PG Music Home Forums General Discussion Off-Topic Mixing Lead Vocal Track Question

I was reading an article where someone suggested chopping up the lead vocal track into parts - example: verse 1, chorus, verse 2, bridge, etc. Then putting each part on a new separate track where that part can be processed individually. Reason being for more control over the dynamics if there's a soft verse and loud chorus or bridge. And for addressing things for example like if the chorus goes high and is more nasally than the rest, etc; that part can be processed differently.

I haven't tried this because in my case I think it might add too much cpu load. Is it a good idea? Anyone tried it?

Thoughts?

All of what you are talking about doing can be done by selecting a section and processing only that.

I do suggest however that multiple takes, and saving them without processing and then making careful notes about what you did to each segment are a great help with your finished product.

Most of my experience in that type of project came with making training videos with DVD's where we would do questions and and later answers, narratives, and commentary. All of it had to sound seamless, although sometimes it took 8 or 10 takes to get the right answer from someone when a pro asking the question needed only one take.

As to volume levels and the final cut it was always cut and pasted into one track after processing, the whole thing was balanced using headphones, monitors, and finally a test cut on DVD on the actual tv set to ensure everything was right. Very specific notes, workflow, incrementally saved work, and careful attention to detail ensured the final product was spot on.

Others might want to extol the use of normalizing a track at the end. I now appreciate that process more than ever, now that most of my upper end hearing has vaporized. I listened to a CD the other day in my car and I couldn't hear the soft parts at all, but the loud parts were too loud. (I have to turn off the hearing aids in the car, it's cold here and the fan noise is in the frequency I need boosted and it's terribly loud, so I can't hear much with them turned on.)


You might want to try using Audacity and singing and playing multiple tracks into it. Then mess with fade in, fade out, and see how it works, then boost volumes, try compression, eq etc. If you make the thing fun and short it's a great learning tool. I don't see how it would hurt processing times, the only time a track is processed is if you highlight a section and apply an effect to it, or if you render the final result. Another thing you can try is to take your short row row row the boat, and copy it, past to another track and shift it just a bit thickening your voice. Panning left and right.

I think Mac has some stuff on a website he can point you to, the whole 'sound' engineering thing is sort of a very technical thing to some, to others a 'black art', and there are many opinions on how to do almost any part of the process. It's probably like learning Russian. It's easy to say basic stuff, buy you can't read Dostoevsky without years of learning!

Hi Josie,

I haven't thought of putting each section on it's own line. There's merit in that. I'm not sure that it would overload the cpu because technically, when (say) track 2 is playing the chorus, track 1 verse is empty in that region and thus has nothing on it to play. It would be different if tracks 1 and 2 were clones and volume envelopes were used to mute track 1 while track 2 played. In terms of cpu, both tracks would then be using the processor (it's just that one of them is too quiet to hear).

A couple of questions...

1. What are your computer specs and operating system?

2. What DAW do you use?

The reason that I ask is because I use Sonar as my final DAW and it allows me to use envelopes to control most of my plugins as well as my volume. This means that I can vary (for example) reverb between choruses and verses by using an envelope. This could potentially achieve the same effect that your describe by cutting tracks into sections.

Regards,
Noel

I would do that, but first I'd want to listen to the track and see if there are any sections that are significantly softer than others before going to all that trouble.

If it works, don't fix it.



--Mac

Josie,

This is almost ALWAYS what I end up with - not by design necessarily, but because I usually do three takes of entire run-through of a song, and then I comp the lead vocal from sections of those 3 takes. Normally I collect those parts into one track and the main thing I'm doing is just slight adjustments on levels on the individual sections. However, with the DAW software that I use - I can process the 'clips' on a track individually by dragging and dropping effects onto those individual clips, in addition to having a 'rack' of effects to use on the collection of clips on the track. I don't do the individual processing often, unless there's a reason. Sometimes this can be useful to add some interest to the song in general. With this capability, it can keep the track count lower if that becomes an issue from a screen-size management issue. I like to have all tracks fit on one screen, no scrolling, while doing mixdown. Sometimes this is limiting, but normally not so.

