Behind the scenes of all Amped Software programs that handle video files there is a powerful and unique conversion engine. Although quick and simple in use, it hides an extraordinarily complex process that is required when dealing with proprietary surveillance video. David Spreadborough from Amped Software breaks it down for us.
We must begin this article with the CCTV and Video Surveillance System (VSS) manufacturers. Unfortunately, they do not make things easy and, as such, worldwide investigations dealing with video evidence is hampered by undocumented and proprietary encoding methods. You would initially think that as they are recording evidence, they would adhere to open and transparent methods of video capture but alas, that is not the case.
Every company can do their own thing. They can encode and format their video, their audio, and their date and time information, in any way they choose. Most operating guidelines for system selection and usage reach the point of searching and playing the footage – not many continue to investigation, analysis and court presentation. This then explains why the Amped Conversion Engine is required. It picks up where the CCTV manufacturer stops – at the evidence export process.
Amped Software currently has three products that support DVR/CCTV video files (in addition to standard video files and images):
- Amped DVRConv – A simple automatic batch converter for proprietary video files
- Amped Replay – An enhanced video player for converting proprietary video files and enabling basic corrections and presentation capabilities
- Amped FIVE – Forensic Image and Video Enhancement. The most complete set of tools required by analysts to convert, analyze, interpret, restore, enhance and present image and video evidence.
Your role and requirements will dictate what application you need but the same conversion engine is in each one. Every time you update the application, the engine gets updated at the same time.
Let us now look at what goes on with some CCTV exports.
Here we have a single file with a .drv file extension and a directory named ‘Player’.
The .drv file is large in file size so I’ll guess that’s the video. The file extension of .drv should be a clue but not when we are referring to CCTV. Manufacturers can use whatever extension they want to, and a few of them don’t even use one!
So, for one CCTV export this could be the video, but for another, the .drv file may be a program extension. The same extension can also be used by another manufacturer for their video, but the encoding and format could be vastly different. This also goes for the naming convention. Here we have ‘[000001]’. Another company may use the same convention but a different extension. Or vice versa.
With this export we do have a player included in the directory. However, most players are not designed for the needs of an investigation and may not present you with the footage as it was recorded – yes, that really is true!
Many players also need installing and it is common knowledge amongst the forensic video community that installation of any product relating to CCTV can often be fraught with difficulties.
Let us see then how the Amped Conversion Engine handles this file, firstly using Amped DVRConv.
Once you have your settings configured the way you want them, DVRConv can be a ‘set and forget’ application. This means that whenever I drag files into the main window, I know what it’s going to do. I can change these if needed, for splitting or joining files for example.
After dragging in my .drv file, the software does its thing.
- It has identified 7 streams of video. Now then, this is important: these were not written into the file as 7 streams. They were in the file as one single stream! A standard conversion would look at the file and would not understand it. A more detailed converter may allow the user to manually select the contents encoding type. In that circumstance, the result would be a single video stream of corrupted video frames. The Amped Conversion Engine has identified each video frame and sorted them into their respective stream. For this file, that took a second per stream.
- After the clean streams have been created, they are each placed into a container. I use MKV, but others such as MP4 could have been selected in the settings. Each new MKV formatted file has the stream number, and then the letters sc next to converted. SC standing for Stream Copy. The video has not changed since it was in the DRV file.
- If a Timecode is identified for the streams, a .time file will be created for each stream.
- A File Information and Conversion report will be created in .txt format.
Take a look at the resulting output directory.
You have the original file, the clean streams, the newly playable and correctly formatted files and a .time file for each one that can be read by any of the Amped video applications, and finally the File Information Report.
The magic bit is the creation of the clean streams. Whenever a format like this is received by the Amped Support team, it takes considerable effort to interpret the data and create a decoder for it. This decoder then goes into the Engine. What could take days to understand and manually carve, will now take mere seconds. Sometimes this process can be very quick and easy if the customization is quick, or it can sometimes take more time and effort.
DVRConv is also available as a customized version for easy integration into Digital Evidence Management Systems, where it will automatically detect uploaded files and will convert behind the scenes, so your evidence is ready to view.
Let us now move onto Amped Replay and see how the Conversion Engine works here.
Users either have the simplicity of drag and drop again, or can use the DIR BROWSER under the IMPORT tab.
As the video is not immediately decodable, the word Proprietary is displayed in the frame viewer window. With a standard video, the first frame would be displayed.
To get the camera’s identified, cleaned and formatted correctly, it is a simple case of double clicking the file and then the engine starts its work!
Within a few seconds, the files are automatically converted with no unnecessary user interaction required.
The Detected Cameras are then viewable with the buttons under the FILE INFO.
The image looks quite different from how it would be played in the player that came with the video. This is because we are displaying it how it was recorded.
Under the enhance tab, these issues can be corrected.
Lastly, and if required, the detected timestamp can be placed anywhere on the screen using the text annotation under ANNOTATE.
So, we have seen how the Conversion Engine works in DVRConv and Replay, how about the powerhouse of the Amped family, Amped FIVE?
With FIVE, the user gets full control over the conversion process.
Upon initial loading, the Convert DVR box will be displayed giving the user all available conversion options.
In this example, we only have a single file but if we have several of them, we could save considerable time by either converting them all, or even joining them together if they were all of the same camera, but had been split on capture or export. This joining process is known as concatenating.
There are many configuration options here, from selecting the container / format to separating the audio.
When your specific requirements are all set, a single click of the OK button puts the Conversion Engine to work.
Within a few seconds, the video streams are all loaded into individual chains and each one has the date and timestamp preloaded.
All the videos have been placed into individual chains, and each one has their time file pre-loaded.
Within a couple of minutes, I can process all of these and get them ready for presentation within a completely forensic process. You will notice I have placed all my chains into a reference folder, I have applied some filters for correct display and finally completed a Multiview for court.
Of course, if I wanted to now work on some of the camera views and individual frames to restore or enhance individuals or objects, then this is again a simple process of just copying the camera to a new chain and getting to work.
One single file, but the Amped Conversion Engine makes it simple to view and manage the 7 cameras within it.
It really doesn’t matter what you start with and what you need to do. The Engine takes care of the video structure and original formatting and will prepare it for you with the least amount of fuss.
It is only right to finish this article by looking at the two types of files that cannot be converted.
The first one is a file that we have not encountered before. For us to decode the file, we need to customize the converter and validate it. Amped users send us new files every week and we provide upload facilities in all our products.
The final one is the closed, proprietary file. These usually have some form of encryption and can only be played back in the proprietary player. They are the most damaging types of files as they cannot be authenticated independently and, often, the players designed for this type apply some hidden image processing that causes issues in forensic analysis and enhancement.
I remember, only a few years ago, spending nearly a week, decoding, and carving a file to read the raw data. Thanks to the users that upload new files, the developers to write the decoders, and the Amped Conversion engine, this forensic technique is completed in seconds and with complete transparency. The good thing is that this is available, in different forms, at all levels of an organization: the conversion can be set centrally by dedicated technicians with Amped DVRConv, handled in a safe way by investigators and frontline officers with Amped Replay, or by the analysts with Amped FIVE. Everybody has the same issues, but with different needs and skills, and the different products available make it easy to tackle these issues in the proper way, on every side.
For more information about the Amped Conversion Engine or any of our Amped Software solutions developed to assist your organization to analyze images and videos for investigations, please contact us https://ampedsoftware.com/contacts