The "In-Depth Vocal Mixing" Thread

[lolzgreg]

I think mixing vocals is by far one of the more difficult things to do. I know people have brought this up in other threads, but I think it would be great if we could all collaborate to get a really awesome tutorial going on here and get it added to ahjteam's tutorial section. I know, ahjteam, you commented about this lightly in that particular thread, that there is no "all around answer" for this issue. There has to be some technique to get the instruments of a mix up to snuff and have the vocals still be audible but not overbearing.

I find myself using the following chain:
a. De-Esser
b. EQ
c. Compression
d. Brickwall limiting (because the vocals really need to be boosted in volume)
e. Reverb as a send (usually plate with a nice pre-delay).

I'm assuming the biggest issue is making space in the guitar mix particularly for the vocal track.

Does everyone here buss all the instuments at the end to one track, do "surgical" cutting and then have the vocals as a seperate buss?

Anyone, chime in (especially those who have conquered this problem).

 

[ahjteam]

Well, my processing depends entirely what I have in the track and it doesn't nessecarily need any processing at all to fit the track. I don't process just for the sake of processing. The thing is that there are three main categories of vocals and they all get different kind processing:

- male vocals
- female vocals
- background/gang/choir vocals

Then there are style specific vocals that get different processing too:

- clean, melodic singing (the "normal" singing, 90% of the singers)
- high pitched singing (opera)
- grunt, growl (cookie monster)
- shouting, screaming (screamo)
- whispering (emotional stuff)

I usually try to have all the different type of elements in different tracks even tho if there is only one lead vocaltrack. Just split it up.

 

[greyskull]

1176 on the way in most of the time.

Then eq with uad 1073 and maybe a touch of plugin 1176.

Dont really do a whole lot of eqing, match the mic to the singer.
occasionally deess if necessary.

The more agressive the vocal is the more i spank it with 1176, may use all buttons, and may add a tapehead after.
Use medium plate on Tlspace and massey td5 for delay.
For BVs i normally add a subtle telephone effect, i.e filter the top and bottom off but not massive ammounts, helps them sit behind things. that and maybe some dimension D chorus.

 

[gabriel g.]

For all vocals (no matter what style)
sometimes double tracked

1. R-DEesser
2. R-Vox
3. colourful EQ
4. send to AUX bus with reverb and or delay and sometimes a comp.

but I have a lot of new stuff so I´m looking forward to it.

Maybe my new setup will look like this:

1.R-DEesser
2.Waves MaxVolume
3.V-Comp
4.V-EQ

I heard alot of great things about the V-series for vocals and bass.
The EQ totaly rocks on rythm guitars!!!

 

[lolzgreg]

Thanks for a good start, lets keep this going. I'm on vacation 'till Wednesday night so I'll see you fellows then!

 

[Suicide_As_Alibi]

Then there are style specific vocals that get different processing too:

- clean, melodic singing (the "normal" singing, 90% of the singers)
- high pitched singing (opera)

And cross your fingers you don't work with a mezzo-soprano, as they sit somewhere between ^those two^ and all the important bits of their voice tend to clash with the guitars. I hate my singer sometimes...

I think EQ it the most important thing with vocals (especially clean, sung vocals) - tiny changes can completely change the feel of the track. If I can help it I don't EQ them at all unless there are specific problems like:

- Plosives ('P's, 'B's, 'T's and 'D's) - tend to be around 100Hz
- Nasal-ness - somewhere around 1.5-3kHz
- Presence - 4-5kHz
- Sibilance ('S's, 'Z's and 'F's) - 8-12kHz (though breath noise is also in there, so you have to be careful with soft/sexy lady vocals)
- Air (I've also seen it called 'crispness') - 17kHz (but it also adds noise like a bitch if there was no vocal booth)

Compression entirely depends on the singer. If you get a good singer that pulls away from the mic when they get louder, I find too much compression can make the change in distance really obvious.

I don't generally limit sung vocals - I record them as hot as I can, which seems to be enough with compression + a few dBs of make-up for most of my recordings.

Harsh vocals are a whole other kettle of fish though. It's all about making them sound as aggressive as possible, so I'm a lot more... enthusiastic about my editing Heavy compression, lots of EQ'ing, adding distortion... whatever works for that particular voice.

The one thing I always find the local studios here do (who don't have much experience with angry vocals etc.) is that they boost the lows on low grunts and boost the highs on high screams - when I normally find myself doing the opposite. High screams generally needs more chutzpah to make them sound powerful, and lower stuff often needs some high-end to make them a bit grittier. Voices should cover the whole spectrum, but I find vocal fry/screaming tends to concentrate all the power in a narrower area - so they need some balancing.

Steve

 

[Javi Perera]

Hello everybody!


