Moulin Rouge

On the set of Moulin Rouge, Fox Studios, Sydney.

TM: How long have you been recording sound ?

GS: I joined the ABC in 1980, started in the backrooms doing transfers, then in the dubbing suites, even rolling the projector. Then I moved on to the occasional foray into the field as an assistant, recording on a few news shifts.

I then swung boom for Ben Osmo and Chris Alderton, who also were at the ABC at that time. I then concentrated on recording news and eventually moved on to more sophisticated programs like current affairs programs. I spent about 4 years recording 4 Corners and then I went to drama, recording programs like Bodysurfer and other ABC co-productions.

TM: What past projects have you recorded?

GS: Recent projects I have recorded include My Mother Frank, Me Myself I, Passion and Babe2.

TM: What interested you in taking on this film?

GS: My initial interest in sound developed from music.I've always played an instrument and worked in and around bands…even now,so it feels natural to work with music and I suppose this film brings everything neatly together.

Moulin rouge has many challenges across a wide spectrum , all the traditional filmic ones plus musical. It might help to explain what we were trying to achieve in the first instance. Basically,the concern was that crossing from sound recorded on location to prerecorded music and singing would always be jarring. It's not just a quality issue but also a question of performance. An actor singing in a studio will always be significantly different to an on set rendition…even if they are good singers.

The solution appears to be twofold. Matching the technical quality of on set to off set (or vice versa), and, more importantly, focusing on live performance. This basic mindset was the driving force behind my approach to this film.

The second part of this twofold solution lay more in the hands of Baz and his musical director, Marius DeVries. Every musical number has some live components, generally the beginnings and endings. So, for example, some dialogue may slowly become more musical and rythmic until the actual song starts.

These areas were like twilight zones,stuck between full playback and live recording and in fact we almost always did both versions and also hybrids of the two. Baz would never be locked into one or the other because everything is completely performance driven and he recognises no technical boundaries, or logistical, for that matter. This flexibility of attitude necessitated many and varied situations for music cues.

As an added complication, sometimes playback was just Marius playing keyboards off set to give the actors tempo and pitch and occasional timing corrections. So an actor would be saying dialogue and at a given cue they would hear through an earpiece a music cue played on keyboards, start singing and eventually full playback would roll in. It was the hybrid shots that were the real killers.

TM: How involved has the director been in the production of the soundtrack?

GS: Enormously. Baz is just everywhere. He is very keen to get the best of everything. We had long discussions in pre-production about what we were trying to achieve, in terms of music and in terms of dialogue. We talked about overlaps and performances.

Baz is really a performance director. It's interesting, the focus puller is feeling like a sound recordist, because he knows that Baz is going to use takes irregardless of whether they are soft because he likes the performance.

So we are all in the same boat and it's a great way to work because it's all about performance, it's not about, oh well that's soft or that's off mic, we have to go again.

The technical side of things moves into the background once we start shooting, but that's not to say he doesn't care about the technical stuff.

What happens is that he is so thorough in pre-production, we are all encouraged to be completely thorough and to cover every possibility. Once you get on set there are no excuses. Small things go wrong of course.

An example of the way Baz directs is that he won't necessarily call cut where you think he is going to cut. If he likes a performance, he'll just keep going, and he'll keep going off the script page into a scene that is not written. The other day we were doing a scene where the three actors were gallivanting around and the last line in the scene was, "Okay lets have dinner", and normally he calls cut because that's the end of the scene. Then Baz just yelled out, "Just go and have dinner", so the actors look around and get up to go and have dinner. So all of a sudden the anchor's off and we are all floating free and he gets great moments like that.

Technically that's really challenging because my boom swinger didn't know which way to go, I didn't know which radio microphone to put up, it becomes like a documentary. After that, one can never relax.

TM: How much time were you given as far as pre-production to prepare for the shoot? In regards to sets, costuming, etc.

GS: We were very lucky, we were given about eight weeks pre-production. Some of this was by default because we were delayed a couple of times. I think originally we were given six weeks, but this blossomed out for various reasons. It sounds generous,but we needed every minute because we built a whole recording system from scratch and a lot of the technology was new to me.

