At 1:42 AM +0000 1/7/03, olympus-digest wrote:
>Date: Mon, 6 Jan 2003 12:43:34 -0500
>From: "Walt Wayman" <hiwayman@xxxxxxxxx>
>Subject: Re: [OM] Cameras Don't Lie
>
>These are (probably) my last words on the faking of photographs
>for use as evidence in legal proceedings.
>
>Someone mentioned the enormous amount of work involved in faking a
>photograph. But I wasn't talking about creating some totally
>fictional scene out of whole cloth, like something from "The
>Matrix," for instance. What I refer to are subtle changes made in
>otherwise accurate photographs.
Even small changes involve lots of information, and what looks good to the eye
won't pass scientific muster.
>For example, a case we recently tried involved a guy suing a night
>club, claiming he was roughed up by a couple of employees, a
>bouncer and a bartender, and tossed out for no good reason. As
>part of his evidence, he submitted photographs of himself
>shirtless, taken a couple of days after the incident, showing his
>alleged injuries: a few scrapes and bruises and a relatively minor
>black eye and a slightly fat lip. It would take little skill, and
>almost no time, with Photoshop to turn those minor boo-boos into
>some pretty serious looking injuries. That's the kind of thing
>I'm talking about.
>
>Little things can make significant differences, depending on the
>circumstances and exactly what the photograph is offered to show.
>For instance, pictures of an accident scene can be altered just
>slightly, yet sometimes a subtle change is enough to raise doubt
>about the fault of one or the other of the parties.
>
>As an example of such chicanery, you might look at my TOPE 10
>photo. I completely removed two distracting signs, a speed limit
>sign and one directing traffic to a parking area, from the road
>running along the left of the picture. Bet you can't tell where
>they were, even now that you know they were once there. (And on a
>scale of 0 to 10, my skill level with any photo editing program is
>less than 1.)
This reminds me of those Soviet photos from the 1940s where out-of-favor
leaders were simply airbrushed out of official photos. Nothing is new except
the tools.
If these photos were introduced into evidence, and if the presence or absence
of those signs were material, great scrutiny would fall upon that part of the
photos. In this case, someone would just go and re-photograph the scene, and
the jig would soon enough be up. Failing, a scientist or two would analyze the
photo where the sign was claimed to be, and the evidence of a clone and smooth
operation would be found. What will fool the eye is far different from what
will fool a expert armed with scientific tools.
>And as for making a "fake" negative or transparency from the
>digital, you can forget that. Unless you thought to do so ahead
>of time, you'll have about one minute, two if you fake a coughing
>fit, to think up a response to the challenge to your digital
>photographs and the request that the original negatives or
>transparencies be produced, and you sure won't be given a recess
>and time to go make them. Besides, under increasing magnification
>of a piece of film, I would expect to start to see grain
>structure, not pixels, and if I saw pixels, you would be in some
>deep, deep doo-doo.
This is the legal constraint I mentioned -- the person that attests to the
veracity of the photo is the key, not the photo itself. And, looking for
pixels is an example of an expert's test, one that goes beyond what looks OK to
the eye.
>The truth is, people are just plain sneaky sometimes. Keep your
>eyes open. Many, many things are not as they seem.
Yes. I read an article some years ago claiming that something like 50f felony
accusations (not inditements) are knowingly false. That's a lot, and one hopes
that the legal system is very efficient at ignoring such accusations.
Joe Gwinn
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|