David: Hi everyone and hope you’re all okay. It’s the afternoon here in the UK, late afternoon over in Italy and going to be evening time over in the East and then morning over for you guys in the States and Canada. And whatever time it is, if this is going on YouTube, so hello to you too. Investigating Video: The Vital First Steps.
Those final two words there, “First Steps” give you a hint of where we’re going on this webinar. We’re not going to be doing a massive deep dive into anything in Amped Five, we’re not going to be looking at any super fun workflows in Five or anything like that, we’re going to be going right back to the start and looking at those vital first steps.
As Michelle said, yes, my name is David Spreadborough. If you’ve never met me before, I’m a certified forensic video analyst through Lever and I’m the forensic analyst or one of the forensic analysts at Amped Software. A little bit about my history. I was a police officer in the UK for 24 years. The last 12 years was purely dealing with CCTV evidence and forensic video analysis. And then in about 2015, I moved out from policing and went to work with Amped Software, originally as their international trainer, but for the past few years, I’ve been sort of dipping my fingers and toes into various pies and mainly now concentrating on improving the workflow in Amped FIVE and doing things like this, which is great. And a lot of research and development, as well.
Okay. So, the big thing though, is that it’s not just a few of us, we are now a pretty big team scattered around the world. In the US, in the UK and mainly Italy, obviously, being an Italian company, but also other parts around Europe. The whole point is that we are a big group of people that are extremely passionate about what we do and extremely passionate about improving image and video forensics within our community and improving the education and training and knowledge within our niche sector of forensics. And yes, we develop the software relating to it, but education is a massive part of everything that we do and it’s really important to us.
So, video, where do we start? Well, I mean, I think that says it all to be honest, because a lot of the time it is pretty mind blowing. I still get fascinated by some of the things that we find on a daily and weekly basis with video, especially from a proprietary and modified point of view. So this is where companies have then done other things to a standard in order for them to facilitate whatever they need to do, usually within the surveillance industry.
And it is a pretty mind blowing subject and there’s not a lot of information out there from within our community. Yes, there’s a lot of the technicals and there’s a lot of things to do with standards, but from within our community, within law enforcement and intelligence, policing, security, there isn’t a lot of that information, so it can be pretty mind blowing. And the mind is the first thing that we need to look at. When we’re looking at these first steps, it’s having the right mindset.
There are mainly two types of people: there are the ones that are looking forward, and then there are the ones that are looking back. And it’s very hard sometimes to change that mindset, especially when it’s new information, and it can be quite scary. And if you’re looking backwards, you’re sort of saying, “Well, this is what I was just told to do.” There’s no questioning of what’s going on. There’s no questioning of, “Well, is it actually right? And should I be doing what I’ve been asked?” Because the person that may ask you to do something may not know some of the technicalities involved. If you’ve always done it in a certain way. It’s like, “Well, that’s what we’ve always done.”
So you may be in an environment where it’s, “Well, we need to create some procedures. Well, we’ll just create procedures that are based on what we’ve always done or what we always know.” There are the other people and then there are the ones that are looking forward and they form and test theories. And it’s like, “Well, what’s going on here? Is this the reason that this is happening? Why have I got information perhaps in a proprietary player? And then I haven’t got that information in something else, or I’m getting more data in Amped FIVE than I would do if I put it into this player or this analysis tool.” What are the differences and why are there differences?
Have you got forensic rigor? So is it forensically sound, some people would say? So can I backtrack all the way from what I’ve created, all the way to how it started on that DVR or MVR? Have I got a forensic pathway reversing all the way back to how it started for integrity? And you’ve got a fact-based workflow, rather than being subjective, and it’s like, well, that’s what we want to show, it’s the, well, these are the facts. And if you stick with a fact-based workflow, then it’s very easy when you’re further down the line in the courtroom, because everyone would agree, that fact can be proved. There are no issues. It’s reliable. So, having the right mindset is one of the first things.
Second step, really, is having organizational responsibility. It is now massively important because video and digital multimedia evidence is in so much of a law enforcement organization. And all the way from identifying where that footage is to acquisition, storage, handling, viewing, processing, the analysis, all the way to then the presentation of it, there needs to be some organizational responsibility. You can’t cope with that yourself as one person or one unit, because there are so many other connected strands to it.
There are some places that don’t have that. And then rather than sort of getting bogged down with the correctness of it, they’ll miss out some of the processing and the analysis, and then they’ll go straight from storage and handling. There’s no one involved in acquisition, so it’s just like, “Well, can you send this to us? Can you send the footage to us?” So the footage isn’t acquired forensically and then it’s just viewed online; the processing is done online according to whatever capabilities are in that system. And then it’s just presented and then there’s no analysis of it, either. So you’ve got missing gaps in that workflow. Having an entire organization look at it from and have responsibility for digital multimedia evidence, whether it be an audio or video is going to help you. So that’s another vital first step to work towards having that overarching responsibility.
You can imagine if you didn’t have that, you think of a car in a car factory, is that when it comes out, it’s got everything working, it’s got four wheels, it’s got a steering wheel, it’s got everything that a vehicle needs. But there have been various different departments in the construction of that vehicle. And if they’re not talking together and getting everything right, the vehicle is not going to work when it gets to the end. That’s exactly the same, especially with CCTV is that you want to avoid the problems at the end by making sure that that piece of CCTV evidence starts off correctly and goes through the entire criminal justice system correctly.
So a bit of a wrap up on organizational responsibility. Overcoming video challenges is not simply down to a video unit. Managing the acquisition, the handling, and the processing of DME is an entire organization’s responsibility. And a video unit, especially you, and I do feel for you, especially some of you in the smaller video units out there in the world, you can’t investigate that media properly and correctly if the data that you’re getting is already compromised because of how it’s been acquired, how it’s been handled, how it’s been processed, and you can’t do your job if the person beforehand hasn’t done their job. And that’s why you need that sort of responsibility.
Bit of a reminder there to myself, and I think Michelle is going to help me out here. Martino recently wrote the video evidence principles and then published that and presented that to the European parliament. It is a fantastically written document. I’m not just saying that because it’s Martino, the CEO of Amped Software, if you don’t know, but all of these issues he’s wrapped up in a very, very understandable way. And the first part of the video evidence principles is all about organizational help. So it’s not about the technical side of things, which is you and me in a video lab, it’s the organizational side of things. And so the document is split into two. So I’m hoping that Michelle would have put that in the chat and you will see the link there for that document. It is really worth a read: share it, use it, send it out to your bosses, send it out to your colleagues, it’s there for you to use.
