Many in the industry suggest mixing sound for film at 85dB, but in the smaller room environments this can be too loud, causing ear fatigue and other issues. I suggest mixing at 79dB, and when you think you are done, do one or two passes through at 85dB to double check. This should give you a fairly accurate listening environment without the ear fatigue. In larger rooms you may be able to mix at the 85dB level, and in really big rooms you would calibrate each speaker individually. Now when you listen through the program if elements are too loud bring them down, too quiet bring them up. Trust your ears.
If your picture editor is experienced and good at audio restoration you may actually get a dialog track with good levels, and already cleaned up. There is a good chance though that you are going to be left with the job of restoring the audio quality of the dialog track. Here is a short list of the tasks you will undertake, and the order they should fall in:
Importing Session Files
Most likely your picture editor is going to want to send you an OMF or AAF session file with all the audio data. I have worked in three different DAW’s for post production, Ableton Live, Logic Pro, and Pro Tools, and unfortunately only one of the three imports the OMF and AAF files properly…. Pro Tools. Logic is super close. It imports the tracks and the automation, but not the track names, and that can be super frustrating when you have a large session to work on. Ableton Live doesn’t accept these files at all. Pro Tools shows it’s strength in this area. You import the AAF and there you are with all your tracks laid out with the proper names, and all the automation. This can save you a bunch of time.
If the price tag on Pro Tools is too steep, Logic would be my second choice for audio post production, but you will have some cleanup and organizing to do before you really get to start working. Another weakness with Logic is that if you have to insert frames because the director or editor decided to make changes to the picture lock, Logic can be a little finicky with the way it handles the frame count and insertion. For better or worse this is another area where Pro Tools excels. Just type in the frame numbers and you are good to go when working with spotting in Pro Tools.
Normalize your dialog
Level out all the dialog so it falls within the target range of -11db to -10db. This is a rough guide point, quiet sections may fall a tiny bit below this to -11.5db, but I suggest not going above -10db for your max dialog peak. Measure this with the dialog solo’d. I tend to group all of the dialog tracks to a bus for control of all of them in case I need it, this can be handy to use as a side chaining source key for tastefully compressing background music.
Once you have normalized all your dialog to that -11db to -10db range, you will most likely find a lot of undesirable background noise. This will vary from generic noise floor hiss, to crew talking in the background, or crew dropping things in the middle of a scene, or crew footsteps in the background during quiet scenes, or large motor vehicles driving by, or planes flying over head, generators or air conditioners kicking on, or even the mic cable not being taped to the boom and rattling against it… I think you get the picture.
ADR or Looping
After all your restoration work you may encounter some dialog that is just irreparable and completely unusable. You will have to get the actors into the studio to re-record their lines. This is called ADR (Automatic Dialog Replacement) or looping. This is a trick in itself as you need the new dialog to sync up with the lip movements on screen. I will write more about the process of ADR at a later date, but in summary, you will need a quiet studio, preferably a larger room to record in, a screen for the actor to watch the scenes on, and preferably a screen with the wave files visible to them as well.
Foley | Sound Design | Music
Now that you have fixed all the problems with your dialog tracks, you can reward yourself by doing some Foley | sound design work and Music. Make those fight scenes really grab your attention, add footsteps, doors closing, car and other vehicle sounds. Have fun in this section. You can then add music to capture the emotion of the audience. I would suggest that your max peak for your loudest effects be around -5db for the loudest stuff like explosions and gunshots, but mix to your preference.
Audio Mix Level Targets
One of the most popular questions that I see when people are getting into mixing sound for film is; “What should my levels be?”. One test you can do to answer the question for yourself, or to validate the answer I provide here is to pull a .mov or .mp4 version of a Hollywood large budget movie into your DAW and watch your loudness meters as the movie plays. You can see the levels where the dialog sits at, the music, and the sound effects.
Here is what I learned from that experience summarized in one easily digestible chunk. All levels dBFS peak.
- Max peak: -2db (This was absolute on everything that I tested probably because of BS1770-3/A85)
- Loud sound effects (explosions, gunshots): -3db to -2db
- Louder soundtrack or score music not competing with dialog max: -5db to -4db
- Dialog level: -11.5db to -10db
Use these as reference points. Start with getting your dialog levels first, and then build everything else around that. These are not hard rules (aside from the max peak of -2db) but a foundation on which to build your audio mix. I found this to work consistently well for me, and puts me right in the ball park to be compliant with the broadcast standards established by the CALM act (BS1770-3/A85)
You can watch my interview on Audio Post Production and its importance –