One of the first thing that I start to Premix when I get a movie is the dialogue section. The reason is simple. I need to know what is being said, what is the levels, drama, etc is so that later when the music or FX are being premixed, I can make sure that there aren’t clashes. It also helps me get an emotion of the scene before hand so that while mixing, I can made decisions as to wether keeping the dialogue or removing it or fading it into reverbs is needed on a scene.
There are 2 kinds of dialogues that we deal with in todays film world. The Production (Sync Sound) and the Dubbed (ADR). Both of these use a different approach while premixing.
I personally prefer to have the raw dialogues with me during the mix. The reason is that I don’t tend to clean up the dialogues a lot unless and until there is legibility issues and there are issues in ADRing the dialogue. Also, while cleaning the dialogues, if the reference of the ambience and music is not present, there is an unconscious tendency to over clean the dialogue. Cleaning dialogues requires very good experience to maintain consistent tonality. The best way to start out is not to clean, but to edit the dialogue. Why?
While editing dialogues, the editor can get a very good idea of what words have legibility issues, what needs to be ADR’d, what needs to be cleaned etc. It is very important to have consistent track layout. Usually, there are 2 ways dialogue editors provide the tracks. Either split as Lapel and Boom and alternating the clips based on the cuts or scenes (known as checker boarding) or splitting the mics according to characters. (Checkerboarding has a lot of advantages especially in Pro Tools 11 due to the new feature of Dynamic Plugin Processing that saves CPU and allows for better usage of plugins.) Each has its advantages and disadvantages. The advantage of having Lapel-Boom split makes it easier to have consistent tonality on scenes. This is because, usually there is very little requirement to change radical eq in one scene, as the location, micing and the ambience wont change drastically (unless this is an unusually noisy scene or a scene shot over days or time). The disadvantage is the difficulty in trying to maintain the tonal consistency if there is an ADR for a character.
One thing that all starting engineers or mixers alike should keep in mind is the phase coherence between the boom and the lapel. If you are monitoring the boom and lapel together and suddenly find a dip in body or low mids, look for the phase between boom and lapel. There is 80% chance that they may be slipped or even out of phase. Many times in my experience, just nudging to fix the phase between the Lapel and Boom itself fixes the tonality without much EQ. A quick way to check this is using the Phasescope on an plugin (PAZ, Avid, Bluecat, Ozone etc). For this, I route the Lapel and boom tracks to a stereo bus in addition to routing it to the dialogue output. The Lapel would be panned left, and the boom on the right, and a phase scope can be put on the bus master. (I sometimes call the bus dPhase). A quick look on the plugin will show if there is phasing issues. Another advantage is, if there is a tilting towards any side, you can immediately know if the lapel is dominating or the boom is. So, tilt to left means lapel dominates and vice versa. This is a quick way to check balances.
That being said, I do have a preference on the choice of track I would use for a dialogue delivery. This choice is defined by a number of factors. Usually for closeup and mid size shots, I would prefer using the lapels. For a bit of distance I would use the boom if that is clear. I song usually mix the two together, but that is a personal preference. I would mix it if i find that lapel lacks body or if the boom lacks brightness. In this case, the balancing between the tracks is done based on what is less. So, if the boom is better and lacks upper mids or highs, I would blend in the lapel (after phase correction) to achieve this before I start to EQ.
On the count of EQ, I always make sure to use the same kind of EQ on the boom and lapel tracks. The reason is to avoid plugin latency and slippage between the tracks. But now with a better engine in Pro Tools, and better optimisations, that is taken care of. There is another reason, and a very important reason I use the same kinds of EQ. Different EQs have different algorithms to achieve what they do. This means that while doing a boost or cut or Q change, the phase may differ between the tracks if the EQs are different. This is not a consistent effect and can easily slip away unnoticed, but it happens. This can be easily avoided.
I set up my session with EQ on all the tracks and a separate 5.0 output for the Dub (ADR) and the Production. They have different master EQ and Compression. I sometimes use the Console for this and sometimes I do it in the box. This also gives me Separation in the final mix. I don’t have a preference for colouring EQs in dialogues. I always use the EQIII and the Sonnox plugins for EQ. The reason is that they are simple and do the job excellently. For Production dialogues, if there are some that need more bands and Q, use the Q10. I always do subtractive eq on individual tracks and the boost on the master. I try and get as much clarity and tonality using subtractive EQ. It is not a rule though as sometimes you may need to increase body or brightness.
All my cleanup is done using a C4 plugin. The reason is that this is a multi band compressor. It is important to remember, that we don’t need to eliminate noise, we only need to reduce it as much is needed for legibility. Being a multi band compressor, the C4 is a very good tool to maintain the dialogue frequencies while reducing a rumble or high freq wind or leaves. it also keeps the dialogue in its natural location, source and tone, and very little is needed for coverup ambience. This is why I would try and maintain the real dialogues without cleaning all the way till the final mix so that I have an idea of what I need to keep, remove or clean. A good example of this is the dialogues in the film My Friend Pinto. This has absolutely no cleanup of dialogues except for some C4 and eq. All the rest is done using fill ambiences and compression.
This brings me to another very important thing in dialogues. Fill tracks. Fill tracks are basically mono ambience of the same location that is sent to the centre channel. This is very helpful in many ways. It helps cover up for ADR when needed by avoiding the jump in location noise. More importantly, it helps smoothen the different dialogue takes and keeps things in a single timezone and tone. Fill tracks are also very important when you make the MnE (Music and effects mix i.e., the final mix without dialogue tracks for language dubs.) The fill tracks make it easy for the version mixes to sound similar to the original by having almost same location noise. This is also another reason why ambience mixes are better done after dialogue premixes. I have a habit of keeping as much ambience in the centre as possible and spreading out different aspects to the other channels. I will write more about this in a later post. But it is important to be mentioned now so that I don’t forget it later.
ADR also is of two types. Complete movies that are ADR’d and patch ADRs for Sync sound films. They are similar in approach but not as complicated as the Production dialogue. It carries the advantage of being able to pan dialogues if needed as the noise itself isn’t being panned along with the dialogue. (There are 2 schools of thoughts to this also which I will discuss later. ) It may not be apparent, or may seem as a surprise, but it is ADR that tends to have more EQ than Production dialogues. This is because scenes may be dubbed in one go. This may have an effect on the delivery of the actor (he may be tired or sore throat etc.). Good Dialogue supervisors can get amazing performances from actors and I have seen this change the persona of the character on screen too. But as mixers, it is our job to make sure that they go hand in hand with the screen. It sometimes becomes important to cut different eqs on interior and exterior scenes etc which are usually taken care of in the Production dialogue because of mic placements. In the case of an ADR only film, it is very helpful to have the dialogue Character split, i.e. different characters on different tracks. It is equally important to maintain the track layout consistently across reels. This means that if the hero dialogue comes on track 1 in reel 1, it shouldn’t be in track 7 in reel 2. This makes it difficult to mix the tracks. ADRs don’t tend to have radical compression as the production dialogues because they are controlled in recording. That being said, there are instances where this changes.
Nowadays, it is common to have the ADR done on 2 simultaneous mics. One may be a boom and the other may be close miked. Or there may be two mics in slight separation. Many people differ with me on this, but I don’t prefer to mix both mics simultaneously. That is because, it has Phasing issues because miking may not follow the 1:3 ratio rule. Another reason is that tonality differs on both mics and they do not necessarily blend well. The advantage is in using a distant mic for louder scenes and close mics for intimate deliveries, but rarely together. It is far easier to shape the tone of a mic rather than fix phasing issues among the mics.
In the next part, I will write about my mixing methodology and approach for dialogues.