Struggling with vocals in the mix

[Lord Tim]

Just adding to what John is saying, while you might have a fundamental frequency of a voice or an instrument audible, sometimes you'll find that once things are in a mix, certain frequencies - even the fundamentals - start to sound wooly because they're sharing frequencies with other instruments, or even more of the same instrument once they're layered.

For example, I do a lot of metal and typically a rhythm guitar in metal is a pretty chunky sound. Layer 4 rhythm tracks together and you get a huge build up in the bass frequencies that you need to dip so it all sounds balanced again. Then you add bass guitar and kick drums and suddenly it all sounds wooly again. So you roll off the low end to make way for those other instruments. By itself, those guitars sound super thin and weak, but your ears are going to fill in the missing parts in context when it's combined with the other instruments in the mix.

Vocals are the same. You'd be surprised at how much you can roll off of the low end before you start to notice anything going missing once they're in the mix. This doesn't mean that at exactly 300Hz (or whatever) you cut everything off entirely under that, but you can do a gradual roll-off to nothing so you're not putting more stuff into a mix than you'd actually hear in context.

I'm sure I'm not saying anything most people that mix don't know already but when you're first starting out, there's a big disconnect between "this thing sounds good" and "this thing sounds good in context."  It may not be "faithful" to what's down, but I can guarantee it'll sound a lot better and more professional when you find the space for each instrument and get rid of the stuff that's not necessary in a complete mix. Context is the key.

There's some really good advice in this thread that I agree with.

My 2 cents: If you're doing vocal prominent music (pop, etc) then start with the vocal, give yourself a decent amount of headroom for that first, and then bring up all of the other instruments around it. If you're finding it's starting to peak out then drop every track down proportionately. Then start carving things out with EQ, and evening out the peaks either with a compressor or clip gain envelopes. You'll likely find the vocals are very dynamic, so certain words or attacks into words will really crank up the levels, so getting into each clip and just turning down the peaking parts will give you a much more even result and make it a lot easier to slot things together, assuming there's no gear related issues making your life harder than it should be.

[Craig Anderton]

I aim for a consistent vocal level. I'm not a big fan of using compression for this application, I prefer to use compression for "character." So I normalize individual phrases to obtain a high average level in a vocal, without using any dynamics processing per se. Go to Figure 2 in this inSync article and you'll see what I mean.

After getting a consistent level, then I typically apply some limiting or compression, but you don't need as much because the normalization already did a lot of the "heavy lifting." This means you don't hear the dynamics processors working, but you still have a big, present sound.

Also, EQ in the 3-4 kHz range can bring out the articulation in voices, because the ear is most sensitive at those frequencies. However, a little goes a long way - you don't want the vocals to sound screechy. 

There can be too much low end on vocals, which makes the vocals less distinct, and takes up bandwidth so you can't bring up the vocal level as high as you want. I usually add a shelf or highpass filter around 100-200 Hz so that the overall vocal can come up a bit, but whether to do that or not depends totally on the mic and the vocalist. EQ doesn't really lend itself to "universal" tips.

Since you probably will want to use compression at some point on vocals, check out the article Five Compression Tips for Vocals. You'll find more vocal tips in this article in the Full Compass blog. It's oriented more toward narration, but the same techniques work with anything involving voice.

The vocal is the most important part of any song. Unfortunately I don't have any Cakewalk-specific book on vocals, but I wrote one called How to Record and Mix Great Vocals in Studio One. If you look over the table of contents, you'll see that most of the tips translate easily to other DAWs. In fact many of these tips were first developed on Cakewalk projects, and translated over to Studio One for the book.

Hope this helps!!

[RICHARD HUTCHINS]

It does, but its glaringly obvious there's a skills gap at my end! I'll work through it all bit by bit, but it may take some time. I may call my album "voyage of the goldfish" as by the time I learn one thing I've forgotten the other...

[twelvetone]

also, be aware that a signal with high harmonic content sounds much louder than a signal with low harmonic content.

Try this when you have time:
Record a sine wave around 500-1kHz, and then some white noise at the same level.
Notice how the sine wave seems much softer?
Voice has low harmonic content compared to distorted guitar and fat synth pads.

Add the fact that distortion is in a sense a hard limiter and voice is dynamic, you will have the issues bitflipper was talking about.

Edited December 7, 2020 by twelvetone

[Starship Krupa]

Go to YouTube and search for "how to make a vocal pop in a mix" or "how to get my vocal to sit in the mix" or similar and you will find much good advice.

