• Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.

"Normalizing" Dialog for Mixing

Hey, everyone.

A few years ago I fell in love with and began studying the editing and post process. Since then I've developed a pretty good knowledge base for editing, color work, VFX, ect., but sound mixing has always been my biggest challenge.

Though I understand the basic concepts of mixing, my trouble lies in getting my dialog to a place where it is consistent in volume and ready to be mixed with SFX and music.

I know various applications and NLE's have "Normalizing" features which are supposed to do this, but they hardly, if ever, work, and I know that this isn't the best way to go about it. I know compressor/limiters can be used for this, but they are mind-numbingly confusing, and I have heard they should only be used sparingly.

My technique as of now is to export my audio as a WAV and drop it into an application called The Levelator, which does a surprisingly good job, but I know that's also far from an efficient way to work as it makes things difficult to change or control.

Another technique I have used is to simply keyframe the volume of the clip, but that's extremely time-consuming. I'm fine with time-consuming if it gets the job done, but I know there must be a better way that I am just missing.

Thanks for your time, everyone!
-Cliff
 
"Normalizing" increases (or decreases) the level of the individual sound clip(s). Normalizing is not a "quick fix" to get them all to the same balanced volume; if you have multiple sound clips normalize will only raise them by the same percentage amounts. Compression and/or limiting are also used as a quick fix to level out volumes, but this method quite often creates more problems than it solves, most notably increasing the background noise levels.

I will normalize very quiet audio clips to get them closer to the level of the rest of the audio, but that's about it. I will put a limiter on the dialog sub/buss/stem which only gets "hit" about four or five times over the course of a feature film. I only use compression on VO/narration. I automate the dialog volume levels the old-fashioned way - line by line, word by word, even syllable by syllable if I need to.

Mixing audio for film (and music) is a true art form. I do very nice, solid mixes; but I don't hold a candle to the masters of the art (of course, I don't get the same material to play with. :D ).
 
Thanks for the reply!

So the best way is just to get into the nitty-gritty and keyframe the volume? Man, and I thought I left rotoscoping in After Effects, haha!

Speaking of compression, I've read that adding a little bit to a final master bus can really make the mix pop, is this correct?

Also, any books and/or videos you'd recommend for learning more about audio? The best way is always to learn by doing, but ya' gotta start somewhere. It's funny, I've been doing this stuff for several years now, but with audio I feel like I'm back to square one. Always more to learn.

Thanks again!

-Cliff
 
Speaking of compression, I've read that adding a little bit to a final master bus can really make the mix pop, is this correct?

Only if you do it right. I put a VERY light compressor (1.5:1) on the final mix with a 65-35 FX/Dry blend plus a dither and a few other things - hey, a girl needs to keep a few secrets! :D

Also, any books and/or videos you'd recommend for learning more about audio? The best way is always to learn by doing, but ya' gotta start somewhere.

It depends upon what you want to know. A lot of the technical basics are covered in "The Sound Effects Bible" by Ric Viers - mics, processing, recorders, etc. Ric is a wonderful lunatic; check out his website and YouTube videos. "Sound Design" by David Sonnenschein is pretty good as well; very philosophical and artistic. "The Practical Art of Motion Picture Sound" by David Lewis Yewdall is very nuts an bolts, and "The Foley Grail" is a fun, informative look at the Foley Process. I have yet to find a book that adequately covers mixing - you only learn that at the elbow of a master.
 
In my experience, sound, done right, is the most time consuming part of the entire process.

At the Hollywood level they can spend anywhere from two to ten man-hours per linear minute per sound category - dialog, Foley, sound FX, mix -sometimes even more. So a 100 minute film can take 800 to 4,000 man-hours of work. Obviously a dialog driven drama will take less time than an action/thriller, but even the sound design and mix of a drama can be quite complicated.
 
Damn. That seems completely insane, but I can totally understand why. My girlfriend and I were talking about this last night, and she made a good point that sound is, in many ways, a much more abstract concept than picture. It varies so greatly, and can be so subjective. If I'm compositing a gun shot, well, I know what that's supposed to look like. It's a ball of fire. Done. But there are so many more variables in trying to create the sound of that shot. It's amazing. But hey, this is exactly why I love post! haha

-Cliff
 
Well, think about it...

For dialog editing I have to identify each character and give him/her their own audio track. I then have to clean out all of the background noise behind the dialog and strip out all of the sound between lines of dialog.

Then the sound generated by each character has to be recreated (Foley - footsteps, clothing, prop manipulation, etc.). All of the other sounds also have to be recreated - ambience, vehicles, weapons, doors, etc. Big action films can have over 1,000 audio tracks.
 
You get better and faster at it the more Foley work you do.

It always takes me a while to "get in step" if you'll pardon the pun, but by the time I've worked for a day or two I'm doing a very large percentage in sync with the right performance on the second or third take. It still takes a long time, though. You have to select & set up the Foley props, set up the mics and set levels; call it 5 to 20 minutes, depending upon what's happening in the scene. Then you have to do the actual performances. So for a three minute scene with two characters that's 10 minutes for footsteps for character A, 10 minutes for footsteps for character A, 10 minutes of cloth for character A, 10 minutes of cloth for character B, plus 10 minutes each for whatever other misc. props the characters touch. You have to put all of the props away when your done. Then, no matter how good you are, it will require some cutting plus organization, another 15 to 30 minutes.
 
Well, think about it...

For dialog editing I have to identify each character and give him/her their own audio track. I then have to clean out all of the background noise behind the dialog and strip out all of the sound between lines of dialog.

Then the sound generated by each character has to be recreated (Foley - footsteps, clothing, prop manipulation, etc.). All of the other sounds also have to be recreated - ambience, vehicles, weapons, doors, etc. Big action films can have over 1,000 audio tracks.

Oh my goodness. These are concepts that still exist outside my head :(

so much to learn, so much to learn... or... so many friends to make, so many friends to make :D
 
Back
Top