Step three in these vital first steps is controlled acquisition. It is so important and it is getting more and more important now. In the last 6-12 months, we are seeing more and more issues. It’s not especially to do with fake: has it been manipulated? Has it been changed? Has it been edited? These things are now in more people’s minds. We’ve always had issues with regarding, well, has it been edited and have these things been done to it, but the ease that it can be done and the ease that it can be sort of hidden is now getting much easier. So controlled acquisition and then the controlled integrity of that evidence is vital.
We haven’t got time today to go into all the issues surrounding acquisition. So last year, I think it was last year. It might’ve been the year before last, we did a blog series on CCTV acquisition. I think Michelle is going to help me out here as well and put a link to this in the chat. We start off all the way from the basics. What is CCTV? And we go all the way through all the issues surrounding CCTV acquisition from searching for it to just going out and collecting it and recovering CCTV to the big issues of today, which is the public submission of CCTV and video evidence.
Mrs. Smith of the corner shop has no idea about video evidence. She has no idea that her system can export into five different formats. Which one is the correct one? Which is the one that’s going to give you the right timing information? Which is the one that isn’t going to be transcoded? She doesn’t know that. And why would she? And so putting that over to the public is a massive danger and we have to deal with that and understand that. So have a look through that blog series. Again, it’s there for you to use and it’s on the Amped blog.
The benefits of controlled acquisition are quite a few. One of the big ones is the overall speed of the workflow. I spoke to someone a few months ago and they said that now a lot of their detectives are just pulling drives. And so even for one hour from one camera, the video unit just gets a hard drive and their backlog is huge. So they’re being told to go out and recover the data forensically by extracting the hard drive, but that’s then putting so much work onto someone else. And so, the detective’s job is quite quick, of pulling that hard drive, but then the video labs unit is then increased. So, by having the control all the way through and by managing that acquisition correctly and ensuring that, well, okay, I’m going to make a decision on-scene to acquire this correctly using this method and validate it, then in the video lab that’s their workload reduced. So you’re balancing out the workflow and you will speed up eventually. Yes, you do have times where you need to extract that hard drive, it’s completely normal, but there’s an imbalance often.
You’re going to reduce the risk on integrity and authentication. If there isn’t a controlled acquisition, then it is, well, has it changed since the time it was created? Is it a true and accurate representation of that which it purports to be, which is authentication? And I’m not just talking about deep fakes there, I’m talking about the times when timing isn’t authentic, the aspect ratio isn’t authentic, color might not be authentic, that is what is authentication.
And you can then correct footage to become authentic in certain circumstances. You can correct the timing if the timing is off in your acquisition. But if you can control the acquisition, you can then go back and say, “This is a fact, this is what happened.” And if you have a look at the blog post, you’ll see some of the ways that you can do that, or the blog posts.
Also, it dramatically increases your chances of restoration and improved enhancement. If you’re starting off with your best evidence, which has gone through no changes whatsoever, then your chances of being able to correctly restore and enhance that to answer a question is going to be dramatically increased. Because the moment something changes in that process, like a transcode, then you’re going to reduce the quality of that data, and you’re then going to reduce your chances of that license plate being recovered; you know, a good shot of the logo or the face or what the telephone number was on the front of the phone that was captured from the CCTV camera in the lift. Yes, we’ve done it. But any slight change, you’re going to reduce your chance. So having a controlled acquisition and starting off at the best evidence is your best chance to make all of this so much simpler. I know it’s difficult and I know we haven’t got a million people out there doing this sort of stuff, but it is something to work towards.
If you haven’t and all you’re left with is something that has been sent in by someone and you’ve got no idea what’s happened and you’ve got no idea what has gone on, what the device was, how it was acquired or anything like that, put it in your notes, put it in your report. Be open and transparent, like the sentence there says is that, “I have not received or been made aware of any documentation surrounding the video recording device, the acquisition of the received data, any processing of the data before submission as evidence, or the individualization of the data through a cryptographic algorithm.” So you’ve got to then start that by doing the hash at your moment in time, because you’ve never been given that before.
And there’s a keyword in that, and that is data. What you are getting is data and you get the .abc file, the .whatever file from such as such a CCTV system until you deal with that and you manage that correctly, you’re just dealing with a big chunk of data, and the video is going to be in there, the audio may be in there as well, timestamp especially may be in there and you may have other data as well, camera information. But it’s just data. It is digital data. Deal with it as digital data, right from the start. You can’t go wrong.
Step four: standard operating procedures. Let me just grab a quick drink. So, the vital first steps. Step four is having good SOPs: standard operating procedures. They don’t have to be massively structured, and that’s probably not a good thing either, but it’s a good idea to get some consistency in a unit. I’ve been to some places, luckily I’ve been to video labs everywhere, and there’s often quite a big difference of, “Well, I don’t know about this, or I can’t do that, because Johnny’s not in the office.” and that’s what you need to sort of try and balance out from.
And a good standard operating procedure should help with that and can bring up the team’s competency in dealing with video evidence. This is something I put together a few years ago and then Martino sort of jazzed it up a little bit for the video evidence principles. And it’s very very simple. You have to get those three things aligned. You have to be starting off with the best evidence. Remember on the previous one: controlled acquisition. So you’ve got to be starting off with that best evidence. If you’re not starting off with the best evidence, it doesn’t matter how good you are, it doesn’t matter if you’re using the best tools in the world, you’re never going to be able to get the answer that you need.
And Martino has put here, sorry, hitting my mic, the forensically sound result. So, same tools, you could have all the skills in the world, and you could have the best evidence, but if you haven’t got the right tool, you’re not going to get that answer.
So, having a good SOP to make sure that you are controlling the acquisition and the evidence, you’ve got good skills and knowledge across the board within that unit, and you’re using the right tools, that’s going to give you the most and the best opportunity. And by having those standard operating procedures, you’re having that forensic rigor, you’ve got that structure, you know that you can go all the way back and all of your evidence throughout the whole of your chain is going to stand up to scrutiny, is going to stand up to questioning and other people, especially I know a lot of you are going to be within law enforcement, from the defense sort of standpoint, I can say, yes, that’s been done correctly, that’s been done correctly, I can understand what the person’s done there, I can understand that decision. There’s nothing to question. I do a lot of case reviews and when it has, and I can see everything structured and it has got forensic rigor, there’s nothing to question. It’s perfectly fine.
The other thing is then having a good structure within that standard operating procedure is having a good structure. Now, this is a bit of a modified workflow from another procedure that’s UK-based. This is my modification because in the official one, it doesn’t have these two very important points, it’s been missed out. These are the two important points and they are massively important. I can’t sort of stress this enough. A working copy has to be a direct bit-for-bit copy of the original. A working copy can’t be another version of it, like a screen capture or a transcode or a clip or whatever you call it.
