Transparency anyone? Why the lotto scratcher ticket approach is no good for forensic technology?

Considering all the things going on in the world today, a little actual, real transparency is a welcome sight. We are often asked “what makes Five a better solution than X?” or “compare Five to X, Y, and Z”. In this context, transparency is a fundamental requirement in being able to weigh the options out there and make an informed, intelligent decision. That is the lotto scratcher approach – the contents are a mystery.

The problem is that (with the exception of us) even finding actual specifications is a futile endeavor. Granted, we have it easy compared to our competitors because we produce our own software without being dependent on a third party application or hardware unit to function. When we publish specs, it is specs on our product; so you don’t have to guess where something ends and an Adobe, Avid, or an Acme (with apologies to Dr. Wile E. Coyote, Forensic Scientist) product begins. This blurred line becomes problem when you have to actually go to court with something, but more on this later. This is what I mean by it being easy. We can publish our specifications on our software since we created it – completely.

This is easy for us, but not for our customers. When you are confronted with the lack of coherent or complete specifications, how do you make a comparison, an informed decision, or understand your needs as they relate to a product? Also, how do you chart total cost of ownership? If you have to buy an update from one element of a “plug-in” based application, how does it affect the host or other plug-ins? Do you have to update hardware, too? What happens if you don’t want to update, but your hardware dies? Will Acme still make or stock replacements for the older technology to match what you have? If not, will eBay be your parts source for older GPUs or other hardware? How do you explain that to management, or worse, in court (yes, your honor, I fixed this machine a month ago with parts ordered from some guy on eBay in China…)? Not easy, nor transparent.

The other obvious problem with plug-ins or other types of third-party dependent solutions is how do you document the scientific value for those parts that are basically commercial software for modifying images or video? If you use software plug-ins or collections of scripts for Photoshop, how can you explain to a jury how Photoshop’s tools worked or refute the defense’s claim that things weren’t “Photoshopped” to make the defendant “appear” guilty? If you need it, will Adobe release data? Have they ever before? To make the answer apparent, it is obviously not going to happen.

Also in the future, with Adobe hosted products going to the “cloud” for the Adobe parts, can your agency change their internet use procedures to allow you to have your forensics machine online 24/7, or once a week, month, at random? Since no one is clear how this will shake out, it is pretty obvious that Adobe is doing this for their core “creative” customers who “have asked” for such things at your expense. Did Adobe ask you how it would affect your use? No one knows how this will shake out, but it isn’t the best way to deliver a product to users that can’t go online due to IT policies.

Since the specs for most competitors are very sparse, it must be said that not all technology displayed or claimed is equal either. When someone leads their specs with something like “Demultiplexing” as something new or innovative, you need to look at a calendar and realize it is 2013. Ten or fifteen years ago, when video tape was king or even ten years ago when analog was all there was; multiplexed recordings were a big problem. With the exceptions of cold case detectives or hard working Postal Inspectors, when was the last time you got tape that needed a $30K “new” solution? Granted tape still does come up, but it is getting less frequent, and we offer tools for it (like demultiplexing) too. The difference with us is that you don’t see us leading with it in our advertising as an innovation. We realize it is becoming seldom used, so we leave it there just in case it is needed.

Some claims of “innovation” out there can be likened to a car company listing “now with a CD player” as an innovation for today. That would wow folks in 1990, but a little tired, boring, and common for 2013. Everyone’s software and Windows Picture Viewer adjusts contrast and brightness, so why highlight it as if it is something new or something you developed? More or less everyone’s software does deinterlacing of one kind or another. It is basically old technology. Does anyone else actually explain how it works or provide the scientific references for it? We can’t and won’t claim any kind of innovation for these types of tools, but we at least provide you with information you need to help you use them properly and explain them to a jury. Interestingly, these are common “innovations” claimed by many competitors in their sparse specs. While there was a time and a place for such things, new evidence technologies require new tools.