Stuart Waller and I did a lot of research into digital recording and came up with a great system. We had many meetings with music people and sound post to discuss the sound pathways and sync issues, it's great when everyone is accessible so early in a film.

David Pearson had his hands full organising costume and art department issues. Because he does the actual radio micing, that job falls to him naturally.

TM: Were there a lot of challenges in regards to sets and costumes?

GS: Nicole presented the greatest challenge, she's basically not wearing to much on the top third of her body and a lot of what she wears is either very tight or very skimpy, so she was a challenge to radio mic.

Harder than the radio mic, was trying to fit her for an earpiece for playback. This is due to the fact that the actor now has to wear a second pack for the receiver. We experimented using an induction loop, but we found that they don't work very well below the neck, so we had one made that sits under Nicole's wig. It was wound using about 5 strands of really thin copper wire and there is a very thin almost invisible wire that runs down the back of her neck or where ever it was out of shot.

Even with this setup we experienced problems because quite often I would need a days notice to put it on, the wig goes on at 5:00 am in the morning and stays on all day. Unless we were told we are definitely doing music and need the earpiece, we would be caught out.

What we did was go back to basic electronics really and wired the whole set. It's the same system in principle that's used in cinemas for the hearing impaired,one giant induction loop incorporating the whole building or room. In our case we used simple telephone wire wrapped around the set and connected it to an amplifier. Stuart would wire up each set a few days before shooting and we never had a problem.

He even designed a small junction box to facilitate connections which we dubbed "The Waller". Place your orders! I have to say that in this situation, I've never heard better results from earpiece work, always plenty of gain, negligible interference problems and best of all….no unsightly bulges on svelte actresses.

The other actors were all pretty standard in terms of costumes, although I tend to think period costumes are generally more difficult to work with than others. It must be the bulk and stiffness.

TM: How much input did you have in modifying the sets for sound in pre production for this film?

GS: Modifying is not really the word for it, what we attempt to do is make the set sound less like a set and that's difficult when sets are constructed mostly from wood and other cheap,fake materials. I often wish I was a construction engineer because I come across the same set of problems on every film and I'd love to have one fool proof solution.

Sets built on the ground are generally not too problematical for sound,but raised sets are truly awkward. Getting rid of the boominess from underneath is the first priority and that's largely a question of money. Density is expensive and construction people also are loathe to do something too solid because it's then hard to pull the set apart for camera position.

Nevertheless, sometimes you have to insist and on Moulin Rouge we tried hard to balance the cost of soundproofing against the cost of losing valuable performance dialogue. Same old dilemma. The art department were always very conscious of sound issues and were good enough to change the motor on the windmill three times for me and on critical occasions wind it by hand.

TM: You mentioned earlier that the sets you were shooting in were very restricted for space, very little room to move. How has that affected you?

GS: That's not my problem mate, that's David's (Boomie), laugh. You need a talented boom swinger, and I've got two of them in David Pearson (1st boom) and Stuart Waller (2nd boom & sound assistant). We've had Mark Wasiutak in as well to help.

The second half of the film we had Leandro Righini join us and he proved to be invaluable. It is tricky because the lighting is really intricate and the sets are incredibly busy, in fact some props are actual antiques worth bucketloads of money.

TM: What would be the percentage of boom to radio microphone you are recording?

GS: At the moment probably 80% boom, but as always it depends on the shot and with anamorphic lenses it can get quite ugly. Not so much in headroom, but just in what is in the shot.

The first scene we did was as complex as we're ever going to do. It was 6 actors, 6 radio mics, 2 booms, the whole music playback debacle, an 8-page scene, with an adlibbing section in the middle of it.

It varies; you just have to pick the eyes out of it. What makes it challenging working with Baz is, you just know that it doesn't matter what the shot necessarily is, the performance is the thing that will really matter to him. So even if it is in a wide shot, if he likes that performance, he'll likely use it that way.

I think that makes it harder for every one, because continuity really becomes paramount at this point, and continuity for sound is probably one of the most important things. It's easy to mic some one, it's infinitely harder to mic some one consistently through out a whole scene, matching shot to shot.

