A Film Editors Guide To Better Sound

There are times in the wild wild world of documentary production where the role of sound editor falls on you - the picture editor.?? Be it budgetary reasons, time-restriction reasons, or maybe even an actual desire to experiment and stretch a little - sonically speaking. Whatever the reason may be, it happens. So this one's for you timeline-titans who want the quick cliff's notes on how to get the best sound you can without being a sound engineer. Here we go.

Where To Start

Dialogue is the fulcrum point of your whole film. The level of your voices inform the level of everything else. Make sure these voices are hitting the nominal target for your distribution channel – which for broadcast is -24lufs average, for netflix is -27lufs average, and web isn't really agreed upon, but I like to approach it a bit louder and let voices drift around -20 /18 lufs average for trailers and ads, but documentaries can still follow broadcast standard for all intent & purpose.

Since dialogue makes up the majority of your film, this should keep the overall average in the right window, while leaving room to get loud with sfx or music ramps in certain parts.

If you do get loud, it's safe to give yourself a range of 6-9db (lufs - not peak) above your voices momentarily. Anything above that might be too loud...maybe artistic?...but probably too loud. If your meter is pumped but the sound is more quiet that you woud like, it is a good sign that you're not monitoring bass correctly, and should EQ some of bass out so you can turn your sound up to sit where you want.

For anything not bass related you should be able to hear what "too loud" sounds like once your voices are set in the correct place...another trick for this coming shortly.

NOTE: that these levels should be considered before and after you do any processing on the voices. You need to hear the voice at the correct volume before knowing how to process it, but then your processing might affect the level reading and it will need to be turned up or down again as a final measure.

Shaping Your Voices

Part 1 - EQ

I made a video on this concept here -> https://vimeo.com/845450147?share=copy

The trick to smooth and natural EQ is to use bands together to create a contour. On film sets the lavalier microphone is really only ever placed around the neck or the chest. Between it's small capsule size and that placement area there is a build-up of "resonant" and "nasally" frequencies between 200hz and 1800hz that can be massaged out naturally by combining wind-band stokes with your EQ. Save acute hi-q cuts when you are taming a room resonance or offensively resonant sounds. In other words - broad-band for tone shaping - narrow band for surgical correction.

Part 2 - Dynamics

Dynamics processors like compressors or limiters can help speed up the process of taming dialogue a lot. There are so many creative ways we use these tools to add punch and stability in a mix - but you don't have to concern yourself with that.

If your voices are clean and dry (not so much room reverberation) it is typically most transparent to use a fast attack and a fast release (maybe 2ms attack and 14ms release). I like to use higher ratios closer to what a limiter would use (7 or 10:1) and adjust the threshold to only clip the loud parts of the dialogue, but not see the meter move in average parts. Limiters have these characters naturally, so you may find a limiter to work just fine on its own. Compressor's are essentially more discrete and flexible limiters, and if you're using a compressor you can follow what I mentioned above. Note that this is just my opinion on what works quick for natural dialogue in the context of film. Get 2 audio heads in a room and they can talk about compressor preferences until the lights burn out.

If you see you gain reduction meter move on every bit of dialogue it's likely pumping the sound too much, and you may not notice this on your monitors. If you just see it moving when the voice gets louder (and the voice still sounds natural when it does) you should now have a more stable range to move the voice volume up and down without sacrificing the nuance and inflictions of your talent's performance.

If the recording is noticeably reverberant you may not be able to use compression at all without pumping either the voice or the room in an unnatural way. If you do try to use compression this is where a slower release time might work better, but might not. You can manage room reverb with iZotope Rx, or some other 3rd party plugins, or leave it– or call a pro! ??

Listen on your laptop or phone... at least once.

Headphones and full range monitors are amazing for nuance, sketching, and doing the major work... but unless you're in a perfectly treated room with trusted gear they can also over emphasize and skew certain parts of the spectrum that will mask big problems in your mix. In all honesty, it is almost impossible to get a good read on level-decisions with headphones. This is where a narrow-band, compressed, listening system can be like taking off rose-colored glasses. The way I phrased that might sound fancy, but simply put - your laptop and phone speakers are narrow band, compressed monitors. Listening back on such a system without the bass or lower mid range filling things out can really show you very quickly if your music (or sfx for that matter) are too loud relative to your voices, or what elements are competing for space in the listener's mind.

There are plugins like "mixchecker pro" by audified that can emulate this on headphones and full range speaker systems, and they're okay. Heck, I use that particular plugin sometimes, but nothing beats the real thing. Changing the localization, medium, and monitors that your film's sound plays from all at once is a huge perspective shift you just can't get any other way.

Using A Reference Mix

Reference tracks are like sonic compasses, bringing you back to true north when you've been looking at your screens for 10 hours and you can't figure out what "good" even-am anymore.

A reference track is essentially an audio print from a piece of content similar to the one you're working on - and ideally that you admire – that you can use to A/B as you get into the weeds of your own project.

The only trick is...drumroll please. ??

you have to make sure you align it's average level to the same average level you're aiming for in YOUR mix. (see part 1 about where to start).

To do this: solo your reference and find the cleanest section of dialogue you can, and adjust the volume fader until your meter is hitting where your target loudness (lkfs) window is.

Our ears have natural features that color sound differently at different volumes, so it's psychoacoustic warfare if you are going to A/B two sources that have different average loudness.

Using Dialogue as your anchor will also give you a great awareness of where the true-peak reading is on your reference, and also what it looks like on the meter when they get "loud" - so that you can know where your boundaries are too.

TLDR: Use a reference. Also make sure it's adjusted to the same relative volume you're aiming for. It's important to check in your editing suite, because online platforms and streaming platforms all have their own algorithms that won't necessarily be playing back the content at the same level it was uploaded at - or at your target for distribution.

Your Monitors Are Probably Lying To You

Likely the hardest and most expensive part of mixing is creating a space that won't lie to you. And that's okay, you can get away without it being perfect. Just don't make the mistake of thinking that you're gonna make a mix that translates on your bose companion II speakers and a pair of close-backed AKG headphones. It might sound great for you there on that system, but not for anyone else where they are.

Two companies that can help with this are Steven Slate Digital and Sonarworks. Steven Slate's VSX headphones & software are getting rave reviews for emulating honest and true mixing environments, and are incredibly affordable. Sonarworks does much of the same, and also has tools for calibrating speaker systems in your environment. Once again, these can help you get closer to true and hear nuance - but when it comes to hearing dynamics of soft ambient Foley in your film, or doing noise suppression (audio Rx), you'll still want a small compressed speaker to help you double check the right differences in levels...maybe those bose companion's could help you there too, hehe. One day I'll write an article on the port-less single driver reference speakers I love to use for this purpose. Connect if you're interested!

That's it For now!

I may continue to add to this article as more tips and tricks transform into words in my head, but that's enough for now. These are some of the biggest cardinal considerations to check off if you're managing your own mix. Of course, we who do this for a living have all sorts of tools and libraries to make the process sound more transparent and reach the right pocket fast - but if you're stuck managing it on your own in Fairlight or something, these are some major things you want to consider.

If you've got any questions or curiosities that weren't addressed here I invite you to connect and reach out! I'm always happy to talk to shop with kindred colleagues. Take care out there. ??

Adam J

BBsoundfactory