I have not actually seen [REC], the 2004 Spanish outbreak horror that provided the blueprint for 2008's Quarantine; but the idea of remaking a highly improvisational "you-are-there" subjective camera flick brings up an interesting conceptual problem. Namely, do you reshoot what ended up on the screen or do you try to recreate the conditions that produced the final film?
To clarify things, let's abstract this problem from the films at hand and look at a hypothetical situation. Pretend, for a moment, that you've been tapped for a remake of The Blair Witch Project that is "faithful" to the original. For now, ignore the question of whether or not the film should be remade and ignore the fact that, even in context of Hollywood's chronic novelty-drought, it's still too soon to be disinterring that particular flick. Instead, think about what it means to remake that original flick.
Famously, The Blair Witch Project was created by giving a trio of actors a couple of handheld cameras, giving them a character design rather than a script, dropping them off in the middle of the woods, and then messing around with them until the produced enough raw footage to make a movie in editing. As I see it, you've got two distinct and somewhat mutually exclusive choices before you.
Option 1: Reverse Engineer the Final Product
Your first option would be to take what ended up on the screen – which was a combination of improv and cutting room choices – and work backward to produce a script and storyboard that would lead, if followed, to a recreation of the finished original. This would, in essence, plan for the accidents and randomness that formed the first one. Unlike the original flick, your film would actually have minimal improv. The words the original actors made up on the spot would, in your version, be carefully scripted. Mercifully, you can also avoid hanging out in the woods and being miserable for several days: you won't have to just shoot the crap out of everything and dig a movie out of it all in post. This would give you finished product that was indisputably faithful to the final film; but, be honest, it seems kinda to miss the point, doesn't it? It is less a faithful remake than a sort of po-mo simulation: a perfect copy of an imperfect original that wipes out its imperfections. Arguably, there's nothing for viewers inside the original film itself to suggest its origin story. On a pragmatic level, our awareness that BWP is a sort of experimental, open-process flick is really a product of the marketing surrounding the film – making of docs, interviews, content from other reviews, etc, - rather than result of the content of the film itself. The original BWP could have just as easily been a cheapo indie horror with a tight script and deftly handled shakey-cam work, and the whole origin story could have just been another level of the film's Byzantine publicity machine that was, at the time, as famous and popular as the actual film was. Still, on a more emotive level, its hard not to say that this remake would be somehow both extremely accurate and completely wrong-headed.
Option 2: Recreate the Conditions that Produced the Original
This option involves finding three actors that are willing to be abused, loading them up with camera equipment and power bars, and the getting them lost in the words. You hassle them semi-randomly for a number of days, gather up their footage at the end, and then see what you can do with all that coverage in editing. On a literal level, this would pretty much ensure profound levels of "infidelity" on the big screen. Unless your actors were conspiring, of their own accord, to recreate scenes from the original film, then it is unlikely that you get much of the same dialogue, few of the same shots, and little of the improved "plot" that occurs in the first film. You could very well end up without, say, the famous hyper-close-up of the repentant and horrified Heather apologizing to viewer. That said, there is something about this approach that feels more true to the spirit of the original, regardless of the fact that it guarantees the lowest possible level of "accuracy."
Of course, you could try to fuse the two approaches. I think you'd find that the combo didn't work. I assume that the more you genuinely allowed for the second option, the harder you would find it to believably apply scripted segments to the flick. That's my guess anyway.
I bring all this up because I found this question the most involving thing about the well made, but very thin Quarantine.
For those who have seen neither version of this flick, the film involves a local news reporter, Angela, and her cameraman, Scott, on a gig profiling the firemen at a local station house. Angela is gamely played by Jennifer Carpenter – a.k.a. Dexter's sister – whose playful tomboyish looks work well throughout light-hearted set up of the flick, but work against her when the fit hits the shan. The two fireman leads are played by Johnathon Schaech, sporting a truly impressive 1970s-porn-star-worthy moustache, and Jay Hernandez, the dreamboat hero of the picture.
The firehouse tomfoolerly that gives the film an up opener comes to a screeching halt when the firemen and crew are called to respond to a medical emergency at a slightly decaying apartment building. Turns out an old lady in the apartment building has turned nutso. She attacks her rescuers, killing one of the cops who responded to the 911 and putting Mr. Moustache the Fireman on death's door. Seems that she's got a hyper-virulent form of rabies that makes you go all rage-virusy within minutes of contracting it (unless, apparently, it would be more dramatically powerful for you to inexplicably hang about with the sniffles for a little bit before becoming all nom-nom-nom on people flesh – the virus, oddly, has a real flair for building tension).
To make sure that the residents of the apartment complex are truly and thoroughly screwed, the CDC and slew of military types put the building under quarantine – truth in advertising! – with standing orders to meet any effort at escape with extreme prejudice. The CDC/military industrial complex also cuts their television cable, their power, and cell phone reception. You see, being trapped in a building full of rabid psychopaths isn't panic-inducing enough. The CDC really prefers its subject to be crawling-up-the-walls, brain-each-other-with-hammers crazy.
Once we've handled all the grunt work of establishing the cast and unleashing the virus, the flick kicks into overdrive. About two-thirds of the way in, Quarantine becomes a ruthless and minimalist plot engine. This relentless narrative drive is both a blessing and a curse for the movie. It's hard to deny that the film is compelling on a gut level. The seasickness of the camera work, the score-less soundtrack of breaking tings and people, the crying and vomiting – it all has a numbing, gripping power. The scares hit when they should and the occasionally successful human counter-attacks get the blood racing. The creature design is also properly creepy. Plus, we've got a rabid monster child – evil children are horror gold.
Unfortunately, what it isn't is involving. The moment director John Erick Dowdle takes the breaks off, we lose almost everything that pulls us into a movie: Characterization evaporates, most of the visuals become a frenzied blur of action, dialogue increasingly tends towards a wall of incomprehensible shouting the general volume of which becomes the chief method communicating the mental state of the characters. In fact, with its rigorously delineated field of play, concern with keys and weapons, literal use of higher and higher floors/levels, and the sense of an over-the-shoulder floating camera, the whole thing resembles a super-high rez survival horror video game.
Quarantine is like one of those angular, minimalist pop dance singles from the late 1990s. Dance-floor oddities like Missy Elliott's remake of "I Can't Stand the Rain." On one hand, the aggressive inorganic feel suggested some experimental post-rock work, but the ethos driving the thing was pure disposable pop. The same odd dynamic is at work in Quarantine. The meticulous simulation of subjective camera work (this film is more like Cloverfield in that it is a very carefully staged simulation of what BWP made its method) requires an almost absurdly fanatical attention to craft. In it methods – the lack of a score, the sequential shooting, the reliance on elaborate and subtle lighting schemes – the film flirts again and again with something truly weird. But all this curious flirtation with experimentalism doesn't distract it from its main goal, which is to act as a thrill ride. That it ends up a dance single and not a noise rock jam is, I suspect, a boon or a bust depending on each viewer's inclination and mood at the time. It's nothing game changing, but it's got a good beat and you can dance to it.