I meant to provide two specific examples to some factual errors.
In one espisode they took a cast of a knife wound and came up with the shape & size of the murder weapon. I didn’t realize this could not be done until I read an article somewhere (may have been here, even) where a real CSI stated that they can’t do that.
Once they showed someone using a plastic drink container, like a 1-liter jug that soda pop comes in to silence a pistol. I can tell you from personal, direct experience, that this will not silence a handgun. Back in 1974, a friend and I tried this very stunt with a .45 automatic. We had to cut some small slits in the mouth to get the pistol in far enough so we could hold the plastic against the slide. We may have been stupid, but neither one of us wanted any part of our anatomy in front of the barrell! It took three tries to get this to work because in addition to the bullet, there is a lot of hot, fast gas coming out of that barrel as well and it kept blowing the jug off. We finally put the end of the the third jug against a hay bale and help on and this time we were able to keep the thing attached when the pistol fired. Folks it was still LOUD. I don’t know if it was any quieter than without the plastic jug, but we could not tell any difference; it still rang our ears. I’d say “This myth is busted.”
I love CSI—well, the original Las Vegas one, anyway. Still can’t stand David Caruso and the whole Miami crew, and i’ve never been able to get into the New York version.
Anyway, as someone who plays around with digital images a bit, what chaps my ass most about CSI—and indeed about most such shows—is the sort of stuff they claim to be able to do in enhancing digital images from security cameras, etc.
The amount they are able to blow up a digital image and still get sufficient resolution to determine a license plate number or some even smaller detail just boggles the mind. Someone needs to tell these morons that digital images can only be enhanced so much, and that it is physically impossible to go below the pixel level to get detail.
“Numb3rs” recently fell prey to this also. Their quote was something like “given enough time, mathematics can enhance any hidden detail in an image by blah blah blah…”
After all, why is the government wasting all this money on new fancy spy satellites, when they can just loft a cheap polariod camera into orbit and enhance the snapshots after they get developed at Costco.
One thing that really irked me in the Miami version (which I can’t actually stand - Vegas is much better) was watching Kim Delany use a pipette. She held it about 8 inches away from the tube she was pipetting into. Honey - this isn’t like pouring Moroccan tea - you don’t get points for distance.
It was something that was free and easy to get right, but they didn’t bother.
Interestingly, they seem to do it correctly on the Vegas version, so perhaps it was just Kim that wanted to be different.
They’re suckers for spending the effort getting the cameras into orbit. The way to go is to float a helium-filled mylar balloon and take pictures of that from the ground – and then use enhancement to take “upskirt” pictures of women stepping over puddles.
My personal favourite magic zoom is in Bruckheimer’s Enemy of the State. Start with a low-res frame from a black-and-white security camera. Zoom in on a guy with the shopping bag. Zoom further in on the shopping bag. Casually use the mouse to rotate the bag 180 degrees along its Z-axis, so you’re looking at the side that faces away from the camera. Then apply an algorithm that extrapolates the shape of the object inside the bag from the shading on that side.
Every time they do a test with the black light, they know what the stain is immediately. That’s the perps spit, while that one’s where the baby pee contacted the rug when the baby-sitter dropped a diaper last week.
They tend to find people by medical records at an unrealistic success ratio.
I saw Grissom take an uncut diamond, run it through a machine, and get a full grading report. Now, your grades of color and clarity are only relevant for cut diamonds, the grades are determined by expert analysis, looking down through the top under 10x magnification. An uncut diamond is ungradable.
They also compared the size of blood spatter and determined that it was spit blood. That’s cool and all, but am I to believe for an instant that I can spit blood on a wall and have ALL the dots be 2mm in size?
A minor point, but recently they were shown tightly wrapping a car in stretchy plastic wrap to “preserve evidence” until they got it back to the lab for analysis. In a couple of hours of Miami heat, the plastic wrap would totally fuse with the paint and be impossible to remove without taking the paint off with it. They would never do this.
To expand, a machine can’t grade a diamond. Diamond grading is all by human beings (which you alluded to when you said “expert analysis”, but that may not have been obvious to everyone). I watched my diamond engagement ring being appraised, including a grading of the center stone - it takes a while, and is done by comparing with a set of standards. It’s really very subjective, which is why anyone would bother to have a stone they buy with a certificate re-checked by their own gemologist.
Their databases are insanely comprehensive and stupidly relational; they don’t ever have any trouble identifying anything, because it’s all in the database - a plant fragment? - the computer can tell you the precise locations it grows in central park. Nondescript machine screw? -by examining the pitch of the thread and the tooling marks, The computer can identify that it was made on the afternoon of the 4th of April 2003 on production line 6 in some tinpot factory in the czech republic
They don’t ever have any trouble tracking and cross-referencing phone records, credit card bills, dental patient records, lists of restaurant patrons etc - they just ask the computer and it finds the information for them.
ARRGGHH! I just about had a stroke when I heard that. Larry (supposedly an astronomer!?!) said something like, “There’s no theoretical limit to the amount of detail you can retrieve from an image.”
GAH! Yes there IS!!! Real astronomers call it R-E-S-O-L-U-T-I-O-N. Airy disks? Gaussian, Lortenzian or Sparrow spread? Any of this ringing a bell, Larry? SOUND FAMILIAR AT ALL? LARRY?
Possibly the grain of truth on which they were basing that: If you had infinite intensity resolution, and knew the precise convolutions that went into blurring an image, then you could, in theory, exactly deconvolve and reconstruct the original image. In actual practice, though, you don’t have infinite intensity resolution, and don’t know the precise convolution, so information is in general lost in the convolution. Sometimes you can still get some improvement: For instance, before the corrective optics were installed on the Hubble, the astronomers working with the data had a very good idea of the point-spread function, and so were able to largely correct the images. But there’s always a limit, and those corrected images (corrected as well as legions of astronomers and programmers worldwide were able) were still noticeably inferior to the later images after the optics were fixed.
Under that logic, if you took any random amateur photograph of the night sky, and kept focusing and enlarging, focusing and enlarging, eventually you’d get a picture of the back of your head.
some bad science but Ny and LV are both good shows, miami can suck my nuts, David Caruso is if anything even worse than Keanu Reaves. and I seriously didnt think it was possible to get even close to reaves in the “I am really a zombie reading lines off a teleprompter” department.
I don’t know if this counts as bad science or just ignorance. One episode hinged on the burn marks an ejected 9mm cartridge left on the baddie’s clothing and skin. I’ve had a .45 cartridge go down my shirt before, leaving burns on neither skin nor shirt. I’m pretty sure one that bounced off my collar wouldn’t leave a mark, much less be hot enough to sear flesh through my bling.
A few years back, I saw a brief piece in People magazine about the real head of the CSI lab in Las Vegas. William Petersen based his character, Grissom, on this gentleman, and even gave his character some of the real CSI guy’s quirks (for instance, the real Vegas CSI boss has a huge collection of insects).
The People interviewer asked the real CSI man how realistic the show is scientifically. The CSI man laughed and said (I’m paraphrasing), “Well, we really CAN do almost everything you see on the show. The most unrealistic thing is the time element. It would take me 6 to 8 months to do most of the tests they do in a few minutes. But that’s the nature of a weekly TV show. They have to resolve everything in an hour.”