Well, the DAW that I use is the SAMPLITUDE 10, and my chain of effects for vocals is the following, from the beginning to the end:

1.- sometimes AUTOTUNE ( though I prefer MELODYNE )
2.- EQ
3.- Compressor / Limiter / Gate ( I normally apply very high levels of compression, the limiter in such a way that all the voices sound with the same intensity, and the gate in order that it leaves the vocals more "dry" ).
4.- DeEsser
5.- AUX to use delays, reverberations ...

For me, the main problem that I have at the moment of recording vocals is that, when I listen to my recorded voice, it doesn't sound as I hear it inside my head. Is there any way of achieving that the voice in the recording sound like the intracranial voice? Can it also depend of the kit that it is used in the recording and the chain of effects?

 

[darthjujuu]

1176 on the way in most of the time.

Then eq with uad 1073 and maybe a touch of plugin 1176.

Dont really do a whole lot of eqing, match the mic to the singer.
occasionally deess if necessary.

The more agressive the vocal is the more i spank it with 1176, may use all buttons, and may add a tapehead after.
Use medium plate on Tlspace and massey td5 for delay.
For BVs i normally add a subtle telephone effect, i.e filter the top and bottom off but not massive ammounts, helps them sit behind things. that and maybe some dimension D chorus.

*scribbles down shopping list and buys $20 scratch ticket*

 

[C_F_H_13]

Here's just an example of what I did this weekend.

-Female lead vocal ala Feist. Recorded with a u87, neve 1066, distressor, and 1176.

PT processing:

1.PT EQ3-doing some minor corrective eq...pulling out a few wierd low mid honks and a few odd whistles in the top end.

2.Waves Desser- She needed it

3.Waves C4- I have a patch I created for vocals that's similar to the pop vocal preset, but a little more suited for the plugins that follow in my chain.

4.Waves Rvox- Just a little bit of compression here...maybe 3-4 on the loudest parts.

5.Waves L1- CLA posted a patch he uses in an article of SOS some time last year and I tried it on vox and actually really liked it. So usually all I do is drastically bring down the amount of limiting as compared to the preset.

All further processing was done OTB, and basically included some minor eq, a plate verb, couple of delays and some mild pitch modulation.

 

[if6was9]

Hello everybody!


Well, the DAW that I use is the SAMPLITUDE 10, and my chain of effects for vocals is the following, from the beginning to the end:

1.- sometimes AUTOTUNE ( though I prefer MELODYNE )
2.- EQ
3.- Compressor / Limiter / Gate ( I normally apply very high levels of compression, the limiter in such a way that all the voices sound with the same intensity, and the gate in order that it leaves the vocals more "dry" ).
4.- DeEsser
5.- AUX to use delays, reverberations ...

For me, the main problem that I have at the moment of recording vocals is that, when I listen to my recorded voice, it doesn't sound as I hear it inside my head. Is there any way of achieving that the voice in the recording sound like the intracranial voice? Can it also depend of the kit that it is used in the recording and the chain of effects?

No you just have to get used to how you're voice actually sounds- gear can't take you there!

 

[dcb]

i think from the first take on you now whether a vocalist has a unique voice,
that he knows how to work with. i think you can only put emphasize on great aspects of someones voice. if there is nothing great about the voice, no neumann in this world will help you ;-)

anyways, here is my setup that i use:

k&m popkiller
at 4040
mindprint envoice mk II (limiter!, 10 K bppost 3 db)

(my) vocals are limited before d/a ,

in the mixdown i add 3 types of compression, first is basically
limiting the shit out of it (attack 1.5ms, release 10ms),
then add slight compression to even out the dynamics a little more,
then add something like the rvox or so to crush them even more

i also tend to eq vocals this way : take away some 300 db Q 1,0 -3 to - 5 db, this will help to sit much better
in the mix. then i add some 10 k, after dessing (which i most of the times dont need to do)

never forget to "ride" your vocals in the mix, this is probably
the most important thing about mixing vocals !
automate the vox so they sit nice inside of the mix all the time.
presonus faderport is a cool tool to do that btw.

if your interested, listen to the exterminating angel on my soundclick page,
its basically only the described recording / mixing chain.

http://www.soundclick.com/bands/page_songInfo.cfm?bandID=885206&songID=7521985

 

[Plankis]

No one's using saturationplugs? I read this thread: http://www.ultimatemetal.com/forum/production-tips/465371-vocal-distortion-plug-ins.html a while ago, I think that TL Saturated driver works good to tame some dynamics. I try to have it first in the row so it simulates a preamp. Am I completely wrong doing so?

Reverb is also something I struggle with. Really hard getting a good mix of it, It's always too much or too little... I can't understand how pros get so dry but still so "airy/smooth/random words for evenness and fits perfect in the mix". A really good mic seems to be the only solution in my none existing experience.

