SOUND ON SOUND

[roses2021]

Hi, so I'm pretty new to vocal and music production. I've run into an issue where a lot of different parts of a song all differ in volume (sometimes the instrumental in a part of the song is louder somewhere than in all the other places and I can't hear my voice as clearly etc).
And I'm wondering if the only way to help that is to go in and manually change the volume of certain vocal regions or is there a more professional way that would get my vocals to be all on the same level? I wouldn't want it to sound super unnatural though, as in there sometimes are parts in a song where you want your vocals to be softer or quieter and would e.g. a plugin for the constant volume level take that away, so that even the soft parts would be as loud as the rest of the vocals?
I'm using LogicPro btw

Any sort of advice would be much appreciated! :)

[BJG145]

Hi, so I'm pretty new to vocal and music production. I've run into an issue where a lot of different parts of a song all differ in volume (sometimes the instrumental in a part of the song is louder somewhere than in all the other places and I can't hear my voice as clearly etc).
And I'm wondering if the only way to help that is to go in and manually change the volume of certain vocal regions or is there a more professional way that would get my vocals to be all on the same level? I wouldn't want it to sound super unnatural though, as in there sometimes are parts in a song where you want your vocals to be softer or quieter and would e.g. a plugin for the constant volume level take that away, so that even the soft parts would be as loud as the rest of the vocals?

Automating or "riding" the level for vocals is a fairly common production technique. There are plugins to assist with it, eg Waves Vocal Rider, but I expect many of the mix engineers here will be used to adjusting levels manually with a fader or a mouse.

Other parts of the vocal levelling process are a good mic technique, and compression.

PS Welcome to the forum...

[Sam Spoons]

Back in the day the producer/mixer would sit at the desk and manually adjust the levels of various instruments in real time as the multitrack mix was recording to a stereo tape machine. Sometimes it was all hands on deck with various assistants, musicians (and, possibly, the studio cat) rehearsing the fader moves for a complex mix. These days, as BJG145 says, it's most likely done by automation.

[Fishnish]

If you're recording vocals in sections it can be be easy to get a mismatch in level or timbre, especially if the vocalist's position relative to the mic has changed between takes. There's lots of ways to deal with this, but sometimes I find it useful to put each section on a separate track then adjust faders and eq to get a general level and tonal balance between the sections and a rough version of the dynamics -which sections need to be lounder and which ones softer, then bounce it all back together onto one track for more detailed processing, effects and automation.

[The Elf]

There are actually multiple parts to the solution. Let's assume the performance is as you want it...

First there are the loud and quiet parts of the song. These I typically handle by 'multing' the vocal onto several tracks and switching between them for each section.

Then there are the loud/quiet words/phrases that either jump out or dip under the mix. For these I will cut the audio and apply a level adjustment.

Then there are loud/quiet syllables. For these I will apply a bit of a volume curve in the 'Part' (in Cubase-speak - I'm sure Logic has similar) and ramp the changes to hide the effect. For 'P' and 'B' plosives I will cut the audio and apply small ramps to hide them.

I will often use two or more compressors as inserts, one to catch the loudest peaks and another to smooth the overall performance. It's around this stage that I may add a de-esser, and I will place this such that I leave room for a bit of HF lift and also such that any further dynamics processing doesn't over-emphasise the sibilance. Getting the tone of vocal is just as important as the level, so I'm open to static or dynamic EQ and applying quite brutal EQ to parallel compression to get the effect I'm after.

Then there is the overall dynamics of the vocal. I want to the vocal to effortlessly float over the mix like a musical god/goddess having the divine ability to always be heard above the melee. For this I will use some compression (often some parallel compression), then some automation.

Sometimes I will cut all the sibilance from a vocal and give it a separate track, so I can balance it more easily.

I'm not a huge fan of Vocal Rider, though I have used it occasionally. I always automate on a Group Track, so that I can adjust the levels ahead of the automation.

Phew! Hopefully I've covered most of it. Of course, the reality is that it is any and/or all of the above and it may change between song sections.

[desmond]

Welcome to the world of mixing!

It's a strange, fabulous place full of mythology and quirks and tools and accidents and tricks and trade secrets, many of which you'll discover upon your lifelong journey to reach the destination of "You're a Good Mixer Now!"-land...

[Watchmaker]

I want to the vocal to effortlessly float over the mix like a musical god/goddess having the divine ability to always be heard above the melee.

Well said!

Welcome to the forum roses2021. As you can see, getting the vocal right - or any front part - takes a huge bag of tricks and lots of experience. Maybe Artificial Intelligence will remove the need for human intervention at some point, but then why bother learning the physics, engineering, and most importantly, emotional, aspects of music?

A key to getting stellar vocals is the actual performance. It can't be stressed enough that it's best to get it right at the source. For people like me who haven't devoted the time to be a great singer, I accept my singing for what it is and move on. When recording others, I work to get the best take possible in the amount of time available with the goal of minimizing the number of steps necessary to make it pop. Note the word minimize, it will come up a lot in engineering :-)

As my wife keeps telling me, everything worthwhile takes effort. Besides, pulling your hair out trying to get vocals to sit consistently throughout a piece of music is fun if you want it to be :-)

[blinddrew]

For me it depends a bit on the total problem to solve. If I'm only trying to sort out some volume discrepancies then a bit of basic automation in parts (bring the verses up, choruses down etc), followed by a couple of runs of fader automation on specific syllables, and finally a bit of parallel compression to keep the weight on the quieter sections.
If I expect to be doing more creative stuff, e.g. adding different effects to choruses or verses, then I'll mult it out like The Elf says above.

[Exalted Wombat]

Try all the things that have been suggested.

Then start over. Re-record, with all the things that MIGHT go wrong in mind.

[BJG145]

Mixochist

Noun

Someone who uses cannibalised speakers on kick drums, records directly to an Otari, and fuels arguments between band members, to get some grit into it.

[Sam Spoons]

Mixogonist

noun

An engineer who who thinks musicians should not be allowed in the control room

[major_mixing]

Just use compression and Vocal Riding (automation of level) and you will get really contact level for vocals:)