TM: In a situation like that where you a running multiple microphones does recording across 8 tracks of the DA98 make it any easier, or do you still pre-mix a large amount of the production audio?

GS: It helps but it also hinders. What we are doing with the system we have developed is that I mix a composite mix track for the picture editor. So track 1 of the DA98 is my mix of every shot. Then after that I split off microphones to different tracks.

With playback music coming in we split pre vocals onto one track, pre music background onto another track, and time code from play back source onto another track. So the tracks get filled straight away, and with a couple of radio mics you basically have 8 tracks recording.

It helps to split microphones off, but it also can be a bit prohibitive, it becomes a bit difficult. We all like to have good sounding rushs.

For example, if I was doing a shot that started off too wide with an actor coming through the door, and we couldn't get a boom in there, I would put a radio mic on the actor, and then as the actor moved into the foreground I would get rid of the radio and cross to the boom. Now when I am mixing for the composite, I have to get rid of that radio mic, other wise it is going to phase and introduce other problems.

But for the final product it is better to leave it there in case they want the choice of actor's radio mic dialogue at a later date. Because Jill, the picture editor, cuts using the composite audio track from day one, this track must be clean and clear.

I must admit I didn't want to do it that way; I got talked into it a little bit, because I could see it would be giving me problems occasionally.

It's been okay and it is a good discipline. I've always been happier to mix two mics together myself than leave it for some one else. I will leave it if I think it is a problem and I can't get it right, but if I think I can get it right I'll do it. Some dialogue editors love that and others don't.

TM: What were the main technical challenges faced in recording this film?

GS: Matching the quality of the music recording ,multi camera coverage and on set comms. As far as the quality was concerned, I think the system was up to scratch, great preamps, digital mixer etc. I think the dialogue sounds great on the boom and also the radio mics.

The coverage made it devilishly difficult at times, the wider cameras see so much more extras and prop action, it becomes impossible to lock down. Meanwhile the close up camera rolls on oblivious.

On set comms was complicated by the musical nature of the film. The logistics of the music side of things can get really ugly very quickly. So many people need to hear different things all at the same time. We regularly sent different submixes to Baz ,Marius, actors, music continuity,video split,boomies, Simon on playback, producers and even other video crews from EPK.

Digital mixers really help with this, I was able to run eight auxiliaries all with individual submixes. Very handy. The system we designed needed to do all this and more.

Here is a broad overview of the setup……. The core of the system is built around three units,a Spirit 328 digital mixer,a Tascam TMD1000 digital mixer and a DA98 DTRS. All three were connected by two TDIF cables. A protools 882/20 was then used to play back music and run directly into the Tascam mixer with analogue connections. The Spirit was my principal mixer and it had four working banks of faders.

1. Bank one,channels 1-16,mic inputs,booms etc.

2. Bank 2,channels 17-32,TDIF returns from DA98 and TMD1000

3. Bank 3, master outputs to 8 busses and auxiliaries and fx sends

4. Bank 4, midi controllers via MOTU interface from protools. Everything therefore appears on my main console, even fader controls for the protools playback music in case that needed pre mixing.

The flexibility is enormous because auxiliary sends can be tapped out from any point on either mixer, remember, the TDIF allows the two mixers to appear on each others banks. The Tascam digital mixer was primarily the PA/Music mixer with protools patched directly into it.

Another challenge was simply the variety of setups which we had during the shoot and I must say that one of the most valuable functions of digital mixers is snapshot memory. With the Spirit mixer I have a library of pre-programmed snap shot memories, all labelled with different scene numbers. The snap shot memories remember all the parameters, including word clock settings, sample rate setting, equalisation and routing pathways. Occasionally Baz would want to rerecord song lines within protools on set for immediate use. I can send microphones out the fx lines via SPDIF from the Spirit into the Pro Tools, so we can record directly into Pro Tools if we want to so we utilised this function a number of times. Simon Leadly,the music editor would then do a quick edit and hey presto! Instant playback. To top it off,we would burn a CD of the new version as well.

TM: How many auxiliaries are you sending all up?

GS: 8 in total. 4 auxs out of the Tascam and 4 auxs out of the Spirit. There are two PA systems fed by the mixer monitor output and the master stereo output. So really all up I am sending 10 if you include the PA sends.