 

[dcb]

No one's using saturationplugs? I read this thread: http://www.ultimatemetal.com/forum/production-tips/465371-vocal-distortion-plug-ins.html a while ago, I think that TL Saturated driver works good to tame some dynamics. I try to have it first in the row so it simulates a preamp. Am I completely wrong doing so?

Reverb is also something I struggle with. Really hard getting a good mix of it, It's always too much or too little... I can't understand how pros get so dry but still so "airy/smooth/random words for evenness and fits perfect in the mix". A really good mic seems to be the only solution in my none existing experience.

the trick is to use only as much reverb as needed :

turn down your reverb until you do not hear it (-30 db)
then turn it 1 db up. mute and unmute it.

does the "airyness" of the overall sound collapse,
when you turn the reverb off ? then you have just the right
amount! if there is no difference - turn it up 1 db and check again,
whether your vox sound more 3d.

unless you want to use reverb as an effect, always use as little as possibble. a great reverb will work when its used very subtle, while a cheap one will consume much more space in your mix.

also try stereo delays, they make your vocals very very large.

 

[Fox Mulder]

1.- sometimes AUTOTUNE ( though I prefer MELODYNE )
2.- EQ
3.- Compressor / Limiter / Gate ( I normally apply very high levels of compression, the limiter in such a way that all the voices sound with the same intensity, and the gate in order that it leaves the vocals more "dry" ).
4.- DeEsser
5.- AUX to use delays, reverberations ...

Why'd you wanna apply a de-esser and gate there, man? It's a good idea to apply the gate first in the chain followed by a de-esser, and then the auto-tuner.

 

[Fragle]

the trick is to use only as much reverb as needed :

turn down your reverb until you do not hear it (-30 db)
then turn it 1 db up. mute and unmute it.

does the "airyness" of the overall sound collapse,
when you turn the reverb off ? then you have just the right
amount! if there is no difference - turn it up 1 db and check again,
whether your vox sound more 3d.

unless you want to use reverb as an effect, always use as little as possibble. a great reverb will work when its used very subtle, while a cheap one will consume much more space in your mix.

also try stereo delays, they make your vocals very very large.

+1 to all that.

i really like the sound of a stereo delay with just a hint of reverb vs. mostly verb all the time.

other than that, i'm mainly dealing with growls/screams, and for that i tend to compress the shit out of it - and then some more. after all you're not dealing with a nice female opera voice, this is pure aggression!

moreover, using quite a severe highpass will kill some low noise from the mic stand etc, and also will get rid of all the mud you (i, that is) get from recording a sm58 handheld.
otherwise i tend to scoop some low-mid mud around 3-500hz, some gentle boosts around 1,5-3k for articulation depending on where it sits, and a highshelf to add some sparkle.
i guess this largely depends on the mic (and the singer) you use though.


and while automating fx-send levels is *the key* imho, i sometimes wonder how the big names just *know* where to put these effects in an enhancing way, not just for the sake of it.
getting the right spots sometimes is a difficult task imho...some sections just scream "epic reverb here", but pro recordings are full of this stuff, which in my book is one of the major aspects of the "pro sound" vs amateur style.


and btw, i pretty much never use a gate on my vocal tracks...i always go in there and edit out the unwanted stuff.
actually, i remember taking some time and cutting vocals to the grid, especially on really aggressive stakkato-like parts.....i really hate the sound of all over the place (rhythmically (sp?) speaking) vocals, like on kataklysm's in the arms of devastation

 

[TheDarkProject]

For me, the main problem that I have at the moment of recording vocals is that, when I listen to my recorded voice, it doesn't sound as I hear it inside my head. Is there any way of achieving that the voice in the recording sound like the intracranial voice?

Refurbish a pair of human skulls into studio monitors?

 

[Javi Perera]

Refurbish a pair of human skulls into studio monitors?

Or making an IR of my head with Voxengo Deconvolver?

 

[Javi Perera]

Why'd you wanna apply a de-esser and gate there, man? It's a good idea to apply the gate first in the chain followed by a de-esser, and then the auto-tuner.

I apply them cuz I use a Soundblaster Audigy 2 NX soundcard, and it sounds like shit, causing lots of noise. Also, I record my vocals with a SHURE SM58 and, I don't know why, it makes a lot of strident sybillances when I pronounce "S". I'm thinking about buying an EDIROL UA25 EX cuz I can't stand this anymore...

 

[darthjujuu]

this thread is gold. *bookmarks*

 

[Plankis]

Thanks for all the good advices! I'll try em all when i get home again.
Also Stereo Delay: Do you add it after/before/parallell with the reverb? I find having it before gets a more natural sound to it.

To continue making this thread epic, what are popular mics? I got a sm58, haven't used it much at all. But when I've used it I've always felt that it doesn't blend with the mix easily, it's either too loud or just inaudible.