A working copy is a direct copy of the master and you can hash validate that. Very, very simple. After you’ve got your work, so this is after you’ve done your acquisition, after you’ve done that, the other box there, Point 9, Produced Generated Exhibit is a huge box. There is a massive amount that goes into that part because that’s then, okay, here are my files that I’ve recovered. These are the questions that I have. These are the tasks that I need to do. How do I get to that point? What do I need to do to that multimedia to get to the next point, to answer the question or complete the task? If it’s just a timeline of a person moving between systems and cameras, well, how do you deal with each of those exhibits? How do you deal with each of those items? So it is a huge part and unfortunately it’s missing from this document. That’s why I’ve added it in.
So let’s look at that one part of that generated exhibit. Here it is. Here’s the sort of rough workflow. And it’s broken down into four main parts: analysis, processing, presentation, and then your conclusions. Your conclusions may just be the completion of the task, because it may be as simple as “Can you tell me what the frame rate is?” And so you may go directly from analysis, all the way to completion of the task, but you’ve got to do that analysis. And a lot of the time, the analysis is missed, unfortunately.
You can break down the analysis into three parts. And for those people that have seen any of my webinars or been on any of my training, you’d have heard me say this before, but exhibit data, and then visual and oral. Exhibit is, who created that item? When did they create it? How did they create it? All the questions surrounding that item of data, and that could be a huge pot of data. It could be a hard drive, or it could be one file, and then everything in between. What is the data? And you may need to do some data extraction in order to be able to work out what that data is. Is it a set of ABC files, or is it a set of MP4 files? Are they standard MP4? What’s the codec? What’s the frame rate? What’s the resolution? It’s the data that’s going on there and you’ve got to analyze that. Are all my files the same? Do they have different frame rates? Are they variable? Are they constant? All the questions.
And then, what are you looking at? What is my visual representation? Can I see that there are some effects in the video that I need to correct? So, and sound as well, especially if there are car tires screeching away or gunshots or anything like that. And, are there any sync issues as well? Very, very common.
So, that analysis is going to help you with your integrity and your authenticity stage and answering those questions for integrity and authenticity. Then you’ve got the processing. So you’ve got then the restoration and the enhancement, and you’ve got to do that within the image generation model. Image generation model is then reversing the errors in the image. So if you can reverse the errors, you’re going to get the higher result and you’re going to get a more accurate representation. i.e. it’s going to be more authentic.
And then obviously measurement, but you can’t measure something until you’ve done the restoration. Think about aspect ratio, think about some of the more modern now HD and QHD and 84k and 8k that are actually getting recorded at half-width and then the player or the system will then expand that. Sometimes there’s some issues there. So, until you do the correct restoration, you can’t do the measurement.
Then we’re moving on to the presentation. What are you then going to create? And this is your new exhibit, so these are your generated exhibits. Think about the previous workflow. So it could be an image, could be a series of images. It could be a video, it could be audio, and it could be data, it could be a spreadsheet. All of these things as well, how am I going to present that data in order to complete my task or answer the question?
And the question may be, where did the person go? Is this person this person? Whatever the question is, it could be that, well, is that reliable? Can I rely on that clothing representation? So can I rely on that clothing representation? Can I rely on that clothing representation? So if I can rely on the clothing representation, is there anyone else in the area that is wearing similar clothing representations?
So that’s your conclusion, and then you would have the report. So this is your forensic video analysis workflow, but it’s going to be not just one exhibit, is it? It could be two or three or four or hundreds, depending on the size of the case.
But think about that standard operating procedure. And if everyone is doing the same sort of thing, then you’re going to not have the risk of people missing things. And so having a good departmental and personal workflow will help you because it does turn into muscle memory. And you do the same thing every time and you look for the same issues. And the more issues that you find, you remember those issues, and then you look for them in the next one and the next one and the next one. And there are always issues with proprietary video.
So on a basic evidence trick, you may have something like this. Just checking the time, okay, we’re a few minutes late, but we’re running all right. Basic evidence trick. You’ve got your initial exhibit, then you’ve got your working copy, and as I said, it could have four DAV files. I’m just using DAV here as a bit of an example, because I think most people sort of recognize them as being a slightly proprietary video format. But that creation of that working copy, and then perhaps then the creation of another copy that you’re then going to be working with, depending on what your internal structure is, there’s the Copy & Verify tool within Amped Five. I will briefly show the Copy & Verify tool. Again, we haven’t got time to sort of go into it too much, and we’ve dealt with that under a couple of other videos and blog posts, as well. But we will have a brief look at it.
So there’s your Copy & Verify tool, so then you’ve got your working set of files that you can then work on. Now, we’re never going to overwrite those files, but it’s a good idea to have sort of a backup of a backup. So you’ve got your master and then you’ve got perhaps a working copy sat on a server somewhere and then you’ve got your files that you’re then going to work with on your workstation.
Then comes the Convert DVR bit is, well, okay, from the DAV, what have we got? And we could then end up with a standard and the SC on the Convert DVR bit, the SC is for stream copy. So I’ve just copied the stream, and then we’re going to have a log file for that process. We’re going to have a log file for the formatting process.
So we’re going to have four new video files, we’ve got four new log files, and you’re probably going to have four new time files, as well. And you could have four new audio files all linked with those. But you’ve got a pathway back to your working copy of your DAVs. And from your working copy of your DAVs, you’ve got a pathway back to the master. And you should have then some documentation on how that master was created, where and when, by who, etc, etc.
All the way at the end of this workflow, so all the way at the end of the evidence tree, you’ve got what you’re creating, so what you are then presenting, and they’re going to be different exhibits, obviously. So, you might have then one MP4 video because you just need a video and it’s going to be half of perhaps MKV2 and half of MKV3 that you’ve then timelined together and done some stuff to it and some annotations and whatever you need to do in order to answer the question or complete the task.
And you may then have a PDF image sequence as well of whatever incident it is and it’s different from the video, so it’s going to be a different exhibit. And then you would have a report dealing with all of those.
But there’s a box in between these stages, and it’s a big box, and in that is a billion things. And we are forever learning. I am forever learning to say, well, this is the question, this is the task, this is what I’m starting with, this is pretty easy, you know, you can get all that sorted, this is what I need to do at the end. Well, how am I going to get from that bit to that bit? And you will never stop learning. And it can take a little while on that initial stages of your video investigation to just say, well, I’m just going to have a look at that. And I’m just going to have a look at that. And well, If I’m going out to a video and I need that video to be displayed in a courtroom, for instance, well, I’ve got 4k video and I may have this camera and this camera.So I’ve got a 4k video here and a 4k video here. Well, how am I going to deal with all of that? That’s a lot of data. Do I need to crop? Do I need to resize? If I resize, would I then need to do a magnify? There’s all these questions.
