how do you approach vocal intensive editing sessions?

[joeymusicguy]

Considering this isn't metal, i dont expect a ton of realistic answers but

im curious. when working with a "poppier band" who has nothing but singing... and im talking real singing. how do you approach the editing?

the type of projects im asking about usually contain your standard main take (usually over lapping in several spots, in pop music), your double for every thing in the main... then harmonies, stereo double sections, low octaves, backing vocals (sometimes consisting of a main, a double, and a harmony for one part)...

currently i comp the takes together from the session (find the best, cross em together). then bounce those, (its usually per line) then pitch correct the bounces (graph tuning, this takes loads of time), then make any final adjustments, and carry this out on all takes.

then the grueling process of cutting up the pitch corrected edits per sylable (when possible) for quantizing to grid by hand.

i kinda learned how to do this on my own, and a "tip" from the guy who edited n'sync's biggest record... so i know it can't be total bull shit. im just very concerned with my methods because its literally pushing projects way past deadlines to do all of this to every single part / track / song.

its hard to explain to a label, yeah it literally takes like 6 hours to edit this. and thats just about a whole working day after you just finished 8 hours of tracking another band prior to attempting to do these edits

are you guys doing this any faster? im starting to dread recording new projects with these thoughts in mind... (this song is awesome... but oh man thats going to take so long to edit...) its very frustrating!

 

[Morgan C]

Any reason why you have to graph tune EVERY take? Can't you just do that on the main take, and then do the harmonies automatically? As long as the attack isn't set too high and the notes hit right, you probably won't even notice it on the harmonies (specially with that many!)

 

[Lasse Lammert]

I usually only edit (timing) the vox when a sylable is really off.

 

[paladin shredder]

I usually only edit (timing) the vox when a sylable is really off.

+1

or, one thing i've done before is get an sm57, record me tapping the correct rhythm of the vocal line on the mic, and editing the syllables to line up with those transients.

edit: there you go marcus

 

[egan.]

I think you are probably doing more tuning than most. I also don't think I've ever quantized vocals much beyond "this is too early" or "I need to make this double tighter."
It's also true that one n'sync style records there is a second (3rd, 4th) guy coming in and handling that stuff possible overnight so that you can work in the AM.

 

[Psych05]

Joey, with the insane amount of vocal editing you do, I think you really need something like Melodyne or VariAudio (Cubase 5), wich is basically the same, just Steinberg's ripoff of melodyne. But it's integrated in the DAW which gives it the advantage IMHO. Regardless which of the two, it's just SO MUCH FASTER and easier than doing it the autotune grahical mode way. And you can edit pitch and timing in one window. That smokes Autotune so hard I still cannot believe it took me so long to let go of it.

http://www.celemony.com/cms/index.php?id=plugin2&L=1%253FL%253D1
http://www.celemony.com/cms/index.php?id=plugin3&L=1%2525252525C2%252525252593
http://www.steinberg.net/en/product...ubase5_newfeatures/cubase5_newfeatures_2.html

 

[aramism]

joey funny you should menton this because i worked the other on my most overproduced pop rock project and we did tons and tons and tons and tons of vocal layers and when tracking was done and the band was on tour and i started to dig in and start mixing i was looking at like 40 channel strips of just vocals and kind of scratching my head. i was a bit intimidated for sure. it took me a day just to go back home and write down some ideas as i said to myself "i need my own workflow" oddly enough i looked at that fall out boy producers edition project file and saw how those were arranged and decided the way my workflow brain works i would divide all those vocals into banks. oh and to make matters worse the band had two singers a male and female with totally different timbres and vocal ranges.


so i kind of grouped them for editing and mixing. first i edited the time stuff in logic whether it was time stretching or comping some stray takes that were up in the air during tracking then once i had everything where i wanted and arranged in a certain order i ran what i had to run through melodyne and burned down those regions then just made 1 region per track to make it easier visually. ran some leads through an outboard comp i have and the rest (adlibs, overdubs, background) did in the box. its use preference here but my labeling was along the lines of "fvxlead1" "mvxlead2" "fvxadlb1" "mvxadlb3" "fvxharm1" those were my 3 basic banks where i would treat them and arrange them leads, adlibs, and harmonies, which included backing vocals etc.

but i dunno i guess this ended up being useful for me but the whole idea was to treat everything in groups, including the editing. pitch correction happens last since melodyne is a PITA since you are recording into it so i'd prefer to have the whole track edited and glued to one region then run it through melodyne and bounce it again with the pitch correction. melodyne also has time correction also so i will use that very gently on some parts it cloaks itself very nice so it doesnt sound stupid, but if the song is done rigidly to a tempo this feature comes in extremely handy.

as far as the time spent... yeah it took me longer to edit and sort vocals than it did to mix the whole song.... its definitely a gruesome i'm sure for bands like fall out boy and pop shit there are like 10 second engineers just sitting there editing

 

[HeadCrusher]

Joey, with the insane amount of vocal editing you do, I think you really need something like Melodyne or VariAudio (Cubase 5)

I was gonna suggest the same. I recon you use Nuendo.
So the update to 5 shouldn't be too expensive.

 

[Skinny Viking]

then the grueling process of cutting up the pitch corrected edits per sylable (when possible) for quantizing to grid by hand.