Our innovations apply to the current digital world, both for images and video. Things like offering both 32 and 64 bit versions as standard to cope with higher definition images or videos from the current crop of megapixel cameras, or offering several options for stabilizing, tracking, and deblurring video where the user has total control over all settings and frames. Those are innovations in the current context, not some warmed over ‘90’s tech. The same can be said for “specialized hardware” as this isn’t really needed these days. Unless your needs are to process and fuse multiple sensors from a high-def feed of a UAV in real time (but then you have an air conditioned shipping container that says “Northrup-Grumman” on the side air-dropped to your site), standard commercial PCs are up to the task of running FVA software perfectly. The need for your average PD to have special media servers, bloated non-linear editors, special boxes or magic beans has come and gone when Intel made the Pentium 4. Commercial hardware is more than capable today to get the job done.

Since we cannot be everywhere at once or visit each agency in person, we also offer Five as trial software that a law enforcement agency, military, or other government user can download and run on their computer for a specific time period to test drive and see how it works. Since Five runs on any standard PC, it can be tried and tested very easily. We don’t need to send out anything that requires truck freight, multiple Pelican cases, or agency T&E agreements with financial guarantees that need to be signed by upper brass. We’re not akin to certain politicians who believe that you should find out about something after you have purchased it, so we let our government customers try it first. You can see, test, validate, and explore prior to going through the whole procurement process only to find that what you bought doesn’t meet your needs. We provide help and tutorials built-in to make it easier to explore functions (Note: We know that tutorials and examples are not a replacement for training. We offer comprehensive classroom training as well, taught by the best instructors – bar none). Purchasing decisions can be seen as career decisions to some people in some agencies, whether it is for you or (worse) your boss.

Some competitors do not list specifications or offer trials at all. To put it in terms of lotto scratchers – you may get lucky and get what you want, but you have to buy ticket first. Fine for $1 in your own time, but for agency purchase? Not wise. That is why we let you test drive before buying and also why we publish specs that you can download from our website.

Transparency is in our presentation of our reports as well. With a couple of clicks, you have a report generated in Amped Five Professional that provides all relevant aspects of the project, including filters used, settings, and snapshots if you enabled them in the project. Great, huh? But I can get log files in Photoshop. Yes, you can get log files in Photoshop or keystroke logs in other products and we offer that in Five, too, but it isn’t what we consider a report. It is helpful for us as a developer to solve technical issues, but of little value to you, and certainly not scientifically or forensically valuable.

Our idea of a report is one that can be used to back up your project in court with science. We include not only the settings, filters, and snapshots but also the scientific references for all algorithms from peer reviewed scientific journals, text books, and papers that have been published and accepted in academia. Our reports include all information necessary to comply with tough standards such as California’s Evidence Code’s requirement for cited, peer-reviewed scientific publications. Competitor’s keystroke logs or “reports” with no references can’t get close and aren’t really up to the improving standards of most courts.

The other criterion for something to be scientifically relevant is the repeatability and the reliability of the results. Our software’s results are based on mathematically precise adjustments that allow for the reliability / and 100% repeatability required by contemporary evidence codes. The steps and settings are all displayed in the report so results from Five can always be repeatable, as the project can always be reconstructed from the steps documented in the report. This makes it so even a prosecutor can reconstruct the project, step by step and get the same result using the project as a road map. A keystroke log and a long set of notes is great to analyze some things, but to explain it in court would prove to be a potentially painful experience. It certainly won’t win any friends in the prosecutor’s office when they have to re-explain it to a jury, and will take at least five times the amount of time it took you to generate a result from the actual video forensic work. Why not do it quickly, precisely, and correctly – and then use the extra time to work more cases?

The bottom line is scientific value and merit of technology. If:

• You can’t explain it.

• You can’t repeat it.

• You can’t find a reference for it.

• You can’t identify it or determine how it was used.

• Or you can’t understand if it is even present in something you purchased or are looking to purchase.

If any of these statements apply, what you are looking at is not based in forensic science, it becomes forensic opinion.

Since fiscal years are ending in the next few months and budgets are coming together, before you spend your agency’s limited taxpayer funds on products from our competitors, how transparent are they? If you can’t even get a coherent or complete spec sheet, and know what is offered, is this really a reasonable approach or a wise use of funds? I guess only if you hope you’re lucky.

There is a better way to buy forensic technology.