective solution that ticks all the boxes.  To support defence lawyers, SiraView has developed a very low cost version.

The current position

As the UK National CCTV Strategy set out “CCTV plays a significant role in protecting the public and assisting the police in the investigation of crime”

But, CCTV evidence can only be of value, if law enforcement agencies have the relevant skills to extract and playback seized footage from the many and varied systems that are available in the market.

The difficulties encountered in recovering and replaying CCTV footage often lead to significant delays in dealing with crimes.  This is usually because of the high volumes of CCTV being seized, or because law enforcement agencies are trying to find ways to view the footage.

As recent as the Boston bombings has shown, there has been an indication that it took analysts many crucial hours to trawl through hundreds of hours of footage from the many CCTV systems that were available.

Traditionally specialist police forensic departments have carried out this activity with trained staff but as austerity bites and the number of technical staff is reduced or as importantly not increased in line with the volume of CCTV to be viewed, bottle necks are developing.

Should forces not be looking at a triage protocol where potential evidence is viewed quickly, and locally, by officers or civilians who may not be experts in CCTV, and where the expert technical staff are left to deal with the more difficult tasks?  However, a force wide viewing system also lends itself to viewing vast amounts of footage quickly, particularly in the event of a major incident, something stand alone solutions would not do.

Using independent 3rd party laboratories, moving CCTV discs around by courier or in police vehicles all add to the delays, are inefficient and costly.  Using less technically competent staff without appropriate equipment, often downloading decoders off the web without expertise and training, can run the risk that evidence may be lost, damaged or subject to legal challenge.

It is clear that a new approach is needed to overcome many of the technical and cost barriers that prevent law enforcement staff from reviewing key evidence to assist in speeding up the detection of crime.

The problem falls into two main categories, retrieval and replay.

Retrieval

The first problem is retrieval.  It has been the case in the past that law enforcement agencies will often employ specialist technical support to retrieve and process digital CCTV footage. However, this is less likely in the current economic climate, where often untrained staff are required to retrieve and replay and process CCTV footage.

This leads to significant problems not only with format variation, but also the ever growing need for significantly improved quality.  Exporting video footage from various CCTV systems is often a lengthy task as DVR storage grows; because it is now very cheap to store months of High Definition CCTV footage on a hard drive, which represents terabytes of data. The law enforcement staff that go out to recover this footage often end up seizing hardware as exporting the footage would take too long.

This whole process of knowing how and what to seize in order not to compromise the integrity and continuity of the evidence is a skilled job.

The United Kingdom National CCTV Strategy said There is inadequate training in place for all staff engaged with CCTV. There are currently no uniform training standards that apply to all. It has been identified that the proper training for all users of CCTV is crucial to its successful deployment and effective use.”  Law enforcement organisations would be advised to have training from specialist companies who design and deliver Home Office courses in the Best Practice methodology.  There is often only one opportunity to seize the footage so it is essential to make sure it is done correctly.

Replay

Even when the footage has been successfully retrieved, it is often the case that the wider Criminal Justice System (CPS, lawyers and courts) have difficulty in playing back CCTV; each different system will require the manufacturer’s software to be installed on a PC and staff training to understand it, before footage can be played.

Why do we have so many formats?

The challenge for the CCTV industry has not been as simple as choosing an existing media format to use. Being digital and easy to create a new format we may be currently looking at several thousand different variations and it is very unlikely that no one single format will dominate the market.  Many manufacturers have therefore created their own propriety playback systems without the realisation that this would become a problem for the law enforcement agencies that have the sometimes impossible job of trying to view all this footage.

The term proprietary refers to exclusive methodologies used to maintain commercial control over the DVR format and prevent third-party copying of a particular file.  Some proprietary DVR systems offer more than one export format, such as the native format as well as a standard container format such as AVI however some manufacturers are reluctant to offer anything other than their proprietary format.  The reasons cited by manufacturers is that they want to maintain control over the file by making it conversion proof vary, but most companies will make “tamperproof” a sellable feature, while disregarding the work required by law enforcement staff to view the video.

Most observers would agree that CCTV systems have been developed in a piecemeal fashion.  There has been relatively little national or international coordination and agreement in terms of strategic direction and regulation of formats.  It has tended to be the case that CCTV was the preserve of relatively small and medium sized companies, and it has only been in recent years that the big multinational companies have come into the market; so we have not to date seen the dominance of a few big companies and the corresponding decrease in different formats.

Where are manufacturers going with formats?

At the present time the situation is almost certainly getting worse.  As CCTV manufacturers push the complexities of algorithms to encode the video images, whilst adding new functionality, this ultimately leads to new variations of existing formats, adding to the list of formats in use.

To illustrate this Sira invited manufacturers to tell us what formats they currently work with.  We asked the question “Based on the equipment you manufacture, distribute or install what transcoded export format standards do your DVRs currently use?.  As a guide we have highlighted the results of the survey below.