And those initial stages of your investigation is that you’ve got to then figure out your workflow, and that’s where your competency comes in and your training with the software, and the time. It’s not a quick and easy job. It’s not something you can just say, #Okay, I’m just going to do this.” A lot of people would then just take the video down from a 4k to full HD. But in doing so, are you then affecting the question that you’re trying to answer? Think about dark clothing at night, reflective objects, perhaps under IR. And then by reducing that, you’re limiting the evidence that you’re trying to get out. So you’ve got to think about the question and the task at hand, and then how you’re going to deal with that evidence.
Remember, authenticity: is it authentic? Is it authentic? Is it authentic that I’m showing this? And has my processing actually affected the result. So you’ve got answer that.
Okay, tight. So, over to Amped FIVE, but I just want to sort of wrap that lot up is that having that structure and having that base knowledge and having the base questions really right at the start of the investigation will help you. And so, well, okay, this is what I’m always going to do. Always going to do this, and we’re going to do this way, I’m going to structure my working folder on my system. So I’m going to structure my directory in a certain way and this is how I’m going to work. And it makes things easier because then you’re not questioning that each time; you’ve got a good workflow and it will work. Everyone has a different way of doing it, but in general, it can be a good standard structure to start off with.
Okay. Quick drink, and let’s see if we can, how are we going to do this? There we go. Escape.Here we are in FIVE and anyone that’s been on any of our webinars or watched any of our things before, this is usually where things go wrong.
All right, first thing is, I tell you what, let me just go to “Details” here so we can see that a little bit better. Yes. And first thing is, we just saw a couple of thumbnails. So, first thing is, in your system, I’ve got a standard video contained here, look, I’m seeing the extension, remember you can turn extensions on and off in Windows Explorer, so I’m seeing the extension, .avi, that one’s a .mp4, and then we saw the thumbnails, which would suggest to me that Windows and DirectShow is being able to decode that video information.
So again, first things first, we’ve got an AVI and we’ve got a big long file name here. We can see we’ve got NVR and then an IP address and then a name, and then we’ve got these numbers here: 2019, 0319, et cetera, et cetera, et cetera. Okay, I’ve got a pretty good idea that that’s going to be perhaps the time to and from. This is really important if you’re using some automated system for members of the public to send in that video evidence. Make sure that those automated systems are not changing the filename or structure in any way. This is forensic evidence that someone is submitting that hasn’t probably been acquired in a very forensic way, but let’s not get bogged down with that for the moment. But this is forensic evidence and so it shouldn’t be changed. File names are vitally important. And so make sure that if there is some system in place, make sure the file names are being retained because it can help you.
Let’s just drag that into FIVE for the moment. Dragging it into FIVE brings up the video loader. Drag & Drop in Amped FIVE is different from loading a video. You can see we’ve got “Load video”, okay? Drag & Drop does some checks and we’ll look at some other things for you, as well. So there are different ways of loading video into FIVE for different reasons. Try and use Drag & Drop or any of the automated processes more than you can because we do a few checks and we then can take over a few things. We’ll always give you a question. If we need to do something, we’ll always ask you. So always consider Drag & Drop first.
I’ve loaded the video in and we can see some information, we’ve got visual information. The first thing I’m going to do is just see if I can scrub. Yes, I can. And you know, everything is looking fine. I just got some notes on my other screen here to keep my brain in gear and I’m now going to just open this up in “Advanced file information”.
So we’ve got this information here, I’ve got the video. The first thing I’m going to do is, well, okay, what am I dealing with? We’ve got, If I just close my assistant for the moment and then go to “View” and “Tools”, we’ve got basic file info here. So I can see that we’ve got some basic file information here: it’s h264, 25 FPS and the amount of frames. But remember, you’ve got the advanced file information here. First thing I’m going to do before we go into this is just start my frame analysis. Because that’s going to take a few minutes and then we’ll come back to it.
All right, so we’ve got visual information, I can see the scene, I can see when we’re talking about authenticity, is it where I think it is? Think about Google Maps, as well. I use Google Maps a lot. So yes, that matches, especially if you can’t get to it. And is it night? Is there any timestamp? Is there a pixel-embedded timestamp? And when we’ve loaded it, have we extracted any data timestamp? And then if we’ve got a pixel-embedded timestamp and we’ve got a data-embedded timestamp, and if we’ve got the timestamp in the file name, is everything matching? Is everything saying, “Yes, okay. It’s the same time. That’s the same date and time. That’s the same date and time.” Or are there differences?
If we have a look in “Mediainfo”, where are we going? Yes, Mediainfo. You’ll see that we’ve got our Mac times, we’ve got file creation, modification, or let me just put that there for the moment. Now, some of you are probably thinking “Oh my goodness, what’s this “Missing Coded Pictures” and when did that come up?” I am using a new version of Amped FIVE that is coming out very, very, very soon. So, I’m going to be giving you a bit of a sneak preview to some of the things that are coming out, including this bit of analysis here, which is really handy. So there we go. You’re privileged to see some of this. But going back to Mac times. Look, we can see here that we’ve got file-last-modified time. And sometimes you get encoded time, as well. And if a player is using this as the start time and then the duration of the video in order to create a timestamp, this is saying that it was on the 2nd of April, 2019, 8:26 in the morning.
But if you have a look, and you might not be able to see, but if you have a look at my file name, it’s the 19th of March, 2019, so a few days before now. So if you’re using a system or if you’re using a player or another video tool that is using this to create a timestamp, it’s probably going to be wrong. So just be careful of that. Again, it’s the power of file names.
Yeah, one thing that we find a lot is network, when something gets written that first file that gets written on the PC is because the PC client CCTV system has said, “Okay, I want this bit of data” and then it brings the video data over using a sort of a network transport protocol, and then it gets containerized on the computer. And so that’s when the file thinks it’s been encoded and written. Well, it was on the DVR a few days before. So always be careful of those dates and times and how it’s been acquired. And if there’s any, I’m sure there’s going to be some collision guys out there, but that’s another thing that we’re seeing about a 50-50 change in that 50% of the time, the original timing data is retained when that network access has been done and 50% of the time it’s original. So it’s keeping the original timing structure. I know, it’s a minefield, isn’t it? Remember that emoji? Head blown.
So here we’ve now done our frame analysis, and we’ve got some issues. We can see that we’ve got pts, not applicable, not applicable, not applicable. What’s all this going on here? AVI doesn’t allow for that data. The timing information is in the container, it’s not in the stream, it’s in the container and AVI doesn’t allow for that. So if you’ve got one of these raw streams that are in an AVI, this is why you’re going to see “not applicable” here. Because it’s not in there, it’s not in the container. So it hasn’t been able to be read from the container.
Another thing to point out, and it happens a lot in proprietary video systems, especially over a network, is you do get missing coded pictures. Something has happened in the creation of this video file. And look at this, we’ve now got 51 missing coded pictures. I’ll now do a check for you and tell you, and then you can copy this data out if it’s going to be relevant to you for you to make a decision.
