Some vocal mixing technique I thought you guys might be interested in

[bgrotto]

Since a lotta GSers frequently ask about getting their vocals to sound clear and present, I thought I'd take the opportunity to share a couple screenshots that might offer some clues as to how we do it.

I've attached a couple screenshots of a recent mix I did; these are the two emcees' lead vocal parts. I'm hoping to give you guys an idea of how much of the sound is VOLUME AUTOMATION, not just plugins and hardware.

The plugs that are greyed out were actually in use; I opened the PT session on my laptop, but my iLok lives at the studio where I do my mixing, so I don't have a means of accessing them. Anyway, the basic rundown on "George Lead" vox is:



Rvox - this was doing about 3-5dB of compression. It actually beat out my hardware comps; I tried a Distressor, an 1176, and a Tube Tech LCA-2B. All of the hardware compressors - when combined with the various hardware EQs I tried - added a bit too much grit, which compounded a fuzzy midrange problem in the original track. The Rvox provided some fairly transparent compression without too much fussing about.

H/W Insert A1 was a Neve 1084. Its settings looked like this: GAIN 0, HP70, LP18k, +3@110, -1.8@350hiQ, +1@15k. The +/- values for each frequency ARE NOT IN DBs! It's a shorthand I use for recall notes; basically each dot on the Neve's faceplate is equal to a value of one. So, the boost at 110Hz is +3 of those dots, not +3dB. This is probably a good 6 or maybe even more dB; I'm not sure.

The Neve was chosen because of it's pleasant top end boost and smooth, thick low mids. As you might be able to tell from the low shelf boost, the vocals needed some help on the bottom end, and similar boosts on the other EQs I tried (which included an API 550b, Wunder PEQ1, Chandler TG channel, and Auditronics PEQ82) didn't offer enough coloration, which required still larger boosts, which created noise floor problems and some odd resonances in the low mids.

After that was a DeEsser, taking off a few dB here and there as needed. Note that the DeEsser is PRE-VOLUME RIDE! This is actually not ideal, generally speaking. Had the vocal had more severe sibilance problems, I would prefer to automate as much sibilance out as possible BEFORE hitting the deesser. That would require a more complex routing arrangement, though, which I wasn't keen on. Plus the vocal sound didn't necessitate it.

After all that came the volume rides. I was able to compress less with the Rvox, knowing I'd be doing a fair bit of riding. NOTE THAT THESE VOCALS WERE RECORDED WITH COMPRESSION TO TAPE! I don't know how much, as I didn't cut the tracks, but it's definitely there. In some cases, the tracking compressor kicked in too hard or with too long a release, which required volume rides to bring up some of the tails of words and phrases. In other instances, the emcees' mic technique was a bit dodgy, so sudden large drops in level required boosts of entire words or phrases.

I often ride prefer to ride vocals INTO a compressor, using the desk's inserts instead of PT H/W inserts, but for some reason, that approach didn't work on this mix. Always be ready to change your approach to suit the song's needs!

"Sean Lead' Vox was processed via:



Rvox again! Same deal; just smoothing it out a hair in anticipation of the volume rides. Once again, pre-compression rides weren't workin'...the emcees' voices are EXTREMELY similar (turns out they're actually twins!), so maybe that'd account for the similar signal paths.

Next up was an API 550b. Settings were similar to George's Neve: +4@200shelf, +2@700, -4@1.5k, +6@5kshelf. Like I said, their voices are similar! The API offered a bit more aggressive mids, which helped a lot on Sean's smoother voice.

The Digi III EQ was a cut to the low mids...maybe even as high as 500 cycles, to remove some boxiness that built up when the doubles were added in.

Next up was the RDeEsser; same deal as before.



Both these vocals were actually very well-recorded and well-performed; the amount of volume automation is TYPICAL of such a session. It's rare for contemporary pop styles to require any less, more often they require more!

So next time you're asking yourself why you can't get the vocals to pop or remain present, turn down (or turn off!) the compressor, back off the EQ, and try some volume rides.

EDIT: I should mention that the vocals were really not happening before I got into the automation. The plugs and hardware were "clean up", the automation was the polish. Now they're crystal-clear and sit in the mix rather well.

[Nahuel]

I'm always reading your posts about automation but everytime I tried to use some I ended deleting my envelope cause I was not getting the desired result and had to make a beat right after it to get rid of the headache.

I have to say I dont have a good fader controler. These pics are a very good starting point for me and gives me a more defined idea of what needs to be done.

Thank you very much for that example.

[dhiltonlittle]

Originally Posted by lpuma

I'm always reading your posts about automation but everytime I tried to use some I ended deleting my envelope cause I was not getting the desired result and had to make a beat right after it to get rid of the headache.

I have to say I dont have a good fader controler. These pics are a very good starting point for me and gives me a more defined idea of what needs to be done.

Thank you very much for that example.

you don't need a fader to automate....these automation moves are "moused" in are they not? if not, you've got some smooth hands bgrotto!