Adjusting relative volume (pulling down the other faders in your case) is the first step. Using processing is the next. By processing I mean EQ, compression (and/or clip gain adjustment as Craig does), reverb, chorus, whatever.

When new to this, as you are, iZotope Neutron Elements can really help get you started. Some see it as a crutch, I would characterize it more as a shortcut. You can use the presets or the wizard at first, then later, when you understand things better, the individual processors in Neutron are top-notch. The Elements series regularly go for deep discount or even as a freebie or part of a bundle, so watch the Deals subforum on here or keep an eye on Plugin Boutique.

Just now they're running a deal where if you purchase anything, you can add MTurboCompLE to your cart and get it for free. PB has many things for sale for under $10, so it's a great way to get a high end compressor for peanuts. Each of MTurboCompLE's modeled compressors has a "Compression" knob that adjusts all the other parameters for you, so it's good for newbies.

An epiphany for me when I was first learning this stuff is that sometimes mixing can be counterintuitive. Cutting certain frequencies, compressing dynamic range can actually make an element in a mix stand out better.

[John Vere]

2 hours ago, Starship Krupa said:

When new to this, as you are, iZotope Neutron Elements can really help get you started. Some see it as a crutch, I would characterize it more as a shortcut. You can use the presets or the wizard at first, then later, when you understand things better, the individual processors in Neutron are top-notch. The Elements series regularly go for deep discount or even as a freebie or part of a bundle, so watch the Deals subforum on here or keep an eye on Plugin Boutique.

Just now they're running a deal where if you purchase anything, you can add MTurboCompLE to your cart and get it for free. PB has many things for sale for under $10, so it's a great way to get a high end compressor for peanuts. Each of MTurboCompLE's modeled compressors has a "Compression" knob that adjusts all the other parameters for you, so it's good for newbies.

Absolutely. I totally score free stuff there all the time. Last 2 freebies were izotope Nectar and Ozone 9 elements for free. Both because I bought a $ 8 VST.  

I have tried them both but found the Nectar when you use the automatic gizmo made my vocals sound worse. And actually quieter. So I guess that is hit or miss. But it is straight forward to dial in your own settings. It's very subtle which is a good thing. Don't go looking for  "wow that sounds 100% better" it's very small difference. 

Same with Ozone. I slapped it on a few songs in the master buss and the automatic gizmo didn't change a dang thing that I could hear. I think my mix was so perfect it baffled the software  

[marcL]

17 hours ago, Kurre said:

I don't like this deleting of subtones and overtones. The end result is then a false representation. (Even if you can't hear it.)

And to delete anything lower than 150 or 250 hz?

"From wikipedia:  The voiced speech of a typical adult male will have a fundamental frequency from 85 to 180 Hz, and that of a typical adult female from 165 to 255 Hz. Thus, the fundamental frequency of most speech falls below the bottom of the voice frequency band as defined above."

Female Singer Type: SOPRANO/C4 - C6 (261-1047 Hz)   MEZZO-SOPRANO/A3 - A5 (220 - 880 Hz)   ALTO/G3 - G5 (196 - 784 Hz)

CONTRALTO/F3 - F5 (174 - 698 Hz)

Male Singer Type: COUNTERTENOR/E3 - E5 (164 - 659 Hz)   TENOR/C3 - C5 (130 - 523 Hz)   BARITONE/G2 - G4 (98 - 392 Hz)

BASS/E2 - E4 (82 - 329 Hz)

Explain yourself, please.

That's exactly how I thought in the beginning of my mixing trip! But in the mean time I learned a lot from professionals and the internet.

There are 3 things I want to add to the roll-off subject:

1. Usually you do not reduce a lot of frequencies during the recording (only in the mix)! Maybe only the low end a little bit to avoid rumble (that's the reason why a lot of professional microphones have a low end reduction knob).
2. If you have a song part with pure vocals (e.g. opera singer) or only with skinny instrumentation (e.g. only piano), then you may leave more of the low frequencies, because it is not overcrowded in the low end. But even in this case some minor roll-off of low frequencies may improve the vocal sound, try it!
3. If there are a lot of instruments, then it is  definitely not possible to make all the overlapping frequencies of vocals and instruments hearable in a recording. That's a fact. (it's alike being in a room with many people that are all talking, you are not able to hear everything.) So it is you, the mixing engineer who decides which frequencies of each instrument/vocal has to come through. The overlapping frequencies are unhearable, so taking away the same frequencies of other instruments makes them raising on the targeted instrument (e.g. vocals). Test it in a mix, taking away frequencies of instruments will make your vocals louder without raising the volume slider! Surely, you should not overdo the roll-off, otherwise it may sound too thinly. And as others said, the amount and frequency you roll off depends also on your taste and on the instrument/vocalist's frequency spectrum.

