When can a photo be trusted?

Fourandsix on TwitterFollow @Fourand6
« FourMatch: A 15-year journey | Main | Introducing Our First Product: FourMatch »

We might not want you to buy FourMatch

Of course, that really depends on who “you” are.

There’s a set of people for whom FourMatch can be a hugely valuable tool, and we certainly want to sell to as many of those people as possible. But then there’s another set of people who are intrigued by the idea of detecting photo fraud, but for whom FourMatch isn’t really the right solution. We’d love to sell something to those people, too. But not yet.

All of us have had the experience of being fooled by a manipulated photo. That’s why everyone can appreciate the value of being able to detect when photos have been manipulated. That’s also why a small startup without any big venture capital has been able to generate a fair amount of press, even before we’ve had a product to sell. People are genuinely interested in the problem.

But it’s a complex problem, and the reality is that there is no magic bullet that will tell you everything you might want to know about the editing history of any image. Instead, there are a variety of evolving techniques and technologies that can provide you with important clues. The value of these clues will vary based on the scenario you’re dealing with and the workflow constraints you may have.

We don’t see ourselves as a one-product company. We see ourselves as existing to solve a collection of related problems. In this early stage of the company, we’re cherry-picking those problems we can most effectively and quickly solve with the technologies at our disposal. As we progress, we’ll add new features and new products that address other problems for other people.

So what about FourMatch? Who is it for, and what problem does it really solve? Somewhat bucking expectations, FourMatch isn’t exactly a tool for detecting image manipulation. In contrast, it’s a tool that can provide compelling evidence that an image has not been manipulated. It doesn’t really solve the problem of how-can-I-tell-if-this-random-photo-is-manipulated? Instead, it focuses on providing confidence that an image is real.

This capability is most critical in legal situations. How can you convince a jury to trust photographic evidence, when they know how easy it is to manipulate it? Even if you fail to see any visual evidence of tampering, how can you know that the photo wasn’t modified by a skilled editor? This is where FourMatch comes in. It provides objective evidence that a file was not touched by any software application since the time it was first captured.

Certainly the potential uses of FourMatch are not limited to legal scenarios. But it is definitely most useful in situations where you can require or request that people provide you with unmodified camera originals. So, if you’re a news organization, you’d want to ask your photographers to send their originals for verification, or track down the source of a social media photo so that you could ask them for the same. If you’re an insurance company, you’d want to require claimants to submit the original photos from their camera as evidence of damage. In these scenarios, a failure to pass the FourMatch test is a clear sign you may need to ask the photographer some probing questions. But, if you’re unwilling or unable to put such submission requirements in place, FourMatch’s utility will be reduced.

It still will have some utility, though, because we’ve aimed to push this particular technology as far as we could. So, even when a photo fails the FourMatch test, the application looks at the exact way in which an image failed and provides some insight into what that may mean. These insights are less conclusive than a positive result, but they can point you in the right direction for further investigation, and, in some cases, they may still give you sufficient confidence that the image is still “real.”

What FourMatch is not, however, is a general consumer tool designed to tell you the truth about every photo around you with just a click. We’d love to achieve that in the future, but it’s still a ways off. In the meantime, we’ll continue to pick off problems as we can, and slowly bring some trust back to photography.

Oh, and, one last thing: That $890 price point should also be a pretty clear signal that this is not a consumer tool. That price accomplishes two things. It ensures that some of the folks with unrealisitic expectations will be unlikely to purchase it on impulse. And it also ensures that, as we begin this journey by focusing on a narrower set of problems for a narrower set of people, we can still bring in enough revenue to fund solving that next set of problems. For the people we’re aiming FourMatch at, we aim to provide more than enough value to justify the cost.

PrintView Printer Friendly Version

Reader Comments

There are no comments for this journal entry. To create a new comment, use the form below.

PostPost a New Comment

Enter your information below to add a new comment.
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>