-Scott

FWIW,
I work kind of like Scott. The assembled vocal ends up on one track.

I may hard write a gain change to a passage here, tune a passage there, or De-Ess a section by hard-writing it, but I don't end up with a single vocal part on multiple tracks *unless* I want to (for example) put a delay on one phrase for a certain effect or something that I don't want affecting the rest of the track.

I know this is how it's done BUT if you're going to sing live I'd try and do the vocal as one take otherwise this piece meal thing is going to be a crutch that'll hurt you in the end.
just my opinion.

I use automation (programmed behavior for faders, effect parameters etc). Like Scott I start with a couple of dry takes on different tracks for each instrument and cut/paste one complete instrument (or vocal)track out of those. This way you end up with one drum-track, one bass-track etc. These tracks and if necessary the used effects are then automated.

I do like so many above. I sing a lead vocal track three times and take the best psrts of each into a forth track. Maybe on verse 2 i miss a word, of do not like the phrasing i add in that line from a track i do like One i have a full track i delete the others , and listen to the complette track, then if anything is still bad i redo that part only.

Interesting responses.

I usually do several takes and then use the best one maybe swapping a word here or there with one of the other takes - the ability to edit that easily is still miraculous to me - but I keep my lead vox track on one track.

Noel, I'm just using Realband. I have an old trial version of Reaper I never figured out LOL. Sonar Le, Cubase Le, and Ableton came with my Emu but I deleted them
because Realband came along at the right time for me and it to grow together, it's the only DAW I have real experience using. grin. My computer is getting long in the tooth. It's a Compaq running XP. You've eased my mind about the cpu use should I ever actually try that idea.

John,

Singing live is nearly all I do. That doesn't mean I'm perfect. My comps usually have only 5 or 6 parts to them; verses, chorus sections, etc.

It's not a crutch. It's a technique. Plus, if I want doubled vox down the road - I've already got the takes to work from. Much of my stuff is very light singing - breathy almost. Doubling is an easy way to give meat to the vocal without it being something other than sotto voce. I like how my voice sounds doubled when I sing quietly more than if it is not doubled.

It's as much of a crutch to do this in recording and mixdown, as amplifying an acoustic guitar for live performance.

-Scott

Hi Josie,

About 10 years ago I had an occasion to go to a professional sound studio with a friend of mine to record a song that he wrote.
He was not a great singer and had to be egged on with lots of confidence comments.
Well the sound man just said to him that he was only getting the levels correct and could he just sing away.
The sound man was telling fibs (lies) because he was recording my friend as he was singing.It was his way of relaxing my friend.
He recorded about four or five takes of the song and then the magic happened.
He took each and every word (yes every word) and chose the best one for pitch and volume
(if, out of all the words in each track there was one that fitted better but the pitch was off, he would adjust the pitch)
He then rendered (placed all words that were scattered around onto one track) first saving the split file in case he needed it again.
He used a slapback echo on my friends voice that was very subtle and the joy on my friends face as he listened was a picture that is burned into my memory.

So Josie, if it's alright for a professional to do it, I'm sure for us to copy that isn't a bad thing.

Best regards
Michee

For some reason I've gotten to where I do use separate tracks for my verse & chorus, but not each one, I'll just have one for verses and one for chorus'. I do like the control it gives me on volume levels.
But, now that I've started using Reaper for mixing and recording any live tracks, I'm not seeing the need as much since you can put splits / in any track and then have level, panning & fx controls over any of those splits in the track. (very cool) Once you play with the program for awhile (OK, 20+ hrs in my case) and do a little referring to the manual, it's actually pretty user friendly and very stingy on CPU usage.

Greg

The practice of taking bits and pieces of multiple takes of the same track and piecing them together to make a single track that works throughout is known as "Compilation".

Or, in the vernacular of the Studio Engineer, "Comping".

I like to use Volume Envelopes for this task, turning all the other tracks down while at the same time turning the one track desired up. This avoids the tedious cut and pasting, pops and clicks when it doesn't cut right at zero crossing point, and perhaps best of all is nondestructive, you can always go back and change just one part of it to one of the other tracks if such is needed while building the mix.

