Digital Evidence Video Frame Rates
Digital evidence video frame rates are an often overlooked topic when discussing the tools used to capture digital evidence. From body worn cameras, to interview room cameras, and everything in between, and beyond, we forget to evaluate devices based on their Frames Per Second (fps) rate, and from the start of any new endeavor, handcuff our agencies to the limitations of what those devices can handle when capturing data. In this article, we will explain common fps rates available on the market through a wide range of devices, while discussing what the pro’s and the con’s of those frame rates entail.
Our Brains and Motion Blur
We’re going to get into a little science, before we get into a little more science, so hold onto your chairs.
First, it’s important to understand that our eyes, and our brain, cannot process visual information in frame rates, like cameras, phones, tablets, and computers do, and thus we can’t quantify in frames how much visual information our brains can process. What we do know is that our brains have a sensitivity, or rather limits, to the amount of visual information it can process successfully before the brain starts assuming things, or it develops a headache like a migraine, to get you to look away, depending on what it is you’re viewing. When it comes to visual standards, we evaluate our vision through use of the Snellen chart, and most people with normal vision are dubbed to have “20/20” vision, meaning those people can read a letter from 20 feet that most people should be able to read at 20 feet. There are some people that have 20/10 vision, meaning they can see the same thing at 40 feet, that the average person can see at 20 feet. Later, we’ll explain why this context is important.
Now that we’ve discussed some basic information concerning human sight, let’s discuss what happens as the eye tracks motion. So, as you might have already guessed, as the eyes track motion, they’re naturally drawn to that motion, except for something that is well established as accepted in that person’s mind, in which case they catalog the motion and rarely revisit it. But as they are tracking that motion, be it an object, a person, an impending natural disaster, if the motion becomes too fast, too vivid, or an overabundance of information for our brain, it will inject motion blur so that we don’t suffer headaches. In fact, two common points of visual information in the consumer market that have had to be corrected because of their headache-inducing content, are the modern generation of racing games, that have had to inject artificial motion blur at the sides of their racing courses, predominantly seen in the grass, or “out of bounds” areas, and in the first-generation models of HD televisions. In both cases, they were so efficient at processing visual information, that blurring didn’t occur, and people would report feeling nauseated, sensitive to light, and even experiencing temporary shut downs of memory, particularly short-term memory. Luckily much of these situations were in test groups, so no consumers were hurt in the process.
The reason we’re discussing the human ability of sight, it’s ceiling, and its floor, is to give you an idea moving forward about how the “average” eye is going to interpret frame rates that are produced from the devices you use to capture digital evidence, and how that might play into preconceived notions.
Common Video Frame Rates: 15, 24, 30, 60
We preface this section with the fact that 15 fps is only a common frame rate when it comes to CCTV. But we’ll discuss here with the big three, because it’s relevant to your work, and it ties back to earlier discussion about human vision. The human eye can naturally pick up 15 fps easily. So easily in fact that while 15 fps falls into the popular label of “laggy,” meaning it has a time lag in presenting information, when a video with full information displays across a screen, our eyes will adjust to “normalize” what we’re seeing, to try to smooth out the information. This is interesting, because the information is moving slow enough that our eyes must embellish the information in insignificant ways to make the full picture into a story we can process the full visual information into. The time lag in 15 fps videos tends to expose itself when a video has one moving object, and everything else is stationary and consistent. For any law enforcement agency, we really don’t recommend this frame rate, or anything below it. There’s 7.5, 7, and all the way down to 1 fps. None of these are good for capturing digital evidence, however, there have been exceptions to that statement. Bigger picture, the one time a lower fps does grab an excellent piece of digital evidence, it will likely be backed by three or four dozen times that it doesn’t capture anything when you need it. That’s too big of a failure rate to accept.
The next common frame rate, 24 fps, is really tied to cinematic products, like movies, documentaries, film shorts, and other similar content. This frame rate has been used in cinema for a very long time, and many films still use it. For one, from the perspective of art, this frame allows the film to be separated in the mind from reality. And movies are designed to escape reality for entertainment purposes. It also was a cost savings measure employed by movie studios, due to the high costs of 35mm film. It’s the lowest frame rate that would still be acceptable by the human eye to perceive continuous motion. This could also explain why certain dramatic scenes in movies seem to move so slowly, even more slowly than the actors portrayed the scenes when acting them out. When you consider the purpose of 24fps, there’s no way you could accept this for collection of digital evidence. A crime scene made to look like a movie? That might inform the defense’s argument, right? But it might also miss key evidence, like slight movement meant to retrieve a weapon. In the real moment, the officer can see what is happening so clearly. But the device they relied on to capture the moment may make that situation appear drastically different from what was witnessed. You’re not a Hollywood director, nor should the digital evidence your devices retrieve make you out to be one.
The next common frame rate is 30 fps, and you tend to experience this frame rate on television. Again, we’re discussing the comparisons with media content. Why would anyone want their digital evidence to feel like a standard shot sitcom? Believe it or not, there is a good argument for using 30fps as your standard. In the first two fps outputs discussed, we touched on the fact that the brain must do some work to make up for a lack of true visual information in those frame rates. At 30 fps, the brain is still doing that work, on a much slighter scale. At this frame rate, we’re reaching the range where the eyes can process information for the brain comfortably, while not inherently manipulating things like motion blur, or background visual noise, and so on. Keep in mind at this frame rate, that is still happening, just not at nearly the same level as the previous mentioned frame rates. The difference between 24fps and 30fps when simply looking at the numbers may seem nominal. In practice however, there is a lot more information being seen in those six additional frames. Keep in mind, these numbers are the number of frames you see in one second. 30 frames can give you a lot of information in just one second.
