CSI stuff:Fact/fiction/both

KingOfAquitaine · 8 Aug 2006 at 23:05

Been wondering this for a while now; how much of the techiques employed it CSI are real or completely made up. Reffering to things such as the computer simulations, the chemical analyses + the (lack of) time they take, the cut testing on bits of meat etc

I would assume they fall into 3 basic catagories

1) Definately real. Used by forensics teams world over
2) Technically real/available, but eitehr don't work as well as suggested or cost way to much to be common
3) Made up

Anyone know?

DB_SamX · 8 Aug 2006 at 23:08

Ooh, I'd like to know as well. It's one of these mysteries...

Curiousalien · 8 Aug 2006 at 23:14

My ex-girlfriend did her degree in crime scene science and she used to tell me about the stuff they used. In america they apparently have much worse equipment than we do over here. so thats one thing that is wrong.

When they spray chemicals about the place (luminol? i think this was a couple of years ago) to find blood it glows for ages. in real life you use 2 sprays to get the glow effect and the effect of these sprays fade quickly so you spray them both once, find the trace, set up a camera spray again take photo, you can only spray 3 times for some reason so i suppose the 3rd one is incase of mistakes, the whole process takes a lot longer. Also the chemicals are not that clever and cannot just show up blood they also glow at traces of rust, bleach and poo.

thats all i can remember off hand but there are a lot more photos taken in real life photos of everything! a lot of her course was photography.

will add more if i can remember it. I think that the whole computer thing they use in the show is made up though. Fun show to watch though.

Werewolf · 8 Aug 2006 at 23:44

KingOfAquitaine said:
Been wondering this for a while now; how much of the techiques employed it CSI are real or completely made up. Reffering to things such as the computer simulations, the chemical analyses + the (lack of) time they take, the cut testing on bits of meat etc

I would assume they fall into 3 basic catagories

1) Definately real. Used by forensics teams world over
2) Technically real/available, but eitehr don't work as well as suggested or cost way to much to be common
3) Made up

Anyone know?

I suspect a lot of it is rubbish made up to make something normally incredibly boring look exciting

Such as the recovery of data from a wiped hard drive by laser reading the exposed platters I saw in one episode

Some things like the "poke a stick in the bullet hole" are I beleive based in reality.

Tru · 8 Aug 2006 at 23:53

As with most productions like this they have technical staff who advise and correct them. They add some pazzazz for effect but generally all the chemical stuff is at least based in reality, chemical names etc will be accurate and I imagine the machines and methods will be similar.

The most outlandish thing I see them do is ZOOM, ENHANCE. They used this technique to get the accurate image of a killer, by zooming and enhancing on the pixels in the eye of someone he photographed.

Beansprout · 8 Aug 2006 at 23:56

I watched CSI today for the first time and thought the good attempts at realism were, er, good

Won't talk about the crapness of the actual thing though as that's another thing

TheVoice · 9 Aug 2006 at 01:30

Most of the procedures used in CSI are reasonably accurate, although as mentioned already, they're not always as precise as the TV programme makes out. The speed at which evidence is collected, processed and then followed-up on is also blatantly unrealistic. In reality, it'd be a couple of days before an incident's evidence is fully processed and reported (DNA testing alone takes quite a while).

There's a program on before CSI on Saturdays called "The Real CSI" which gives you a decent insight into how a Las Vegas-based PD handles crime scenes.

Berserker · 9 Aug 2006 at 01:40

Basically, all three. They use real stuff most of the time, but speed things up and/or exaggerate and/or make things up for dramatic effect.

Also, the same goes for the likes of Casualty - they have medical advisers on the team but may adjust things a little for dramatic effect. They have to be careful to avoid anything outlandish though - people have been known to use life-saving techniques from Casualty to save people in real life. :eek:

Raymond Lin · 9 Aug 2006 at 01:47

Berserker said:
Basically, all three. They use real stuff most of the time, but speed things up and/or exaggerate and/or make things up for dramatic effect.

Also, the same goes for the likes of Casualty - they have medical advisers on the team but may adjust things a little for dramatic effect. They have to be careful to avoid anything outlandish though - people have been known to use life-saving techniques from Casualty to save people in real life.

I find ER far more acccurate than Casualty, not that i am a doctor mind. I know most legal programs have a tendency of time jumping, its like "you are nicked" to "30 years behind bars" in the space of a week. In reality, they take months before they even get to trial. And a real court case is never that dramatic, it is all very very civilised.

I have also been told CSI are pretty accurate (according to a few friends who does Foreigsic science, they all LOVE CSI.), like they don't invent a machine just to process an evidence. All the machines they have actually exist, it is some of the bits a dramatised for television.

