We got to sit down with Ren Klyce, Oscar-nominated sound designer (Social Network, The Curious Case of Benjamin Button), to ask a few questions about the technical and creative sides of the sound design process for The Girl With the Dragon Tattoo.
We learned how Klyce (pictured below) and his team created the soundscapes of freezing cold and blind terror for the film. Here's how it went:
POST: You've been working with David Fincher for a while. You did Social Network and I imagine there has been an evolution and a progression in terms of the sound. For Tattoo how did you approach it from a sound design standpoint?
REN KLYCE: “The first initial reaction was of the locations and the temperature. The film takes place in Sweden and in very cold climates. So the first concern was making sure that all the sounds we were stitching into the fabric of the soundtrack were of the correct place and not using American recorded sounds and putting them over images that belong in a foreign country.
“When David Fincher and I were discussing how to collect the sounds, and David was already camped out in [Stockholm], he started commenting on the sounds of the city. He would say, ‘We're in the Slussen district and the sounds of the passing trains are very unique, and you have to collect the sounds and this area has a very unique tone to it.’"
POST: So he was sharing his feelings of the environment?
KLYCE: “Correct, because he was on location. When the film is being shot, the primary concern of principal photography in terms of sound is to only focus on collecting the dialogue as cleanly as possible without any other sounds and the reason for that is later on in editing. When Angus [Wall] and Kirk [Baxter] are editing the picture if there are sounds married to the dialogue and they need to cut from one line to the next you can't have a sound in the background because it'll draw attention to itself. It'll sound very spliced. So when Fincher is focusing on his audio in production with the actors they try and get rid of all the sounds. But later we have to add them all back in and it's during those conversations that he said, ‘You've gotta come out to Sweden and you have to record these trains, these ambiences. You have to record the, ice, the show, the trees with icicles swaying in the breeze, these types of sounds.’ And so I was all set to go out, but then I thought, you know Rune Palving, who lives nearby would be a far better person. So we hired on Rune to go out and collect these sounds. So that was wonderful having someone who was already an expert in the location and who knew where to go to collect the sounds.
“So that was the first concern, the second concern was, as I mentioned earlier, the feeling of temperature. A lot of the film takes place in a cold rickety cabin with gale force winds blowing across the tundra and icicles forming on the door. Every time Daniel Craig's character, Blomkvist would open a door Fincher wanted it to go KRISHHSS like the ice had fused the door shut and he wanted every moment to just feel like, ‘Holy crap it's freezing here.’”
POST: So you really wanted you to shiver when you heard these sounds?
KLYCE: “That's correct. So all of those sounds are to be collected, and there are some big moments sound wise in terms of the larger scenes. There's two larger scenes that we are focusing on. One is a motorcycle chase in which Lisbeth Salander (the main character played by Rooney Mara) is chasing after Stellan Skarsgård's character and it's a very intense chase sequence where she's trying to overtake him on a bridge; it's quite exciting with a lot of cuts. Perspectives from being on the motorcycle or being up in the sky or different angles. The whole chase ends in a very dramatic car crash sequence with an explosion.
“The Highlights on this film in terms of sound would be that moment and a wonderful sequence that Fincher actually shot in a subway in Sweden where Salander is accosted by a thief who steals her bag and they have a scuffle, a big fight on a moving escalator. It's a pure sound moment. Fincher wanted to have only sound carry that sequence and so he wanted the sounds of the trains to kind of start of realistic and fuse into kind of a screeching terror noise but using the sounds of the subways and the escalator and what you would imagine would be going through the main characters head as she's being accosted and then in this fight.”
POST: Did you bring all your own people and gear from the US to Sweden?
KLYCE: “David Fincher certainly did in terms of his filming crew, but in terms of the sound crew no we did not. We only hired one person to go out and spend a couple weeks collecting sounds. So what I did is before David started shooting I read the script and I spotted what I thought would be sounds we would need based on Steven Zaillian's script and made extensive notes like, ‘Try going out to find a cabin in the middle of nowhere during winter and record it or go to a subway station at midnight and record the trains coming in and out or go to a rooftop and record the morning atmosphere, go to this cafe, record these people, record their voices. So we tried to record as many different sounds as we could that were specific to the country. Particularly sounds that you don't think are unique, like the sound of the telephone ring. In Sweden they sound completely different than they do in the US.”
POST: How long was this list? Hundreds of sounds? Thousands?
KLYCE: “Pretty long. There are quite a few cues. I would say at least a hundred cues if not more.”
POST: How many hours of audio did you have to work with?