So, always have a look at that data analysis because it could affect your timing and it could affect, do we need to go back out and try and do another acquisition if it hasn’t been acquired correctly, et cetera, et cetera. Excuse me, hit my mic again.
All right. Let’s just drag in another one. I just want to show you. So, you’ve got your video and you think, I wonder if it’s been transcoded? I want to show you something that I use a lot, and we can automate this now. In Amped Authenticate it has a filter in the video module, it has a filter for looking at this automatically. So if you’ve got a really big video and it’s very, very slight, having it done automatically in Amped Authenticate makes it so much easier. You may not be able to see, I’m hoping that you can see the sort of like, I call this like a parquet flooring effect here. And you can see it here and I’m just going to hit “Play”. Oh, what’s going on here? There we are. There we go. So I’m just hitting “Play”, you can see the little black specklies that are going on. Alright.
What I’m watching here is a transcode of this. So let’s say for instance everything looked as if this was original. Remember in our analysis stage, we looked at the exhibit, we then looked at the data and then we looked at visual, this is one of the things that I look at from a visual perspective. I’m looking at these little bits here and I can see, and you may be able to see it, there’s a bit of a sort of a flash, and it goes blurry, and then refreshes, and then blurry, and then refreshes. Oh, there’s a cyclist, but we’ll ignore them. And then there it goes. Well, that’s your GOP structure. And you can see it change to an I-frame.
Do you see the change at the same time as our frame changes to an I-frame? Does it match? If it doesn’t match, that’s your biggest clue that it’s been, so I’m just going to go to change. There we go. So there we go. So that’s a change. And that is our I-frame. But it’s shown as a P-frame and this is because this is a transcode, it’s been transcoded, it’s been changed, it’s not the original. And so your chances of getting a license plate, your chances of then finding other very, very small pixel-level detail will be reduced because of that transcode. And because of that transcode, have you lost timing information, et cetera, et cetera, et cetera. So your three stages.
All right, let’s move on. I knew this was going to be tight. Again, people know me and I can waffle for hours. Where are we going? Next one. Okay, just dragged it into Amped FIVE again. Remember that sort of stage. Oh, hang on. We need analysis here. What’s going on? If you want to, you don’t have to go directly to Convert DVR. You can have a look at the file first. And look, you can see that we’re just analyzing the file as it is and I can see some information.
I can see that this is the size of the video, I can see that this is the codec, I can’t do any frame analysis because it can’t read how many frames are in there because it’s in a proprietary container, but I can read some of the information that’s going on inside. If I go to ffprobe, I can see that we’ve got a lot of unknowns, not a lot of detail.
And one of the things, and it came up in a question, I wonder if it’s going to be in this. Yes it is. It’s not as bad. It came up in a question in support I think it was earlier on in the week, and this probe score, you can see you’ve got a probe score of 51. If that was 100, then everything is as a standard: we’ve got a standard container, we’ve got standard codecs, and everything is structured according to the world of standards. No problems at all. Yes, probe score 100%. 51 is in the middle. And if you get something like a raw h264 stream, which has got a few little sort of modifications in it, you’ll probably get a probe score of something like 1. Because it’s thinking, “Well, I think it’s h264, but I’m not quite sure.”
That’s what probe score meant, and it only came up in my memory because as I said it was a support question earlier on in the week. But we’ve got some information here, and I think, okay, I know the sort of things that I’m looking for. I’m now going to go to Convert DVR then, and I’m going to open it up. I’m going to put it into a MKV container and then I’m going to press “Ok”. And I’m going to close that for the moment.
So, we’ve done a process in order to get that in, and then this is what’s now in my file, or with my file. We’ve got the clean video string. So we’ve cleaned the video out and that is just the raw video string. I want to highlight the file size. The original .dv4 was 79,450 kilobytes and then the clean is 64,739 kilobytes. There’s a bit of a difference there. Have that in mind, consider these things, what’s going on. Then we’ve got the .time, so that’s all our timing information. We can deal with that, as well. You can see it’s loaded automatically for us as a timestamp. Then we’ve got our converted MKV, which then our clean video stream placed into a standard container. And then we’ve got a log of that process.
If you contact us through support and there are issues for whatever reasons, we may ask you to send us the original file and the log of what’s gone on, just so we can have a look and just make sure. And you will see from the log that it says well, success, yes, that’s a good start, but you can see what it’s taken in and we’re using ffmpeg, no problems here, but we’re using ffmpeg to place the clean into a container, so when we chose the MKV container.
We’re not going directly from dv4 to MKV, we’re doing a process in between, and that is the cleaning, and that is again so important and I’m going to show you why in a moment.
If I then just close that. Let’s drag in the dv4 again. And this time I’m going to “Attempt direct loading”. And you’re probably thinking, “Oh, this is new. I haven’t seen that before.” It’s another new thing. Yes. And this is so you can then say, don’t worry about any cleaning, don’t worry about anything else, if you direct-loaded it, what would happen? So if you direct-loading it with FFMS, then what will happen?
And as you can see, I’m now reading the .dv4 with FFMS. I haven’t got the timestamp because we haven’t done that important cleaning process, but we’ve got more frames. The original had, where are we? Down the bottom here. 11,379 frames. The dv4, 11,394 frames. Well, those few extra frames don’t don’t account for all that data. Remember the data size? So, what was going on there? Let’s deal with the data size first. There are always reasons and it might take you a little bit of time to find out the reasons but there are always reasons to stuff.
Here’s my clean file. Here’s my direct-loaded file. You can analyse any of those, and if I went to my clean file up here and I went to the original file, I can analyse the original file as well. So even if I didn’t have to load it, I can analyse it there as well. So, here’s my clean in my MKV. I can right-click that and I can go “Advanced File Info”. The buttons are everywhere for the reason so you can do whatever you need to do whenever you need to do it. So that’s the reason why you see all of these things and you see Convert DVR boxes everywhere and Advanced File Information boxes everywhere because you’re always thinking and there’s always questioning, well, what’s going on there?
So, here’s the MKV, and I can see all my hex data here. And let’s do it from here, so let’s do it from the original file, so Advanced File Info. So now we’re going to open up the dv4 and then I’m going to go “Hex View”. Okay. So this is the dv4 file. I’m just going to scroll down a bit and I can usually spot the big bits. Let me just make it a bit bigger. My eyes are going to go in a minute, I can see. Where is it? Let’s scroll down a bit more. Oh, there we go. Oh, wait, let’s go down a bit more there. There we go.