Powertracks can do this via the Pianoroll feature, which has a dropdown for Audio Volume CC insertion with the mouse. I'd like to see it get changed someday such that we can draw the Volume Envelopes right overtop of the Waveform View as done in many other topnotch DAW programs.


--Mac

My opinion is, it depends. The only songs I've had to go nuts over is top 40 stuff, where effects can change from part to part. And it's not just vocals either- it can be any instrument and snare is a big one too. But it's highly dependent on the song, arrangement, and the genera. Being able to automate effects on one track can cut down on having to split everything out too. Just splitting up vocals to multiple tracks won't stress your cpu persay, unless you already have your system up against the wall with lots of effects and tracks. It's not the extra vocal takes and tracks, it's the effects on those tracks that will eat up cpu. And that's why whenever possible, it's better to use effects as an aux rather than an insert. If you use a reverb on an aux buss you can send any number of tracks to it and then automate along the way how much will be sent or not for each track. This is better than having four vocal tracks with four separate reverbs on each track [inserts]. That sometimes is exactly what has to be done though, but when you can, use one reverb [or delay, etc] on an aux buss and send all the vocal tracks to it, each one being adjusted appropriately. It'll save tons of cpu power.

For years I recorded my vocals no less than four takes, comping one part together for the final. I've gotten away from that though. I mostly did it for a safety net with pitch. If I was pitchy on one phrase, well sure enough there was a good one between the other three. And maybe even a more inspired performance to mix and match from too. I do this by default when recording other singers. I get the great take done and then have them 'throw it away' a couple more times and see what happens. With the great take done and the pressure off, sometimes a really inspired performance can pop out. But for my vocals, I've let go of multiple takes. I've also let go of auto tune- a great tool that is VERY labor intensive to use properly. What I do is record one section of the tune and stop and listen carefully. If pitch and performance is good, I'll move on. If it's not, I'll punch the fixes or redo until it's right. When the vocal is done, I have one finished track that I no longer have to futz with in post. I don't have to comp or tune or anything. It's one track ready to be mixed. It cuts down BIG time on extra work. It's called getting it right on the way in, instead of noodling with endless desicions after the fact. It's liberating and frankly, it comes out much better. The key to doing that is the headphone mix. Next time you're doing vocal, put only a smidge of verb on it for inspiration and mute every instrument track that isn't absolutely necessary. What you'll get is a much better tuned vocal right off the bat and usually a much better performance too. After that, the business of splitting out tracks for multiple effects and comping and all is just different ways of mixing. No one way is right or wrong. It's comes down to what works at the time, and I find that the best way is usually the easiest and quickest way to get where you need to go.

So like I said....it depends!

Dan

Sorry but if you can't sing/play a simple 3or4 minute song end to end you shouldn't be presenting a CD as you.
I'm just an old guy that believes in practice over technology.
Yes I use technology in the form of RB but I do sing and play end to end.
I've also done a number of recordings for others on Bass and refuse to use punch-in.

Whatever. Some of the worlds best live performers take months making takes for a CD and have almost no involvement in the process until they hear it near completion.

Some of the same people then do a 'live' DVD but the sound is mixed from multiple performances.

I do object to using the recording and then going live and lip-synching.

John, Did you ever hear the interview Neil Tenant did for the BBC, when he described what it was like recording with Dusty Springfield?

He got the Pet Shop Boys track all set up, gave her the lyrics, adjusted all the monitors etc. Dusty sang the first word of the first line and promptly moved back from the mic. Devestated, Neil went to find out what was wrong, bad key? bad balance in the phones, hopeless song?...

Dusty said everything was fine, but that was just how she recorded - one word at a time!

Now if its good enough for Dusty....

Like I said it's my opinion & actually it's not really an opinion. Yes many "pros" do this I guess but they are in a different boat sometimes. Differing circumstances may dictate that they do this. To me ,at the level of experience of most of us here I surmise, I wouldn't be proud of something I had to piece meal together because i couldn't do it otherwise.What's the point in that?