Meridian · 9 Aug 2006 at 06:21

Mostly 2). Real forensic science is a lot more "maybe", "there's a link but it doesn't mean much", "it's probable" etc. Very little of the real thing is certain. And the results from most of their scientific results (especially the high tech ones) are much cleaner than in real life - I like the way that blowing up a picture doesn't lose definition.

Oh - and you can't get a DNA profile out of an Agilent 5890/5971 GC/MS, which CSI seems to be able to do. And it comes with £4000-worth of control and analysis software, so why do they always do manual control?

And in this country we wear scene suits.

M

Borris · 9 Aug 2006 at 06:59

There was a slashdot article on this a year or two ago.

One poster worked out the real world financial costs of the CSI team's experiments, averaged across various episodes.

It turns out that, irrespective of the innaccuracies in technique & timing (artistic license, I suppose - altough many seem to take it as gospel), they would have blown the entire annual budget of an average forensics lab within the space of a couple of weeks.

Takhisis · 9 Aug 2006 at 07:23

An author I read, who does actualy work and do research in the "CSI" field (forgotten the name of it) has a test/quiz on her website:
http://www.patriciacornwell.com/home.html
(Flash into but there is a link at the bottom to the challenge)

alexakasloth · 9 Aug 2006 at 07:50

Tru said:
The most outlandish thing I see them do is ZOOM, ENHANCE. They used this technique to get the accurate image of a killer, by zooming and enhancing on the pixels in the eye of someone he photographed.

I've always found that one funny. How can they enhance it all the time?

Borris · 9 Aug 2006 at 08:35

$loth said:
I've always found that one funny. How can they enhance it all the time?

I prefer the Bladerunner version, wher he can not only enhance a grainy photo, but travel around corners in a 2-D image.

kiwi · 9 Aug 2006 at 10:36

$loth said:
I've always found that one funny. How can they enhance it all the time?

i'm pretty sure i read somewhere that the CIA ir similar have a very expensive algorithm that allows them to blow up images with no great loss of detail. doubt the CSI could afford it though.

Scott Salisbury · 9 Aug 2006 at 10:49

TheVoice said:
The speed at which evidence is collected, processed and then followed-up on is also blatantly unrealistic.

I think if they kept the time lines realistic we would be in for very boring viewing. I know everything they do seems to be done in one day, but I like to imagine that it is over a period of weeks.

What REALLY makes me cringe in CSI is the computer systems and their blatenly made up interfaces. When they load up a database to find someone, for example, and as it is scrolling through it beeps, "beepbeepbeepbeepbeepbeepbeepbeep", as it is searching. Err, just no.

And as it is searching you see photos of everyone flash up for a milisecond and flashes through them, it wouldn't do that.

When they zoom into a photo, it doesn't just zoom in, it "beepbeepbeepbeepbeepbeepbeep"'s in.

Other than that I find it enjoyable.

Tru · 9 Aug 2006 at 10:54

kiwi said:
i'm pretty sure i read somewhere that the CIA ir similar have a very expensive algorithm that allows them to blow up images with no great loss of detail. doubt the CSI could afford it though.

I'd be interested to see any hard facts about this. You can enhance/sharpen images to a certain degree, but for the most part the computer would be creating detail based on very little information. I just can't see how an algorithm could take, say, a 64*64 block of pixels and increase the detail.

Fusion · 9 Aug 2006 at 11:00

Tru said:
I'd be interested to see any hard facts about this. You can enhance/sharpen images to a certain degree, but for the most part the computer would be creating detail based on very little information. I just can't see how an algorithm could take, say, a 64*64 block of pixels and increase the detail.

Agreed. Fractal interpolation is a relatively new technique that allows the sharpness and definition of images to be better retained when enlarging that with bicubic, but it can't implement detail that wasn't there initially.

Borris · 9 Aug 2006 at 11:00

Tru said:
I'd be interested to see any hard facts about this. You can enhance/sharpen images to a certain degree, but for the most part the computer would be creating detail based on very little information. I just can't see how an algorithm could take, say, a 64*64 block of pixels and increase the detail.

Exter-/Inter-polative / deconvolutive techniques can be used to estimate the true image (for example, removing halos from lights - something that can be predicted).

I doubt that any algorithms are not held in the public domain. The requisite hardware, on the other had, is probably in the league of a military budget (but imagine the WUs one of their server clusters could pump out).

aardvark · 9 Aug 2006 at 11:04

i'm a scientist and so are virtually all my friends and the only time we watch it is if we need a laugh - the vast majority of stuff shown is laugh out loud wrong - does that answer your question?