Look, see all this data here? It’s throughout the whole file. What that is, I’d need to do some more digging. I don’t know. Have I got a dv4 player? Is that a sort of a holding space for audio? Did the system have capability for audio? Remember that controlled acquisition? Has anyone taken a picture of the back of the DVR?
Has it got a mic input in? Has it got any capability in the system in the GUI for recording audio? But throughout this entire file, I’ve got these massive areas of this, and I’ve also got a data timestamp in there, remember? Our data timestamp has been extracted. So, that’s obviously affecting the data, but how is all this affecting the video footage, as well? And none of that is in the MKV, by the way. It’s all clean.
But what about these extra frames? Let’s just close that. So let’s go to our directly-loaded video. How are we doing on time? Not too bad. And I’m just going to scrub to the end because I know there’s a good time to see it. Oh no, I’ve missed it. Oh, there he is. All right. There we go. All right. Look at this. And remember the advanced file info for missing coded pictures. Well, if we do an advanced file info on our clean file now, on our MKV, it’s going to have some missing coded pictures because it hasn’t brought across the damaged frames.
Those damaged frames could be important to you for timing information. If you’re dealing with timing of a vehicle down on this street, you may need those frames. But everything is there, and whether it’s missing coded pictures, or what the differences are, or what you’ve dealt with for whatever reason, all the information is there for you to use.
So, our MKV with our timestamp, that is clean. There is no damage. Everything is there, and in the Advanced File Info, there’ll be a list of what are the missing coded pictures, and it’s because of that damage. And that damage is probably caused somewhere on how that data has been structured with that information.Just as a bit of a point for you, in the player for this format, you don’t see these frames. The player cannot present those frames because of the damage. So you see exactly the same as our plane. But if you play the dv4, you actually can see the damage that is there. So there are pros and cons, you see. And that is the reason why is that, well, okay, I’m going to do the clean, I’m going to get the timestamp, and I’m also going to look at this ,as well. Because you never know.
All right, anything else in that one? This is just briefly tell you, yes, look, if you haven’t seen that before, that’s interleaving, so you’ve got the top field at the top and then the bottom field at the bottom. And so what you’ve got to do is you’ve got to interlace them back together again and then deinterlace and then deal with the timestamp because you’re dealing with a data timestamp as well and then deal with the timing on the frame rate, as well. And it’s all possible within FIVE, but we could probably spend about two hours just on this file alone.
All right, next one. Let’s just close this. Have a drink. All right. So, we’ve dealt with a pretty standard file, we dealt with a couple of standard files: the avi and the mp4. We’ve dealt with a single proprietary file. Now we’ve got an executable. So this could be a file that has lots of cameras in it. And it could be an executable as well that has the player in there and all the video streams as well. And these executables, some of them are pretty good. This is not too bad. Some of you may recognize this as being a lot like this sort of IDIS clip player. it’s branded as many, many different things with all sorts of different capabilities as well. Some do some very weird things, some not too weird.
There’s always some weirdness going on, but anyway, you’re limited by what the player is then giving you. So you want to override that and we want to get to the video that’s inside this executable. So we can just drop that in and again, it’s going to come up, “Do you want to deal with the data that’s inside it?” Yes. Going to choose MKV again, and then I’ll quickly say why I’m going to use MKV, excuse me, after I had a drink.
MKV in practice is probably the most common in flexibility. MP4 has some restrictions; AVI has many restrictions, I don’t tend to use AVI unless absolutely necessary. MKV seems to be the most flexible. Sometimes MKV doesn’t work and it needs to go into MP4. You really have to be flexible. You’ve got to have that questioning approach, not that, “Oh, it just doesn’t work so I’m going to screen capture.” It’s a, well, okay, why can’t I do that? What’s going on here? What is restricting that? And learning from it.
So you can see now that we’ve loaded our four video streams. And these are the four video streams that were in the player. Now we can deal with that. So let’s have a look and see what is now in our folder. And we’ve got the original executable and we’ve got our clean file, remember, our video stream and we’ve got our .time and then we’ve got our working MKV and then we’ve got our log file. And you’ve got these for all the streams.
Now, this is probably a good time to explain that we’ve written these into the same folder that that was. We haven’t overwritten that, but we’ve written them into that directory. Remember what I was saying about with Copy & Verify and perhaps having a backup of a backup? And it’s a good reminder for Copy & Verify. Copy & Verify here. Copy & Verify allows you to select a source, So where’s your master? So that could be a USB drive, could be a network drive, could be wherever, external hard drive or another drive. That’s your source. Then you’ve got, well, where do you want it to go to? And then where do you want to save the log of that process? And then you want to copy the source to the destination and then compare it. And then you’re going to get a log file as well and that will be where your log of that process goes, with all your hashes, and you can select your hashes here.
And then what type of process you want to run So you can either have, well, I want that directory and all the sub-directories and all the files in it as well, or don’t include any of the sub directories, or just have a single file. So if you’re just hashing a single file and copying a single file, you can choose that there as well.
But a lot of the time, it may be two processes because someone may bring in a hard drive and say, “Okay, I need to take that hard drive back with me to my department. Can you copy all the data over?” Yes. Do your initial copy and then do another one to perhaps your NAS or your internal server, because then you’ve then got a backup of the originals that you’ve got, and you’ve got a process of that. So if anything goes wrong with your system, you know how dodging some of these computers can be and you get a hard drive crash or some corruption or whatever. is that within your department, you’ve still got a forensically sound working copy that you can then go back to.
And so that’s your Copy & Verify process. Then we were talking about your folder structure yourself. When you’re dealing with an investigation right from the start, have an idea of what your folder structure is going to be when you go out. I have my reports, which will be a PDF and that’s my analysis report, and then I have another folder linked to that report, which is linked documents, and then I have everything in my linked documents. That is everything is in that folder: all the files I started with, all the files that I’ve created, all my log files, PDFs, image sequence, everything is in there. And so everything stays the same because, remember, when you start using hyperlinks and you’ve got links in your fire projects as well, it’s all relative file paths. So as long as you keep your working directory the same, you’re never going to lose any of those links. If you start moving things around afterwards, it’s just going to be a headache for you. So just try and make it simple for yourself.
All right, so we’ve got our executable here, we can see all our files. What was I going to show in this? We were going to look at sort of a public release of an image very, very quickly. Let’s just drag this down. So our bad guy comes in and, let’s say, we wanted this just for clothing. We can’t really see his face because it looks like a bit of an alien, right? But I think he’s wearing ski goggles or something. Anyway, and if you want to just deal with that, very, very simple, and I could probably keep this here and I don’t really have to deal with much of this, but I’m just going to drag my video loader down. I think, what are we on? Stream seven. Yeah, so I’m just going to drag my video loaded down just down here I’m going to say “Yes”, I’m going to just rename that and put that “Wanted”, oh, can’t spell. And I already saw what frame I was on. There we go. And then one of the things that I think, it takes a few seconds of, I know everything’s been dealt with correctly from the start, I’ve got my full acquisition, I’ve dealt with all my files, I’ve extracted my image here, and now you can do something to it.