It sounds weird but it's not. The way we have it set up is very software orientated with snap shot pre-set menus, which we can recall at any time.

What I send daily are 3 Comtek frequencies, 3 IFB frequencies, we have 4 speakers all run on separate lines (Mackie active speakers), and then we have another Peavy PA system.

I can customise the whole PA section as well, for example if the dancers need more of the thumping bass in the background I can send them a different mix. If they don't want so much of the vocals and just the beat, I can accommodate them. It has proved quite valuable to be able to do a little sub mix for the PA.

Baz gets his monitor feed via me, post Michael Taylor who operates the video split, through an IFB feed, that's Aux 1 on the Spirit. So I send what I want Video split to record and he sends it onto Baz to hear and then when Video split is in playback mode, Baz gets instant playback sound.

Also if he wants to hear something from protools, it's very easy to wind that in. We started off sending left and right to the video split but it got a bit painful to do that, so we finally decided to send just a mono composite feed. Then when the video-split is in playback his audio signal comes back to me and I feed that into my stereo two return on the mixer. I can then call it up and route it to where ever I want. So any combination of people can monitor playback.

Aux 2 is my private line to my boomies. That is patched into a different IFB feed. They generally don't want to hear the music because they hear it through the speakers any way.

Aux 3 feeds the music continuity person's Comtek and they receive their own separate mix. Vocals and a little bit of music.

Aux 4 is more of a public one for the producers and the dialogue coach. The four auxiliaries on the Tascam mixer are used to service the music department in general,for example Marius's keyboard foldback.

TM: What about the track layout on the DA98?

GS: The track lay out when recording on the DA98 is as follows:

Tk 1: Composite mix

Tks 2 - 5: Boom and radio mic channels.

Tk 6: Pre recorded vocals.

Tk 7: Pre recorded music

Tk 8: Time code from the pre recorded material (Pro Tools).

The reason for this track allocation is because in rushs we split down to four tracks in telecine.

Track 1 is composite, track 2 is pre-recorded vocals, track 3 is pre-recorded music, and track 4 is what ever I specify.

I sit in rushes at night in front of yet another mixing board and mix rushes. If we want hear the live performance we wind that in, if we want to hear the pre-recorded material we can wind that in. It is worth doing. There were a couple of bugs earlier on getting the right tracks, but now it is as smooth as clockwork.

TM: Working so heavily in the digital realm, did you experience any problems interfacing any of the digital equipment?

GS: Initially we had a problem with clocking the Spirit mixer. Although it is a great mixer ,it generates less than perfect word clock. Bringing in an external word clock source solved this problem. This was a DBX 586 dual vacuum tube external preamp. This DBX unit is now the master clock source at the moment. This clocking problem is now fixed and Jands have been very cooperative in this regard.

TM: After each take do you replay the DA98 for safety checking purposes or are the confidence heads enough?

GS: I am led to believe when recording on the DA98 in confidence mode you can trust what you are monitoring. At the same time I record a safety copy on a twin dat deck I have set up. I had to use them once because I rolled late. So it works.

TM: How much extra equipment did you have to purchase to record this film?

GS: Not much really,both mixers were mine but the DTRS was purchased by Bazmark. Some IFBs were bought and lots of odds and ends.Stuart can make anything out of a bit of string really.

TM: Are you using the Internet for overseas file transfers?

GS: Video split, the music people and I are all connected via the Ethernet and we do transfer music files. The problem with the Internet is that it is just not secure enough in the end and bandwidth is not great enough.

Marius also ended up staying in Australia so we didn't have a problem with distances. All his stuff comes from Iona, the production office, and it's ten minutes drive.

We were going to set up a server and do all that sort of stuff, but I think it's just a little too early in terms of the comfort factor of that technology. Not just that the connections break occasionally, but it's not a universal language just yet.

TM: Is there any location recording required for this film?

GS: No, it's all studio.

TM: How are the studios to work in?

GS: Yeah, they're alright. There are a couple of little problems. Some of the air conditioning is faulty, when you switch it off it rattles for the next three minutes, which can be a stitch.