Somehow the mixing engineer does something similar to a painter that has to decide what is visible in the foreground. He cannot paint multiple persons/things at the same position (analog to frequency). It is not visible what is behind the object.

[Kalle Rantaaho]

22 hours ago, Kurre said:

I don't like this deleting of subtones and overtones. The end result is then a false representation. (Even if you can't hear it.)

And to delete anything lower than 150 or 250 hz?

"From wikipedia:  The voiced speech of a typical adult male will have a fundamental frequency from 85 to 180 Hz, and that of a typical adult female from 165 to 255 Hz. Thus, the fundamental frequency of most speech falls below the bottom of the voice frequency band as defined above."

Female Singer Type: SOPRANO/C4 - C6 (261-1047 Hz)   MEZZO-SOPRANO/A3 - A5 (220 - 880 Hz)   ALTO/G3 - G5 (196 - 784 Hz)

CONTRALTO/F3 - F5 (174 - 698 Hz)

Male Singer Type: COUNTERTENOR/E3 - E5 (164 - 659 Hz)   TENOR/C3 - C5 (130 - 523 Hz)   BARITONE/G2 - G4 (98 - 392 Hz)

BASS/E2 - E4 (82 - 329 Hz)

Explain yourself, please.

I've never recorded a female voice, but I assume I wouldn't cut it any harder than a male voice (?). Cutting  at around 50 Hz cleans the accidental proximity booms of my baritone voice, I don't want to cut anything more than just problem frequencies which add to volume but can't be heard. But as mentioned, I've never faced those of a female voice. I'm an amateur, not so high goals. Such tricks often  help you out if you have lacking monitoring equipment or environment.

Trying to fit together something like piano, bass guitar or contra bass,  and a male vocal  (maybe a cello,too) makes one really sweat with what to roll off and what to emphasize. All the most loudness creating low frequencies more or less overlap. I remember how I struggled with my first rock song which included a piano. I ended up wiping out practically all left hand piano parts, cause I couldn't make it work with the bass guitar. 

[Lord Tim]

Both @Craig Anderton and @marled are on the money - how much you cut is really dependent on the material around it. If it's a very sparse arrangement and you're going for a natural sound, you might leave more of the low end in there (getting rid of any problem frequencies like rumbles or proximity effects with certain mics, etc.) but in a dense mix, you can certainly stand to cut both male and female vocals a lot higher and more aggressively because you won't really hear that cut in context, and it'll clean things up considerably in the mix.

[Starship Krupa]

11 hours ago, John Vere said:

I have tried them both but found the Nectar when you use the automatic gizmo made my vocals sound worse.

I'm not a fan of Nectar Elements either, which is why I mentioned Neutron. Neutron has presets for vocals that sound pretty good.

And heck yeah to Ozone's wizard not doing anything for you, that was when I realized I was getting good at mastering, when I could outgun Ozone Elements. ?

[Glenn Stanton]

1 hour ago, Kalle Rantaaho said:

Trying to fit together something like piano, bass guitar or contra bass,  and a male vocal  (maybe a cello,too) makes one really sweat with what to roll off and what to emphasize. All the most loudness creating low frequencies more or less overlap. I remember how I struggled with my first rock song which included a piano. I ended up wiping out practically all left hand piano parts, cause I couldn't make it work with the bass guitar. 

this is where a side=chained compressor on the other instruments could help by "ducking" some of the parts while the vocal is happening. might only need a couple of db to make it cut, sometimes using things like RBass/MaxxBass (or other LF harmonic generators) can help to cut low end on instruments by creating some additional higher frequency parts of the signal so you can reduce the instrument level slightly... which may/or not/ interfere with the vocal but it's something to try...

[msmcleod]

If you're still really struggling getting the vocals to sit properly, you could try using Trackspacer by WavesFactory:

https://www.wavesfactory.com/audio-plugins/trackspacer/

Basically, it carves out the EQ in other tracks to make space for your vocals... it's much better to learn how to do it yourself though.

[John Vere]

I ended up wiping out practically all left hand piano parts, cause I couldn't make it work with the bass guitar. 

That's why a good rock and roll piano player keeps their left hand on their beer mug! The exception is Boggie Woogie. Just listen to Asleep at the wheel Route 66 and how the piano bass and Upright bass track perfectly together.