Very, very quickly. Edit, where are we going? Can’t see. Rotate. There we go. it wasn’t badly rotated on the original, but a lot of the time they are, and I see them in the press all the time. It doesn’t take very long to just hit the rotate button and just get the person straight. Otherwise you end up trying to do this with your phone or whatever. It’s crazy. So, a bit of a rotation, a bit of a crop, and let’s just make sure that’s nice.
And if you notice, I’m still dealing with the whole of the video here, so I haven’t selected just that single frame yet. And then just deal with that. A lot of the time you see stuff that’s really washed out, you can put in the levels and do perhaps some colour work, and it takes less than a minute sometimes to come up with a much better image for your sort of wanted. And then you could put some text over the top for whatever reason. But we haven’t really got time to go into that because we’re running out of time. So I’m going to leave that, but super quick and easy.
and then I could stick that in a folder and I could say, “Okay, well, that was my wanted image.” Done. Finished with. Then, when they get identified, and then you’re going to going out and preparing all the video evidence for after his arrest, etc, etc., then you go on to something else.
So hit “No”. Where are we? Always goes super fast. I’ve got another two examples to try and get through. It’s never going to happen, is it? All right. Let’s see. Nearly there. Quick time for a drink.
Okay, here we go. So, here now we’ve got our full folder structure and obviously there would be another folder here, which would be the wanted bit. Oh, that was the bit that I started off at and then you can then carry on working. And we can see this is from the gas station. So there could be other robberies here. I could have gas station and something else and something else. I could have four or five incidents here because then you can start linking those together and say, “Well, this was incident 1 on this date. This was incident 2 on this date.” Do a timeline, etc, etc.
If you’ve got a big timeline and you’ve got quite a lot of incidents in various places, I tend to use one Amped FIVE project for one location and then another Amped FIVE project for another location. So I’ll do my analysis and all my restoration and enhancement measurement, whatever needs to be done in that Amped FIVE project, then I’ll write out videos from that. And remember, in FIVE, you’ve got the FFV1 codec. So it goes into an MKV container and it is lossless, it’s 100% lossless, but it takes about a tenth of the file size than uncompressed would. And it’s scrubbed super fast, renders super fast as well. So it’s a brilliant format and container for then saying, “Well, okay, that’s going to be my interim file from this location. That’s my interim file from this location.”
You can then have an Amped FIVE project for all of your interim files to create your final timeline. So rather than having a massive history of exhibits from all over the place, I tend to tidy mine up and say, “Well, okay, this is this location project, this location project, this location project.” And I’ll usually run that in something like a spreadsheet with some notes as well and stuff like that. So if I’ve got a big incident on the go, that’s how I’d do it.
And you can see here that we’ve got our originals here and you can see that we’ve got file information and the hash code of our cameras. And then under the restoration enhancement, you can see we’ve got aspect ratio, then an undistort, then a resize, because I’m thinking about how I’m going to display this information. So what do I need to resize it to? Automatic colour equalisation. Then we’ve got the timestamp, but look, I haven’t rendered it on the screen. You see, I’ve gone “Disable rendering”. So I haven’t displayed the timestamp there, but I’ve got it as a data source to use.
Then I can choose my range and then all the way at the end, you can see now we’ve got the timestamp over the top, and then we’ve got our file data here, so that’s just for that camera, and then you can just copy them down. You don’t have to do each one the same, don’t forget these are all the same cameras. Again, controlled acquisition, you’ve been out to the scene, you know that it’s all the same cameras, all the same settings on the DVR, et cetera, et cetera. So it speeds up. This takes no time at all because you just Copy & Paste down. No worries.
Then you can join all of them together. Where’s our multi-view exhibit? Let’s just close that. DS1, there we go. And you end up with something like that. And you’ll see as he comes in, how are we doing on time now? Not too bad. You’ll see as he comes in, all of these, because of the work that you’ve done in sorting out each one and restoring and enhancing each one. Where is he? There he is. each one is now accurate. So let’s just leave that there. Okay.
And even though, that’s frame number 1452 from that file, this is frame 730 from that file. But look, you see it’s the same moment in time. These frame numbers are really important because if that is being questioned, then everyone is talking about the same frame from the same file.
I get some compilations sometimes and you’ve got no idea how that compilation has been made. Nothing’s frame-referenced, nothing’s referenced back to the file. I know that that’s the file that has been used, that’s the frame number, and I’ve got a document that explains where that file came from and the integrity of that file.
So, there we go, and you can see the timing, so we’ve got what, 533 on that one, and then, oh that’s 533 as well. 566, there we go, so we can see now we’ve corrected all the timing and it’s all in-sync. Remember, it can’t take the same image. Some of these systems, they can’t take the same image at the same moment in time and then record. So there’s very often you’re going to get that millisecond difference.
So it does take a bit of work and this is just on the same system. Don’t be worried about having real difficulties and having to explain that some things aren’t frame-accurate. Because if you’re using different systems with different time bases, different frame rates, different recording capabilities, getting them all synced up is nigh-on impossible. So you have just got to explain perhaps that you may have a one-, one-and-a-half-second error rate in your syncing, but you have then done a visual sync at certain points, and just explain it. It’s perfectly natural because it’s really hard to do. Okay. So, wanted person all the way through to court and presentation.
Let me just, new project, do you want to start a new project? No. Let’s see how we’re doing. Let’s see if I can do these. Writing. Oh yeah, that’s a bit of a shame. Now, I tell you what, I’m just going to mention this. The test card, which is here I’m hoping that Michelle will put the link to the test card on. It’s really worth it. This is one of the Amped test cards. I am in the middle of building another one. These colours and the greyscale ranges, you can validate yourself because each of these individual components have been created. So, If I go back to here, you can see as I’m hovering over here, I can see that we’ve got a 255-0-0, so my pixel is saying those colour ranges, that RGB value, sorry. And you can see, yes, 255-0-0. So 0-0-255, yes, 0-0-255, I’m just reading over here.
My greyscale ranges, I can see, yes, we’ve got 38 luminance, yes, 115 luminance, yes, 217 luminance, et cetera, et cetera. So, this is all in an Amped FIVE project, so then you can then write the file and you can then write the file like a MKV, FFB1, completely lossless. And you can confirm and validate that it is completely lossless. And then you can have a look at some of the transcoding options. MP4, h264, and the difference in the values. And then you can see those values change. If I just quickly load up another one into this, let’s go test card MP4 rather than the uncompressed MKV. And you’ll see now, and this is what I’ve used, and I used the default MP4 h264 values, you can see now that yes, okay, we’ve got good luminance, so our grey scales are all fine. If I go over red here we’ve got 254-0-0. Well, that was set for 255. So we’ve lost a tiny little bit of red: one value out of 255.