The next most common frame rate, 60fps, is where things get fun! At 60 fps, we’re really hitting the top end of what’s possible for the human eye to experience fluidly, without having to add major amounts of motion blur and other visual noise to avoid those nausea-producing headaches we described earlier. And this is where we begin to experience what is commonly referred to as diminishing returns. You may remember that term from an economics course, in this case it’s describing that there’s a limit to the amount of visual information the human eyes and brain can process, before it reverts to filling in information around what they are focusing on. At 60 fps, we’re likely at our top end of what the human eye can process, without having to send back artificial blurring at the edges of the frame for the eye to focus.
Of course, the true difference in these frame rates comes in the experience of viewing them in action. What you’ll find is after exposure to video shot at 60 fps, 24 fps looks choppy in some instances, if not in most instances. Now, if you discontinued watching 60fps, there is a “normalization” the eye will go through, where if it’s watching 24 and 30 fps video, it will start to fill those details back in, and video shot at those speeds will begin to look normal again. But if you go back to watching a 60fps video, those choppy edges will be exposed yet again.
Emerging Frame Rates: 48 and 120
Two frame rates that starting to creep into commonality, are 48 fps and 120 fps. We’re going to discuss these in reverse order.
First, 120 fps is really something that video game users are familiar with, as it’s becoming a sought-after frame rate for online video gaming. We don’t recommend ever considering this technology for devices, it’s not designed for anything realistic, and works well in video games exclusively, for the simple fact that graphic-dense video games need more frames to process all the information involved, and still give the player realistic movement within the game environment.
The next frame rate is 48 fps. This has emerged from Hollywood directors Peter Jackson and James Cameron. If you saw The Hobbit: An Unexpected Journey, 48fps was used throughout the entire film. And many fans of the title, and film enthusiasts alike did not like the results, because it made the film seem “real” to them. That backs into our earlier discussion about 24 fps, but it also tells you what you may expect from your digital evidence shooting at this rate. While moviegoers won’t like the frame rate when they’re at the cinema, courts and criminal justice professionals may truly find use in 48 fps in capturing and reviewing digital evidence.
Playback and Data Considerations
We’ve discussed several different rates and what they’ve been used for, and whether you should use them or not, simply considering the recording of digital evidence. But there are some other things to consider.
First, let’s consider playback. In playing back digital evidence, we look at the actual playing of the digital evidence, as it was recorded, we look at how it performs in slower settings, sometimes generalized as “slow motion,” and we also look at how it performs while played in a rewind/reverse mode. Why do these various play modes matter? Because we need our digital evidence to sustain quality over different versions of playback, because it may be important to play a video at recorded speed, but then slow a section down, where details are finite, or partially obstructed, and be able to pick up what is happening. Remember, our eyes and brain have a limit to the information they can process, so we must be able to maintain consistent quality in our digital evidence, over all these variations of playback. Why might playing digital evidence video in reverse matter? Because it may help jury review an event with a focus on the details that matter. Watching something heinous once through is one thing, but watching those same actions in reverse can help those deciding to see the details of what transpired, and adding a slower presentation of the incident while in reverse can help people see things like premeditation in activities, or the viciousness a criminal acted in.
If you look at videos captured in 24 or 30fps played in reverse, you’ll find that the subtle choppiness you saw while it is playing at recorded speed, straight through, tends to be even more choppy, and can even make things that seem clear in the “normal” view seem even more questionable in reverse. It’s like the same piece of digital evidence telling two different stories. To put it bluntly, you don’t want two stories coming from your side of the table at trial, because that’s a bad day waiting to happen.
60 fps presents the absolute best quality when playing anything in reverse, or slow motion. The crispness of details, the succinct presentation of information, 60fps simply can’t be beat when it comes to the multiple versions of playback we evaluate frame rates by.
48 fps is certainly acceptable however. It has a finish in these playback modes that would present accurate information to anyone viewing them, though the difference between 60 and 48 is noticeable.
What about data at these rates? Well, in terms of file size, there are differences, but they are not as easy to predict as one would think. For example, if we shot a 60-minute video using 60 fps, and placed it on YouTube’s HD placement, we’d clock in at 8.64 gb. That same video shot at 30fps 4.32 gb. It would appear this is simply double the difference between the two, and that we can expect that every time. Take our same frame rates, and compare to a 60-second video, and the change in file size tells you what you can truly bank on, with 60 fps coming in at 144 mb, and 30 fps coming in at 1.2 mb. Big change, right? True both files are much smaller than our first example, but the disparity is what tells us something. The more frames you shoot, the bigger the file.
When it comes to considering data, your agency has some questions it needs to ask of itself. Is it OK sending digital evidence in that the capacity of fps has been overburdened by information, making for terrible quality on playback, or is it more important to save on data costs? These may seem easy to answer, but what if the data in question is an OIS, and the frames in the video are so overburdened that the gun witnessed by the officers on scene, isn’t very clear in the video? You may be able to see an object, but will the others viewing it see it as a weapon? We know to trust your officers, you know to trust your officers, but will the court? Will a jury? And will frame rate end up costing your agency in civil court? That’s what makes this discussion likely one of the toughest any agency will ever face. Where does the line get drawn when it comes to digital evidence, and the quality standards that must be met. And like we stated at the beginning, frame rates generally have not been considered in this conversation ever. And yet, they really should be.
Conclusions
When it comes to frame rates in the outside world, the arguments and reactions to content will rage on long after we’ve all retired. But when it comes to public safety, we must consider what it is we’re doing in digital evidence collection, and who we’re presenting that information to. We owe it to ourselves, and our personnel, to provide them a thorough discussion of what our standards are, and why they are that way. Because if we leave any part of the data discussion out, we may be setting someone up for failure. And that’s not acceptable.
Be safe out there!