WRITTEN BY KATE FINAN
CO-OWNER OF BOOM BOX POST
We are excited to continue our Music Mixing Basics for TV & Film series. All month long, experienced Boom Box Post re-recording mixers Jacob Cook, Kate Finan and Jeff Shiffman are tackling a wide variety of approaches to post-production music mixing.
As a mixer of a lot of Disney content, I have years of experience mixing musical numbers (or songs, as we call them). As someone who came into post-production sound mixing from a classical music and music engineering and mixing background, this is one of my favorite parts of the job. It brings me so much joy to have the privilege to integrate amazing Broadway-style musical numbers and carefully crafted pop songs of all genres into the stories that kids consume. A great song can be a wonderful extension of the story–not just adding a little pizzazz, but bringing the entire story to new depths.
For the purpose of this post, I’m defining a song as original music that contains both an instrumental track and vocals. This may be sung on-screen by characters, or it might be used to create a montage. I would use a different approach to needle-drop or licensed music, as that often serves a different purpose in storytelling and you will often not receive the same breadth of split elements to work with.
Transitioning Into a Song
In order to transition into a song, I believe that the volume of the dialogue before the song should seamlessly transition into the vocal volume during the song. This should be done by ear and not with the meters. There might be a small uptick in volume that feels natural between speaking and singing, but the singing should not suddenly strike you as being a higher volume.
After choosing a vocal volume which feels natural, I will then adjust the instrumental tracks to be the same volume as the vocals. This will inherently make the instrumental tracks louder than the surrounding score. That’s exactly what you want. The song should really take over the mix to become a bigger and more emotional moment in the story. But, by keeping the dialogue to vocal level consistent, you are making the transition seamless.
If it feels like the instrumental track is popping on too loudly, I will consider ramping into it. This can be especially nice if there is a long instrumental intro with dialogue over it. In that case, you may want to keep the track slightly lower under the dialogue, and then ramp into the full volume just before the vocals enter.
Panning
I like to pan any on-screen characters’ vocals or lead vocals to the center. Again, you want your dialogue to seamlessly transition into the vocals. You want to avoid feeling like anyone popped suddenly to a different speaker.
Pan any background vocals left-right (rather than center). Sometimes I have instances where there are many characters on screen singing at the same time, and they are not really BG vocals (in the sense that is implied when talking about BG vocals in a song). In this case, I will choose the “lead singer” and pan that character center, and then put everyone else L-R even if they’re on screen. In that case, be sure that your lead is actually singing the lead melody and not a harmony. I like to think of this like separating the lead singer in a stage musical from the chorus. This separation really helps to focus your viewer’s attention and spread the mix to use the full space afforded to you.
I also like to request that all vocal effects (reverb, delay, chorus, etc.) for each singer are bounced onto separate tracks by the music mixer, and then I will pan those L-R even if the vocal itself is panned center. This ensures that no phasing occurs when summing, as it sometimes does with certain effects.
If the song that you are mixing has no on-screen singing, and is instead an overlay for a montage, you can consider keeping all of the vocals panned L-R. This will make it drive the emotional story arc more (much like underscore) rather than seeming tied to anything on-screen. Sometimes this feels better, and sometimes it feels too distant. Play around with it!
You can choose to use an up-mixing plugin or your panner to pan your instrumental music tracks if they were not given to you in 5.1 or Atmos (whichever is your mixing format). I, personally, like to use the panner as it gives me more control and allows me to better match the song’s reference mix in the downmix (see below). If using your panner, pan the instrumental tracks back somewhat so that the song is also coming out of your rear speakers. You want your audience to feel fully encompassed by the music. Play around with how far back you want to go. Keep this panning choice somewhere near what you do for the underscore, but I feel that further spreading the song can be a great way to widen your mix slightly for added emotional impact on the audience.
Volume
This is the point at which I will reference the stereo mix I received (track + vocals mixed together and possibly mastered as a stereo file). Panning choices will greatly affect the perceived volume balance, and I always want to be sure that I’m staying true to the original intention of the music mixer in both my final format (5.1 or Atmos) as well as in the stereo fold-down. Now that I’ve spread the song, it will sound different. But, I want the stereo track, my full 5.1 or Atmos playback, and my stereo fold-down to all be perceived to be roughly the same. At this point, I’ll adjust all of the vocals, or all of the instrumental tracks to make that possible.
Remember, your job is to mix the song into the episode and to upmix it into 5.1 or Atmos. Your job is not to actually mix the song–a music mixer has already done that! So, you should not have to worry about applying EQ, compression, or rebalancing of instruments overall.
Once I’ve found a volume that keeps my mix true to the intention of the music mixer now that things are panned differently, and I’ve possibly ridden into the intro of the song, I always leave the song at a constant volume from start to finish. I do not ride the faders like I would with the score. It’s not necessary since I’m not working around dialogue, and I find that it will make the song have less punch.
But, there are some scenarios when I do make volume changes within a song. I will lower the track (all of the instrumental stems) slightly if I find the lyrics impossible to understand. Remember that you are telling a story with the song, and I find the intelligibility threshold is a bit higher for television and films than it is for a pop music track you listen to on the radio.
I will raise the track slightly if it feels like the energy is too low when moving into the song. But, only if I can do so without compromising the intelligibility of the lyrics.
I will also raise individual singers or instruments if they are featured on-screen and it feels like their volumes aren’t in line with the focus they require from the audience. Again, I only adjust this if it doesn’t feel quite right when I play them at a nominal level. This is all about how it feels! Don’t get too cerebral with this, and really go with your gut. Less is more.
Reverb
I like to add a touch of 5.1 or Atmos reverb to the vocals even if they already have their own vocal effects applied by the music mixer. This helps to spread them into the spatialized mix without doing anything too dramatic with your panning. A light touch is essential here. You want just enough to spread your mix and take advantage of your full speaker array. But, not so much that you change the volume balance or wash out the mix. Again, it should sound like the original mix, just fuller and more spatialized.
Transitioning Out of a Song
When transitioning out of a song, I like to employ the same ideology as I do going in. I aspire to have the vocals to seamlessly transition back into the dialogue. This is all about a natural feeling, and not what your meters say.. Again, the singing may be slightly louder (or it may not), but it should feel like the natural difference that would occur in real life between singing and speaking. Likewise, the score will be lower than the song’s track, but it shouldn’t feel like the bottom dropped out of your mix. You may need to start slightly louder than usual on the score than usual and then ease it back. But, I do find that any large fader moves to transition in or out of songs are likely the result of a song that needs an overall volume adjustment. If you find yourself needing to really ramp into a song, or carefully ease out of it, then it is probably set too loud overall. You should have to do almost nothing to make this work if your song is the right volume.
In conclusion, it’s all about bringing the song into the space and applying choices like panning, volume, and reverb that keep the viewer focused on the story rather than noticing your mix choices. A well-mixed song should feel to the viewer like it hasn’t been mixed at all, it just is.