KLYCE: “We have hours and hours of footage with the sound of course, so a lot of it is gold mining and you don't know what you're going to get. There are certain scenes and sequences where the main character, Blomkvist, has been wrongfully accused and being ridiculed in the media for liable, so we're working on a sound right now that's sort of the festering annoying buzz like insects. That is the press and the media and the sensationalism of what television has become, so that's another sort of texture that Fincher is about. Right now we're researching and getting music sent to us that's used on shows that would be the equivalent of Entertainment Tonight or Geraldo Rivera. Those kinds of ridiculous stingers. We're weaving that into the soundtrack as well so that we become aware that this is a huge story. We never see anyone else in the world paying attention to it other than our few main characters, so David really wanted to make sure that it sounded like there was a media blitz happening. So sound really can contribute to that aspect of the film even though we don't see the world paying attention to this media story, we somehow get the sense that the world is quite aware of it.”
POST: Are there cases in the movie where you're hearing things that you aren't actually seeing?
KLYCE: “Absolutely. There are some pretty dark sequences. There's a few rape sequences that have quite intense imagery and quite intense sound that accompanies these rape sequences. So one of the characters, the rapist (Bjurman) invites Salander to his apartment. He's a social worker and she thinks, ‘Oh I've gotta give this guy another blow job so he can give me another 20 bucks’ so I can eat. Really what he does is something much worse. He handcuffs her and ties her to a bed and rapes her quite violently, so all the way leading up to this rape sequence is her journey to get to his home and it starts with her arriving at his home and again the sound of the trains in Sweden, the click-clacking click-clacking click-clacking has a tension, a rhythm from the click-clacks of the train. When she enters and gets drawn deeper and deeper into his lair, into his bedroom the sound of the muted subway can be heard, and it's almost as if that sound is there, it's loud enough that other people won't be able to hear her screaming. It's an organic source, but it's a sound that would mask someone's suffering.”
POST: What's the favorite sound you've recorded for this film so far and what did you use?
KLYCE: “The motorcycle sound. The character Lisbeth rides a motorcycle; it's a custom Honda 350 from 1974 with a dual-side exhaust and a modified exhaust. We knew that the sound was very important to her character and of course the film and for David. She revs it and she expresses emotion with the motorcycle. When she arrives she's very loud with it, even though her character's kind of timid, her voice can almost be spoken through the sound of the motorbike. We actually ended up buying a Honda 350 and we recorded it, and we recorded a Honda 450. Then we asked production if we could get the actual motorcycle that she road and they sent it to us, which was great. Then we spotted the film for every single angle we'd need a recording from and then we mapped out the sequence.
“So say, for example, she starts, she revs it, she drives off way in the distance at 50mph. She whips past us on the next angle and then she arrives. We designed the records to capture each one of these cues in sequence because the tone of the motorcycle is very important to keep consistent. We ended up having four Sound Devices synchronized together. We'd have one person with one set of stereo microphones, two Neumann KM 140s set up with the microphones — start and away. Then someone midway, maybe 50 feet away there'd be another person with another Sound Devices 744 audio recorder with dual Senheiser 416s, one locked off so you get the bi without moving the microphone and then the other microphone would swish pan and track the motorbike. The third person would then be on the receiving end with maybe two K1s once again one locked off and one tracking to receive the motorcycle as it came in.
‘Then, in addition to those three recorders with a total of six mics, we then microphoned the motorcycle itself with four microphones using a 4-channel Sound Devices recorder. We used two Countryman lavaliers, we taped one onto the engine block to get the high frequency clack of the motor the second mic we put on the exhaust port. The third mic was a dynamic mic, I can't remember which mic we used, something that could handle a lot of high SPL, and we put that on the back of the motorcycle rider’s jacket, literally taped to his back. Then the last mic was a lav taped to his forehead so the wind would hit it. So that's 10 channels of a synchronized recording and we brought those all into Pro Tools and we would edit the various recordings together.”
POST: For the actual dialogue do you have a favorite radio mic you use that's really clean?
KLYCE: “Mark Weingarten is the production sound recorder. He uses the Sanken COS11 wireless and that has a great sound. He wires the actors up with the wireless mics and then he has two Sanken analog shotgun mics on two booms, and Michael Primmer, our boom operator, will be booming with one character and then the cable operator will be booming the secondary character on the two shot. Mark will have both characters wired but he'll be doing, usually a five channel mix. So there will be principal, Daniel Craig on a lavalier the CRS11, Mara on the same, two booms and a mix track. So usually five, four channels and then Mark Weingarten will be performing a live mix for Fincher and for the guide track. But we will actually have all four of those microphones separate.