[Nahuel]

Originally Posted by dhiltonlittle

you don't need a fader to automate....these automation moves are "moused" in are they not? if not, you've got some smooth hands bgrotto!

ok, thanks for the info, that's how I tried to do it. painfull job, especially when your not very patient, like me... I have to stay focused and stop being lazy, lol.

[Karloff70]

Originally Posted by dhiltonlittle

you don't need a fader to automate....these automation moves are "moused" in are they not? if not, you've got some smooth hands bgrotto!

lol....what I thought.....

[ryst]

You got that right, Benny. That's how we do it!

[jikky]

Originally Posted by dhiltonlittle

you don't need a fader to automate....these automation moves are "moused" in are they not? if not, you've got some smooth hands bgrotto!

Yea I would say moused in definately. At any rate, great explanation of your process. Mine is a lot similar. I record, do volume automation, drop on the plugs needed for individual tracks, drop on the plugs for groups, send if needed, master processing, and export.

How long did it take you on the vox mix?

Thanx for posting as well bgrotto!

[phillysoulman]

Originally Posted by dhiltonlittle

you don't need a fader to automate....these automation moves are "moused" in are they not? if not, you've got some smooth hands bgrotto!

Come on fellas....Its a skill that all of us "old school dudes" are proficient at.
We never had mice back in the day...except maybe in the echo chamber

[Karloff70]

Originally Posted by phillysoulman

Come on fellas....Its a skill that all of us "old school dudes" are proficient at.
We never had mice back in the day...except maybe in the echo chamber

Point well made and understood as ever, but I'd still like to see you hand fade the auto in bgrotto's pics

[elmolemon]

thx for sharing! thumbsup

[MicB]

nice post

[MicB]

Is it me or doesnt the levels seem low? Was there no gain used while recording the vocals?

[dhiltonlittle]

Originally Posted by phillysoulman

Come on fellas....Its a skill that all of us "old school dudes" are proficient at.
We never had mice back in the day...except maybe in the echo chamber

ha! yeah man, i'm only 30 so i guess i can't consider myself old school (or maybe i can) but i automate everything using faders. i prefer to use a fader over the mouse unless it is something i really need to get into. i've found that even though the mouse is convenient, i get much more "musical" results riding an actual fader.

anyway, good post bgrotto. u da man

[elmolemon]

Originally Posted by MicB

Is it me or doesnt the levels seem low? Was there no gain used while recording the vocals?

Originally Posted by bgrotto

NOTE THAT THESE VOCALS WERE RECORDED WITH COMPRESSION TO TAPE!

[MicB]

Originally Posted by elmolemon

Ya I guess you didn't catch my edit, I still wanna know if you guys would consider those levels low... maybe I record too hot?

[phillysoulman]

Originally Posted by Karloff70

Point well made and understood as ever, but I'd still like to see you hand fade the auto in bgrotto's pics

It s f***ing easy,man!!!!

[phillysoulman]

Originally Posted by dhiltonlittle

ha! yeah man, i'm only 30 so i guess i can't consider myself old school (or maybe i can) but i automate everything using faders. i prefer to use a fader over the mouse unless it is something i really need to get into. i've found that even though the mouse is convenient, i get much more "musical" results riding an actual fader.

anyway, good post bgrotto. u da man

How about that??

[halcyo]

I've never really messed with this sort of automation much, even though I know in my mind that it would be a great way to even out a vocal and get it really nice and up-front. I guess I'm lazy (really I'm just more of an artist/musician at heart than I am a mix engineer I guess)....

Do you find that you are able to make those automation edits by just looking at the waveform, or do you have to listen and work it phrase by phrase on the vocal?

Some before and after clips would be awesome, but I suppose your clients probably wouldn't want us to hear it un-mixed.


I've always wondered why there isn't a plug-in that can analyze an entire waveform region, and automatically make the inverse volume curve on a 'sample accurate', musical level. You could even have simple controls to decide 'how much' compensation you want - sort of like a compressor. I know some people would just rather use a compressor, but as you've described in your original post, they are very different processes with very different results!

[phillysoulman]

Originally Posted by halcyo

I've never really messed with this sort of automation much, even though I know in my mind that it would be a great way to even out a vocal and get it really nice and up-front. I guess I'm lazy (really I'm just more of an artist/musician at heart than I am a mix engineer I guess)....

Do you find that you are able to make those automation edits by just looking at the waveform, or do you have to listen and work it phrase by phrase on the vocal?

Some before and after clips would be awesome, but I suppose your clients probably wouldn't want us to hear it un-mixed.


I've always wondered why there isn't a plug-in that can analyze an entire waveform region, and automatically make the inverse volume curve on a 'sample accurate', musical level. You could even have simple controls to decide 'how much' compensation you want - sort of like a compressor. I know some people would just rather use a compressor, but as you've described in your original post, they are very different processes with very different results!

Use your ears and monitor at a low level.