This is completely uncalled for. Like Lasse said ... fix a word or spot that needs it for timing, but every syllable? Thats way overboard ... I don't care how "pop" oriented the music is. Only way it might be needed is when the vocal phrase is purposely being mechanical or something, trying to achieve a rhythmic effect

 

[Metaltastic]

+1

or, one thing i've done before is get a sm57 mic, record me tapping the correct rhythm of the vocal line on the mic, and editing the syllables to line up with those transients.

Not to be confused with an sm57 toaster oven Cool method though dude!

 

[Ermz]

I'd put the emphasis less on editing and more on coaxing the right performance from the singer. Editing every single syllable is way overboard for me. If the vocal track is at a point where that is even close to necessary I consider it a total failure and will tell the band to reschedule after the singer gets his or her shit together. There is only so much of the load you can take on yourself. At the end of the day THEY are the musicians who should be performing the material, not you.

 

[Metaltastic]

Yes, but Joey has to work with bands like Attack Attack!

 

[ahjteam]

I have three policies to vocals....

a) I prefer to slavedrive the musicians than reserve more time for editing. Make it sound as good as possible with tracking.
b) If there is nothing major wrong with it, don't touch it. They're called "nuances".
c) ...but if the vocals suck, retrack it. And if it still sucks after 8 takes, blame the singer. It's just polishing turd.

What I usually do is that I rather use 8 extra hours in tracking fixing the vocals naturally, than try to use 30 hours to fix it in the mix and get ulcers. I try to not autotune the lead vocals if possible. But I do autotune the backing vox IF they need it. I start the editing already in the tracking (I have one track that is labeled "lead rec" for recording and then there is the "lead edit" where I throw the good takes) and I get rid of the shitty takes in the first place. I usually first record a full take from beginning to the end with all the errors (because in 90% cases the first take usually is pretty much the best take the singer sings, I have no idea why... No frustration at this point I guess) and then we listen to it, leave the good parts and take notes where the singer seems to have it hard and then record those parts.

It took me about 2 hours to edit vocals on 5 songs in the last sessions this way (they had 5-11 vocal tracks per song) and what was the most important that the band was happy. It made the record sound more honest.

 

[egan.]

To everyone suggesting better takes, I'll save Joey the trouble and point out that what he is doing is a production goal as much/more than it has to do with flawed performances. I mean, he literally quantizes everything which is impossible in nature.

 

[egan.]

To everyone suggesting better takes, I'll save Joey the trouble and point out that what he is doing is a production goal as much/more than it has to do with flawed performances. I mean, he literally quantizes everything which is impossible in nature.
To Joey, if you want to continue quantizing I would recommend vocalign. This would at least allow you to quantize one track as a guide and then conform the others to that timing. IME that plugin works pretty well.

 

[seditz740novaprospect]

coming from the majority of my work in the above categories i would say that most bigger studios i have been in have had multiple people doing this.. one engineering, editing , and assisting... as well as the producer overseeing everybody..in cases where im stuck doing it all myself i find it best to get the best possible performance and comp as i go to get the best take for the stack that were working on... if need be punching in word by work to ge tthe right performnace out of the artist yet keep it natural unless they want the processed sound.....and tuning is part of the editing process these genras are really vocal intensive and sloppy vocals will really through off the production so tuning is always almost always presant .. grid mode on autotune or waves tune for me almost exclusivly i had melodyne but i am not really feeling it for my workflow i also use vocalign alot to correct mixing projects i did not track


im currently working on a r&b pop crossover that has 115 takes for one song we comped the best performances as we tracked,pocketing the vox so the performances are consistant with the groove . if there flat or sharp make them redo it till they get it. i dont touch autotune till the song is tracked and the performances are spot on..... i feel that vocals dont need much processing from start to finish if its not sounding good from the start its a wrap start over at square 1 the performance, which is the main goal of a producer. getting the best out of who you are working with........

 

[paladin shredder]

Not to be confused with an sm57 toaster oven Cool method though dude!

dur hur!: i r teh smert

 

[paladin shredder]

oh by the way joey, mind sharing your "tip" you got from n'sync's editor?

also, does anyone else find the irony of a band called n'sync needing their vocals edited to match up in rhythm?

 

[JeffTD]

oh by the way joey, mind sharing your "tip" you got from n'sync's editor?

also, does anyone else find the irony of a band called n'sync needing their vocals edited to match up in rhythm?

Pick up Multi-Platinum Pro Tools by Nathan Adam/Brady Barnett - they go in depth about editing techniques.

The tip he's referring to is hard quantizing/auto-tuning background vocals and editting main performance to grid.


Again, as Egan stated... it's NOT AT ALL about getting bad performances sync'd up - it's about achieving a polished, 'perfect' sound that the genre/times demand. Again, pick up that book. There's a philosophy behind editing that essentially takes a great performance and makes it over-the-top. The same techniques we use to make shitty bands sound decent or even good (autotuning, quantizing) can be utilized to make an already great band the very best they possibly can sound.

this all, of course, leaves out the 'natural is better' argument - we won't get into that.

 

[Charlie E.]

This is why I love Logic's comp feature.
I'm currently tracking vocals and doing it section by section. I'll do about 6 takes on each section, then create the comp on the spot before moving on to the next. If I like the entrance on one take better but the rest of the word doesn't sound as sweet it's so easy to fix it by just moving the mouse of the selected areas of each take. This way I don't have to bother editing by hand. (cutting/moving etc)
I feel really stupid for upgrading to Sonar 8 now.
As a matter of fact I'm about to go to guitar center and buy the melodyne plugin so I don't have to bother transferring my vocals to my PC and use roland's vvocal. RIP Sonar