H.264 (MPEG-4, part 10)72.2%
MPEG-4 part 2  Information technology50.0%
JPEG27.8%
MPEG-227.8%
JPEG 200013.9%
Formats not listed19.4%

 

It should be noted that some manufacturers support more than one format which is why the percentages add up to greater than 100.

We also carried out an analysis of all the examples which have been sent to the Sira laboratory by users within the criminal justice system, and the results were very similar to the information supplied directly by the manufacturers.  The results presented indicate that there is a move towards using H.264 based image compression.  Clearly this is only a snapshot and a more accurate finding may be derived if a larger more up-to-date database had been interrogated and a wider selection of manufacturers had responded to the survey.

So why not adopt a single format based around H.264?

“The National CCTV Strategy  suggests that one way to establish digital CCTV standards would be for Stakeholders to agree on a standard digital video format”    Developing a single format may be an answer but is  highly unlikely,  as already there are several thousand different format variations developed by hundreds of manufacturers.  It may also be the case that if all the manufacturers were forced to use a single format it would stifle innovation.  Even if a small range of formats could be adopted by the industry, the standards are not usually prescriptive in the implementation so there would still be variations.

Even those manufacturers that adopt a common protocol like H.264 our work on writing decoders has not found any great degree of commonality amongst manufacturers with regard to metadata formats which although common and predictable e.g. : time stamps, channel numbers, image information, etc they are not presented in a common way.

It is our view that establishing a common metadata format is an essential part of the CCTV data standardisation process, if reducing the number of formats is to happen.

So what are some of the current viewing options and the issues around them?

  1. The DVR’s manufacturers own decoder and viewer

A lot of CCTV exports are viewed on a CCTV player supplied by the manufacturer which is loaded onto the viewing computer at the same time as the export.   The big advantage is its free and you know that you have the viewer appropriate for the exported footage.

However, some of the issues working with imported manufacturer’s players may be familiar to you:

  • Importing manufacturer’s players onto the IT network is not usually allowed.  Computers are locked down to stop this happening.
  • Different manufacturer’s players may interact with each other and other critical software when loaded on the same computer which may call into question the evidential integrity.
  • Each manufacturer’s player has different functionality and may require staff training to use it correctly.
  • Not all manufacturers’ players have the correct export functionality to interface with court room presentation software packages.

The player will often be written for the computer operating system relevant at the time and is unlikely to be future proofed. The legacy support for older CCTV systems results in the law enforcement agencies, maintaining older computers to enable access to these files, due to incompatibilities with more modern operating systems such as Vista, Windows 7 and Windows 8.

      2.      Downloading manufacturers own viewer from the internet

A lot of exports do not include the player.  It is not feasible or cost effective to source the relevant player; in a lot of cases downloading off the internet is not possible.

Nearly all the problems listed above still apply but in addition you have to identify the exact player from the file extension and hope that you download the right one.  In some cases manufactures may have several different players all dealing with the Meta data in a different way; so how can you be sure of the evidential integrity?

This is not an approach that would be sensible to adopt if you did not have some expertise.

     3.      Purchase stand alone hardware with manufactures players loaded on

On first inspection it looks a good option but you would have to purchase additional stand alone hardware at several thousand pounds a time loaded with players you can download for free off the internet. You still have all the problems of different functionality, questionable evidential integrity etc set out above.  Can you bundle up hundreds of different players and expect them all to work?  If you have tried loading up different players on a single computer you soon realise that some interfere with each other bringing evidential integrity into question.  The hardware is often not scalable for use on a network deployment.  Export capabilities within the manufacturer’s players vary widely and can result in the end user screen capturing the imagery; this can be a lengthy and expensive process.

Maintenance

Each player may need to be separately updated (or downgraded to find a version correct for a specific example). The Windows operating system also needs periodic upgrades which will have to be administered manually. If this system is connected to the internet, it may not benefit from the usual corporate IT safeguards such as virus scanners, and so could be at risk of infection. CDs exported from such a system may also be infected, thus risking the introduction of a virus to main corporate network.  Legacy players can result in requiring the use of legacy operating systems such as Windows 98 / 2000 this increases the complexity in standalone environment requiring dual boot or installation of virtual machines.

    4.      Use a free player like Windows Media Player, VLC and GOM

By chance or design many formats are viewable in Windows Media Player, VLC and the GOM Media player all designed as “free players” to view entertainment type video streams. This is far from ideal.  These players often strip out valuable data like the time and only play one camera at a time.  These players are designed for playback of broadcast video, and will drop frames as necessary to maintain a 1:1 playback speed.  You cannot be sure you’ve seen every frame of evidence.  They are not really designed as players of CCTV and the evidential integrity is questionably.

So if we cannot agree on a single format and manufactures players have different functionality and questionable evidential integrity what’s the answer?