[dhiltonlittle]

Originally Posted by phillysoulman

Use your ears and monitor at a low level.

very true, i've also found that closing my eyes while automating really puts me in focus and helps get things in a good place.

[halcyo]

Oh I agree. You pretty much HAVE to use your ears to be sure that the beginning and ends of phrases work. However, wouldn't it be cool if your DAW could do an 'automatic pass' of volume smoothing automation for you to then tweak by ear? It would save ALOT of time!

[piotr]

Hi bgrotto,

I like your website!

;-)

p.

+

PS

Nice post too!

[matt2525]

Originally Posted by halcyo

I've always wondered why there isn't a plug-in that can analyze an entire waveform region, and automatically make the inverse volume curve on a 'sample accurate', musical level. You could even have simple controls to decide 'how much' compensation you want - sort of like a compressor. I know some people would just rather use a compressor, but as you've described in your original post, they are very different processes with very different results!

No matter how a recording is done, when there's any surgical stuff going on, manual editing of any form always seems to beat automated functions - I reckon it's that way just to make sure there's SOME form of musicality involved in a project!

[Nahuel]

Originally Posted by halcyo

Oh I agree. You pretty much HAVE to use your ears to be sure that the beginning and ends of phrases work. However, wouldn't it be cool if your DAW could do an 'automatic pass' of volume smoothing automation for you to then tweak by ear? It would save ALOT of time!

lol, I've been thinking of that one myself, would be pretty cool, it would draw the automation curve for you and you'd jus have to edit it...

I think I'll have to work at getting it right from scratch... I have to get myself a fader controler, I like faders, I'm not too bad at cutting on the DJ mixer.

[dhiltonlittle]

Originally Posted by lpuma

lol, I've been thinking of that one myself, would be pretty cool, it would draw the automation curve for you and you'd jus have to edit it...

I think I'll have to work at getting it right from scratch... I have to get myself a fader controler, I like faders, I'm not too bad at cutting on the DJ mixer.

do it man! somebody (presonus maybe) makes one automated fader via usb. that would be a cheap way to get into it and see if it is in your workflow.

[bgrotto]

I'll go back and answer individual questions a bit later, but right now, I gotta pizza on the way and some beers to drink!

In the meantime, yes, those rides were mostly moused-in. I have a Behringer BCF I use for hands-on rides, too. These rides were probably about 80% moused-in, 20% by hand. There's a function in PT called "thin automation" that will smooth out hand-written automation, and ease up processor load. Also, since I opened the session in PT LE at home, it changed things a bit. Specifically, I do all my automation rides in PT HD using the Volume Trim (the yellow line), not the Volume (the black line), it seems opening in PT LE simplifies things a bit.

OK..anyway, it's pizza time...i'll be back tomorrow to address some specific points. Glad y'all gotta kick outta that post! Thanks!

[dhiltonlittle]

word! i'm sippin on some homemade margaritas and ordered in cuban.

peace until tomorrow..

DL

[phillysoulman]

Originally Posted by halcyo

Oh I agree. You pretty much HAVE to use your ears to be sure that the beginning and ends of phrases work. However, wouldn't it be cool if your DAW could do an 'automatic pass' of volume smoothing automation for you to then tweak by ear? It would save ALOT of time!

Im afraid it doesnt work that way..yet

[bgrotto]

Originally Posted by dhiltonlittle

you don't need a fader to automate....these automation moves are "moused" in are they not? if not, you've got some smooth hands bgrotto!

Heh..no, my hands aren't *that* smooth. Like I said, it's about 80% moused-in. Different rides call for different approaches. I tend to do instrument rides on a fader, but vocals are a combination of mousing and by-hand.

For drums and other transient sources, I do almost 100% by mouse.

[bgrotto]

Originally Posted by jikkyboy11

How long did it take you on the vox mix?

Thanx for posting as well bgrotto!

It's tough to say, because I tend to mix the song as the song, meaning, the vox mix is as much a part of the drum mix as it is the strings mix as it is the..well...you get the point.

Automation doesn't take too long; once you get the hang of it, you can do a verse in anywhere from five to fifteen minutes. Like I said, these vox were well-recorded, so I don't think it took more than five or ten minutes per verse (there were three from each emcee, I believe).

They woulda took less time if I weren't doing so much mousing-in...automation by hand tends to go faster, as it's more intuitive (once you get the hang of it, that is!).

The whole mix took about five or six hours start-to-finish. It actually woulda took less, but I started one night after I'd been working all day and I was a little burnt; I ended up making some decisions that weren't very good. So I had to come back the next day and do some touchups for a couple hours. Most of those touchups were vocal rides, actually, and they pretty much tied the track together.

I've asked the producer if I could post clips of each iteration of the mix (the producer's rough, my first night's mix, and the final mix). Hopefully I'll get the OK. I'm not sure when it's getting mastered, but I'll try to revisit this post then and post the final mastered version, so you guys can really see what a mix sounds like from A-to-B-to-C.