Down the bottom here, 180-16-16, 180-15-15. So we’ve got a little bit of a change here. So you can see what the differences are, and then remember video mixer. Whenever you’re doing testing to say, “Well, okay, I’m always going to go out to this”, remember video mixer and your similarity metrics and have a look at that.
Video mixer, remember, “Link” and then “Video mixer” and then have a look at your similarity metrics. So then you can then identify how different certain things are, like the original to what you’re presenting in a court. Yes, there are going to be some differences, but is there a dramatic difference, or is it just a very slight difference? If you need that data, it’s there for you.
How are we doing? We’re nearly there. Frame timing. All right. I’m going to quickly show you this and I’m going to do it in hopefully less than five minutes. New project, no. Here we go. I want to show you, oh, yeah, don’t worry. That’s because it’s got audio, but I’m in a virtual here which hasn’t got an audio. So, I’m just going to turn the audio off so I don’t see that all the time. There we go.
Now then, when we’re dealing with frame timing and frame analysis, remember, you’ve got to sort of compare what data you’ve got. I’m just going to go for PTS playback down here. Remember there’s several different ways you can play back video. There’s what, I mean, six to seven different possible ways that durations and frame rates can be calculated. You’ve got to verify which one is the most reliable. You may not be completely 100% accurate, but which one are you going to use and which one is the most reliable? And make a decision on that.
If I go to Advanced File Info now, you’ve got lots of different ways that the data is being presented because the different tools have different capabilities on reading that data. Think about Mediainfo and ffprobe, they’re using the same libraries at the back end, but they’re reading those libraries differently. So you can end up with slight differences, or you can end up with pieces of information in one, but not in the other. Sometimes ffprobe may be able to read that there’s an audio stream, but Mediainfo won’t detect the audio stream. Things like that.
Timing information. What is the frame rate? It’s variable. I’m just going to go to the frame analysis, press “Yes”. Obviously this is an MP4 now, so we have got PTS data within the container and it’s just sort of churning through.
We do a number of calculations now, so it takes a few minutes to churn that information through because then it’s retaining that information at the back end. Remember the coded pictures, as well? But, what I want to show you is this.
First of all, I’m going to go to Mediainfo. You’ve got to sometimes see and report on differences. Now, I’m going to show you something that I haven’t got an explanation for yet, because I only found it a couple of hours ago, and I’ve been using these files for ages, but the more you look, the more you find. And format settings, GOP, Mediainfo is saying 30. There are 30 frames between each reference frame, 30-frame GOP.
If I go to GOP analysis now, M1, n=60. It’s a 60-frame GOP. What is it? Everything in this file is saying that it’s a 60-frame GOP. I haven’t got time to now go into the macro block analysis, difference analysis, and even the hex data for the h264, everything is saying that it’s a 60-frame drop. Why is Mediainfo saying it’s a 30-frame drop? I have not got the answer, but I am going to try and find out. I’ve got some clues, I’ve got some hunches, but anyway, these are the things, and these are the things that you find. And it’s like, right, okay, is that an issue? Do I just need to document it for my investigation or does that affect my investigation? Does that infect the integrity? Has it been transcoded? Has there been something going on in this? So as it’s brought the data over, perhaps it was originally a 30-frame GOP, and as it’s brought the data over the network, it’s turned it into a 60-frame GOP. I don’t know, and so we’ve got to do a little bit more digging.
Now that we’ve done the frame analysis, I can just turn off the audio and the other frames. And you can see that we’ve now got, and it’s going to be coming out tomorrow, we’ve now got some calculations going on, PTS duration computed. So what it’s doing is it’s saying, “Well, what’s the difference between that one and that one.” And there’s my frame duration. Remember, you can display all of that over the top of your screen. One of the quick ways of doing it.
Oh, one last thing, two minutes, is don’t ever get tricked into thinking that packet duration is the same thing as PTS duration. It’s not, even though it’s the same in this one, sometimes your packet duration will be based on your time base, and if you’ve got a constant time base, you can have the same value there, but you’ll have differences in your PTSs. So just be aware of that. So, if you’re doing any sort of speed work or motion body mechanics, use of force, things like that, always be careful on exactly what your frame durations are.
Remember we were talking about presenting all of this on the screen? One of the things I have, I’m just going to close that and then go to “Assistant”. This is only in my virtual, so I’ve just done this just to sort of show you. I have a quick-text assistant and you can see I’ve just got this quick text here and I can just grab this and you can create these depending on what you want. So if you are a collision investigator, you can say, “Well, okay, I want to have this information.” You can use this in Annotate, as well. Obviously, in Annotate, you can have it frame-accurate. So you can say, “Well, okay, I want a countdown between this frame and this frame and have other things going on as well.” The world is your oyster, to be honest.
I’m just going to copy that and then do an add text. So this is one of the things that I would do straight away if I’m looking at a file. I’m just going to then do an Advanced File Info and then whilst that’s churning perhaps I’ll do this. And I’ll go, “Presentation”, “Add text”, just going to ctrl+v and then I’ll stick that left top and then we’ll do a back colour text. And now, I’ll tell you what, let me just do a range so you can see, let’s have alt+i for my input and then when the vehicle goes out, we’ll have that as out, then my add text, and now you can see my sequence number there. That’s my original frame number. That’s my file.
The sequence time. So I’m four minutes into the sequence time, not the original time. Ah, CCTV time, I didn’t put the CCTV time in. You can use add timestamp or you can load the timestamp time, but if you haven’t got any time within the data, then you can add it in if you knew that from another source, but you can then put all the details on that.
And the macros are, there are just loads now, and we’re always adding macros to give you all the information that is available, all the data is available, you can present that on the screen and you can use that data in any way. So you could use countdowns, chronometers now, and all of that information.
And it all goes into saying the final point of this, it’s all for the purpose of you being able to explain that you understand the video. I understand the video, I analysed the video, this is the data that I found in relation to the questions and tasks that I was being asked, this is the information, this information is all reliable. If you’ve had to interpret the information, how you’ve done that, then you’ve got your presentation, your PDF image sequence, your video, your data, your report. Done. And it proves straight away, the moment someone can read that, it says, “This person knows what they’re doing. This person has competency.”
And so those vital first steps are ensuring that you can prove your competence in understanding the video and understanding its limitations. And especially with CCTV, there are all sorts of limitations. You look at the GOP structure there on that one, why is it showing twice? There’s something’s gone on somewhere and it might just be a fault in that system in how it’s presenting that data to Mediainfo, but we don’t know. Thank you. Bye bye.