The law enforcement agencies that are ultimately the end consumers of CCTV footage need a low cost efficient way of playing CCTV. Some of the requirements might be:

  • Easy to use and navigate
  • Plays everything on any computer
  • Deals with multiple camera streams
  • Preserves the integrity of the footage and associated meta data
  • Affordable to implement throughout the organisation
  • Future proof, without additional expensive hardware maintenance
  • Quickly scalable in the event of a mass viewing of footage being required

The solution therefore needed to utilise the existing Windows based hardware commonly available throughout the organisation, without the need to deploy further custom hardware which needs a specialist to operate.

Listening to the police and staff at the Home Office it became apparent that what the law enforcement agencies needed was a universal CCTV viewer that was easy to use and preserved evidential quality.  It should be capable of dealing with thousands of CCTV formats as well as having all the additional attributes that a free player cannot offer i.e.

  • Support for multiplexed video streams
  • Access to original recorded time and date information
  • Support for variable frame rate video
  • Correct aspect ratio
  • Audited handling of data, to preserve evidential integrity

It would enable proprietary formats to be viewed outside of the manufacturer’s software, which often don’t play in the free players widely available.

The universal viewer

SiraViewA universal viewer SiraView was developed by Sira Defence and Security, in consultation with law enforcement staff and Centre for Applied Science Technology (CAST), formally the Home Office Scientific Development Branch (HOSDB), to address the problems set out above and to provide a standard user interface with the functionality required by most law enforcement staff.

SiraView enables users to view many different CCTV formats using a single, easy-to-use user interface without the need to use each of the manufacturer’s proprietary viewers. It preserves the evidential quality of CCTV footage.

As a result users can view footage directly on their PC rather than having to wait for the footage to be sent to a laboratory for conversion into a standard format.  This can often mean a suspect can be shown evidence and charged immediately, rather than being bailed to return when the footage is viewable.

Product methodology and independence

Sira has written its own software to do the decoding rather than producing a ‘wrapper’ for the CCTV manufacturers’ proprietary software.  This preserves the evidential quality of the CCTV – an important consideration for law enforcement use.

Crucially, the fact that Sira Defence and Security Ltd does not manufacture or sell CCTV system hardware of its own means that it is ideally placed to take the universal viewer initiative forward in a neutral, non-threatening, way in partnership with the manufacturers of CCTV recording equipment.

Manufacturer support

Many manufacturers have been particularly forthcoming with support; several have produced SDK (Software Development Kit) which will allow SiraView to play all of its current and legacy formats.

The practical support of major DVR/NVR producers is undoubtedly a pivotal moment for the project and hopefully will result in other vendors following this lead.

Many manufacturers see SiraView as a landmark initiative because they are committed to ensuring – against the backdrop of a growing and disparate CCTV infrastructure – that vital footage can be accessed and reviewed in an effective and timely manner by the police and other users.

Events such as the Boston bombings, and the London riots, have served to both underline the evidential value of digital video evidence and also, crucially, the challenges faced by the authorities when having to sift through thousands of hours of video footage. Consequently, manufacturers believe that a universal viewer like SiraView which can ease this process by delivering one common viewer interface has to be applauded, and is why they support our work.

Benefits to law enforcement staff of a universal viewer

Current procedures often involve law enforcement staff travelling to their nearest video labs to seek help in recovering and viewing evidence obtained from the scene of a crime.  This process is costly in officer’s time and associated opportunity costs, which may require a return visit to the lab to review the footage at a later date.

How much more cost effective would it be if the seized footage could be viewed locally by a wide range of officers on a local computer without incurring travel costs and wasted travel time?  Speeding up the retrieval and review of CCTV and digital camera footage will ultimately lead to time and cost savings with the right process and tools in place.  Officers would not need to be familiar with hundreds of different players with variable functionality and integrity

In many cases officers can obtain an early plea by showing key evidence at time of interview saves times and money.

This approach also lends itself to integration into the growing development of digital case file systems and repositories where relevant clips in the original format may need to be viewed by many staff with the original master disc in secure storage.

Benefits to the courts

Courts can view the footage in its original form.  Court staffs that are not generally technically skilled will become familiar with the common functionally and will be confident in its use.  Complex cases containing CCTV from multiple systems can be easily reviewed, speeding up court proceedings.  Using the original footage without conversion to another format maintains the integrity of the evidence.

Conclusion

In order for the Criminal Justice System to continue to win the fight against crime, whether complex major crime or low level volume crime, they need to have access to the best possible evidence that can be collected quickly and efficiently, and with no risk of the loss of integrity of that evidence. That is the solution that SiraView and 3G Associates offer. Using the SiraView software and understanding how to recover the CCTV footage quickly, will equip the Criminal Justice System (whether police, CPS or the courts) investigate more crimes and bring more criminals to justice quicker.

Leave a Comment

You must be logged in to post a comment.