The music swells. The actors portraying crime lab technicians gather around an impressive-looking computer. They input a fingerprint, and instantly they get a picture of the bad guy, with full identifying details. And then they go to a commercial.
That's the reality presented by the latest generation of TV crime dramas, which often use heavy doses of technology to keep the plot moving. Predictably, reality is different from TV.
"The big difference is that on TV they usually find a single match, which pops up with a picture of the individual," Michael Wieners, chief of the FBI Laboratory's Latent Print Support Unit, complained to LiveScience.
"What the system really does is provide a list of the most likely matches," Wieners said. "There must always be a human being who then looks at the print on the screen, side-by-side with the sample print, to determine which is really a match."
Doing so is not a simple matter of overlaying the sample with the stored file, if only because the elasticity of skin typically means you can't overlay the two images. The technicians must know what to look for, and knowing what to look for takes 18 months of intensive training, Wieners explained.
The other big difference is that the TV shows typically depict the operators simply inputting the fingerprint into their computer. In reality, the image must be carefully edited by the technicians to remove everything that isn't really a fingerprint, such as dirt and digital noise. Failing to do so will reduce the accuracy of the process by about 30 percent, he warned.
And while the TV shows typically depict instantaneous results, running a print through the FBI's database of 53 million files (called the Integrated Automated Fingerprint Identification System, or IAEFS) takes close to two hours, he added.
Full versus partial
The database represents full sets of fingerprints. Criminals usually only leave partial prints, but the system can handle a partial print as long as it is big enough to include five separate distinguishing points, he noted.
However, even if the prints are in the system, they cannot always be matched to the evidence print, since the part of the fingerprint on file that would match the evidence print might be blurred or smeared, he added.
He said that about 26 percent of the cases received by the lab include identifiable fingerprints. These often include multiple sets of fingerprints, since the investigators will submit all the prints found at the crime scene.
Investigators are asked to submit "elimination prints" of everyone who had legitimate access to the crime scene. The prints of everyone else are submitted to the system for identification.
Porous vs. nonporous
Wieners explained that the way fingerprint evidence is handled depends on whether the prints are on a surface that is porous (such as cloth or paper) or nonporous (such as metal or glass.)
Porous material is treated with chemicals that will interact with the sweat and oil of the fingerprint and make it visible, he explained.
Nonporous material is often put into a chamber with a few drops of heated superglue, whose fumes react with the fingerprints, hardening on them and making them visible. This simple process was discovered by accident in Japan in the early 1980s, Wiener noted. But investigators also continue to use older methods involving dusting the prints with black powder and then lifting them with clear tape.