Loudness Basics
I’m going to go over some details about loudness and loudness specs in my own words, if you’re interested in having a bit deeper of an understanding of the subject. This may be a bit technical at times, but hopefully it’s helpful.
There are three broad loudness specification categories that I’m generally mixing in. Theatrical, television, and the web. How we mix for these formats based on creative intent and technical reasons differ. Let’s go through each one.
Theatrical
First up is theatrical, which is usually feature films. If we’re mixing a film for theatrical, we mix to that experience of watching a film in a movie theater as our hero mix. This will be the best way to watch the film, and we have the most flexibility with creative intent because of the controlled environment. There can be other versions we mix afterward, such as a home theater version, broadcast or cable TV, streaming, or airline, which we’ll go over later. Every theater’s speaker output is calibrated (or is supposed to be) at the same level, which is 85dB. This is setup by playing a particular pink noise file through the speakers and adjusting the volume until it is registering at 85dB in the room. What’s important to note here is that this is pretty loud, and the loudest our speakers are set to for any format we’re mixing. This gives us the most flexibility, so we can make loud moments LOUD if we want to. Everything is mixed to taste though, with the dialogue at a solid comfortable level. When you mix with the speakers loud like that, the overall average volume of the mix is quieter than TV, stuff on the web, or music, but the loudest parts can be just as loud. This, combined with how quiet a movie theater is when nothing is playing, gives us lots of dynamic range to play in, which can be fun and very affecting in a theater environment. There are no specifications when it comes to loudness when mixing for theatrical, you mix in that calibrated environment in whatever way the director and mixers intend.
Television
Then we get to television, which encompasses many different particular loudness specs and targets. I’ll start with the home theater mix for feature films which I’ll consider a television format. This is what goes on Blu-rays, etc. There usually isn’t a specific target to hit for this, but the intention is, and should be, different when making this version from a theatrical. Some films will play fine at home from the theatrical mix, if all the dialogue was mixed consistently and it wasn’t very dynamic with loud music sections or action. But often when taking a theatrical mix and mixing it for the home, dynamics need to be reduced, with the loudest sections brought down and the quietest section brought up with anchored, consistent dialogue. We want people who invest in a good home theater to have a good experience that’s close to theatrical, but we also don’t want average viewers to chase their remotes volume too much, and we don’t want important moments that are quiet to be lost in a noisy home environment with background traffic, dishwasher going in the other room, etc. How different this is really changes from mixer to mixer, director to director, and film to film.
Once you get to broadcast tv, cable and streamers, is when we have to hit specific targets for loudness, which exist so commercials and tv shows are near the same level and there is consistency throughout programing (though many specs are implemented wrong so this isn’t the case, which is a whole other tangent, a great article on loudness and the CALM Act from my colleagues can be read here). A common broadcast TV spec is -24LKFS Full Program Loudness with a true peak of -2. Here, 0dB is the loudest a sound can be. -2 true peak means no audio at all can go above -2, 2dB below zero. This is particularly relevant to very loud sudden sounds such as gunshots. -24LKFS is an average measurement taken of the entire show. Many mixers and filmmakers dislike the Full Program loudness spec, particularly when it’s enforced act by act between commercials, because busy scenes with active music and action can drive the volume level of the dialogue down, so the dialogue is inconsistent. We much prefer the spec that has been introduced by streamers like Netflix (and is supposed to be the law on broadcast), which is Dialogue Gated Loudness. This means that the LKFS average level is only monitoring the dialogue level, so dialogue is consistent in volume from show to show, and allows for a bit more freedom around that. Netflix chose a -27 dialogue anchored spec, others have -24. -27 gives us a bit more head room for loud moments.
What this means in practice, is that we set our calibrated room volume a bit lower from that 85dB level we use for theatrical, so that the dialogue sounds good at the level we want, and we mix around that. For a -27 dialogue spec we usually set the room 6dB lower to 79dB, for -24 a few dB lower than that at 76dB. The lower we get, the less head room we have for loud moments. But this is OK because we’re mixing with home TV systems in mind, which changes our mixing approach. We will mix on smaller speakers too, to make sure the mix works on sound bars and smaller consumer speakers.
Web
Mixing for the web doesn’t exactly have rules but there are guidelines to follow. Once something is on platforms like YouTube, Vimeo, Facebook, or Instagram, the audio level of your video competing with music videos, podcasts, content creators, etc. This is important because people will have their computer, tablet, phone, etc volumes set to listen to those things, so you want your video to be loud enough to be appropriate in that context and not so low that it’s hard to hear even when a persons volume is turned all the way up. YouTube has a target loudness spec of -14LKFS. This is loud, but mostly exists for music on the platform, since pop and rock songs are usually mixed louder than that (YouTube and other platforms will turn down the music to -14 if it’s louder). I usually aim for around -18LKFS, but it depends what it is. When you make the whole mix louder though, there’s less room for the things you want to be loud in context of the mix, to be loud, so how something is mixed or mastered for the web can vary depending on the content.
The question that often comes up though is how to present films on platforms such as Vimeo, whether for the public or for festival submissions. My answer is usually to make a web version that is a bit louder than the original mix so that it will play nice on laptop speakers, but not loud enough that it decreases the quality of the mix by compressing loud moments too much, if I think it’s safe to do so for that particular film. This will change from film to film though.
Hopefully this gives you a bit more insight into this topic without adding too much confusion, I’m always happy to answer any questions.