“When we get all those microphones in the timing of all those microphones are not quite in phase. In other words the boom, because it's four feet away from the person’s face, versus the lav, which is six inches away. That timing discrepancy will create a phenomenon known as phasing. It sounds wrong. It's what we call comb filtering; it has a very ugly annoying doubling sound. So very carefully once we get it into our workstations our dialogue crew, by hand, zooms into the boom microphone and ever so slightly, within microns, shifts it earlier in time so it lines up with the lavalier microphones, so that they are no longer out of phase.”
POST: What does he record all those microphones to on set?
KLYCE: “Mark Weingarten uses the Deva from Zaxcom.”
POST: So with all this sound recorded, how many tracks of categories of tracks do you end up with in your final mix project.
KLYCE: “Our final mix right now is laid out fairly extensively. We have a dialogue pre-dub, which is four eight channel dialogue premixes that have all the characters split out on different tracks. Separate reverb mono reverb tracks for each one of the characters and then our ADR is another series of pre-dubs, eight channels each. Then we have a loop group, which is an eight channel pre-dub that is actually two LCRs and two monos. Our sound effects are actually 10 eight-channel premixes; each of the eight channel premixes is actually two premixes; five zero and an LCR and then our Foleys are four eight channel pre-dubs. We have two LCRs for footsteps, four for props, and one for cloth. Music isn't premixed so each cue is laid out differently. For Trent Reznor's and Atticus Ross's score they deliver anywhere from four to 12. I'm mixing a cue right now that has 10 stems. Each one of the music elements Trent will break into bass, fills, piano one, piano two, synth one, synth two, melody, pad, strings that kind of thing. Then we have in addition to that our source music, which is music that Trent and Atticus haven't composed but it's music that happens in the background at the party, the Christmas music and that sort of thing. For those tracks we have the original stereo tracks off of the CD for, example, but then we ‘thump’ each one or ‘world-ise’ each one through different plug-ins to make them sound like they're in a room, or coming off of a radio, or a shitty speaker. Then we blend those all together to create that sound, so a lot of different food groups. I hope that it's not too confusing.”
POST: So is that everything?
KLYCE: “There are actually more channels coming into the console as well from the machines that are called ‘fix machines.’ So in addition to everything that we've premixed when you're working on the film you kind of go, ‘You know what, we need something right there.’ So we have another machine that has 16 outputs that can provide sound effects, another 16 tracks of Foley, dialogue ADR, music, etc. So there are additional Pro Tools workstations that are sort of hanging off the console if you will, that are on the ready to provide more sounds in case something isn't working.”
POST: Where are you doing al of this work?
KLYCE: “We're at Skywalker Sound.”
POST: What's different about working on this film than working on Social Network?
KLYCE: “Sound has a larger role in this film due to the constant changing of locations that Fincher is weaving into his storytelling. Through his visual storytelling, the characters, for example, have conversations that move through space. So you follow a character as they walk through a doorway into the street onto a train and so on and so forth. So that is quite challenging for us because it means that we have all these set-ups that we are required to premix and create a texture for each one. On Social Network, you might be in a courtroom with the deposition scene but then you're in that deposition scene for a fairly long stretch of time. In ‘Dragon’ those long stretches of time are fewer and further between and instead there are many cross cutting scenes between the Lisbeth character and the Blomkvist character, cut back and forth as they go through their different locations before they meet ultimately.
“The funny thing about this film is that we, as the viewer, are introduced to these characters but they don't actually come together and meet until the fifth reel of the film. So sound has to sort of amp up the difference between these characters, and the music and the picture edit, so that the rhythm of this cross cutting when they finally get to meet, kind of 'lands.' So that's quite different in terms of sound structure than Social Network."
POST: What was the most challenging part of doing this mix or was, that it?
KLYCE : “The most challenging scenes I would say are the violent sequences. There are three sequences that are quite violent involving rape, and those are quite challenging because the sound has to amplify the tension and the horror, the fear of the character who is being raped. In one case it's Lisbeth's character who's being raped; in the other case it's Bjurman; we call it the ‘Revenge Rape.’ Those are very difficult for us to work on because no one really wants to see that. It's one thing to watch a film, and you know you watch it all the way through, and you only have to watch the rape once. You can close your eyes if it bothers you, but for us we can't close our eyes. We have to watch it, again, and again, and perfect the sound over and over and over again. Whether it's someone screaming in pain, or the things that are binding this person during a rape sequence, and we have to enhance those sounds. That's been very difficult to do just emotionally because it's not pleasant subject matter.”