When it is windy, the wind howls through the doors, even when the doors are shut.

We had problems with the dimmer racks, but they are all outside now. For the big lighting set ups they are using more and more dimmers and they have snuck their way on to film sets, by default, one at a time, until all of a sudden there is an army of them. If they were all in here you couldn't record any usable sound I think. So we had them shipped out.

TM: How involved have post production audio been?

GS: They are hugely involved. That is one of the best things about this job that Soundfirm and Trackdown have been involved since day one. We have all been in this together. We discussed the whole postproduction path way and what's required from me.

I am not saying they are telling me what to do, but they have a very strong input into how they want the product delivered and I have had input into their side, it's been really good.

For example Roger suggested that I obtain some Danish Pro Audio radio mic heads, which was a good suggestion. I purchased them and they are great.

TM: What would be the percentage of playback to sync recording?

GS: About 45 percent would be music. There is also a lot of talking in it. There is a lot of everything in it. If you think of old time musicals with fast-talking and lots of action and then singing, that's what it's like.

A number of times we will do two versions of a scene, first being totally playback and the second will be live singing. Baz will try anything that will tickle the creative side of people.

TM: How easy did you find the DA98 to use?

GS: Ultimately it was fine, but it's not what I'd call intuitive. The menu system is overly complicated but I guess that's the plague for a lot of equipment these days.

TM: With the complexity of feature film recording today, does a 2 channel Dat have enough tracks? Do recordists need to be looking at 4 tracks as with the DEVA or 8 tracks as with the DA98?

GS: I don't think there is a clear trajectory there; I think it depends on the project.

The last couple of films I have done, My Mother Frank and Me Myself I, I could have done them in mono. It really was one clean track, because I mixed mics together, it was easy to do and it sounded good.

This is a different kettle of fish. This is almost an OB job. In fact on Babe2 you might as well have driven in a couple of sound OB vans.

I guess what I'm saying is that it's horses for courses. It is true however that things like comms and the complexity of sound post have affected us in the field. It's very hard to monitor more than two tracks for film recording, I only have two ears but maybe in the future we'll have more grafted on. I look forward to it…a new breed of super sound recordist.

TM: Do you do much processing of your sound, for example compression or equalization?

GS: No, very little. I roll off a bit of bottom end just to get rid of some of the thumpy stuff and control the dynamics. I will occasionally do a bit of limiting if I feel it is absolutely necessary.

TM: Do you have favourite microphones you like to use?

GS: I am a Schoeps fan. I use Schoeps for several reasons. One is that I think they sound really modern, and I don't quite know how to explain that, except they sound modern.

They are a more forgiving microphone, which is great for some one like Nicole who tends to do really fast head turns, she is quick, even the focus puller is saying, "man, she's quick".

So that helps David, it helps me, we have a better time and we get the results.

I did try one of the Neumann's on her through a warmer pre-amp and although it sounded ever so slightly rounder, it would have been a really hard microphone to record her with because it is more directional, harder to swing, and I also would have had to match every body else as well.

As far as lapel microphones I am using Danish Pro Audio. I think they are great and they go well with my Lectrosonics gear.

TM: Which headphones do you prefer when recording?

GS: Sennheiser HD25, with velvet ear cuffs. Naturally.

TM: Any final comments you would like to make?

GS: You can have all the technology in the world, but when you're shooting, you rely on people, and David Pearson and Stewart Waller are fantastic boom operators.

The simple things, like slipping that piece of carpet underneath the shoes or watching out for that extraneous sound on set. Even the way they talk to the first AD and ask for things.

Baz now understands what our requirements are without us forcing them down his throat.

The psychology of the set is really the most important thing.

Another thing that's worth mentioning is to do with the relationship you develop with on set crew and directors. Increasingly I find myself parked miles away from the action because of the bulk of equipment needed on jobs like this. Although I have comms to my onset boomies, it's not quite the same as being there.

I find that if there is a point to make or a question to be raised, nothing beats physical presence. When I first started drama, I used to sit on the floor, on set, with the nagra in my lap, now people see me in the lunch cue and wonder who I am.

TM: Guntis, thanks for making time to talk to us.