How to make vocal takes sound similar when recording?

[DavidStarz]

Hey there Slutz. I was just wondering what you guys do when you're recording to make all the comped vocal takes sound similar so that the finished product sounds like it was all in one take? Hearing famous people's songs frustrates me sometimes since they make it sound perfect even though it's comped/punched. Do you guys have any techniques or tips on how you make all of your comped takes sound the same? Any ideas should be greatly appreciated.

[ofajen]

Using the Quick Sound Field helps decouple the room and any variations in tone from location, orientation and sound treatment. QSF also minimizes the effects of small movements by the singer relative to nearby reflecting surfaces like music stands. Also, working at a distance of two feet or more minimizes changes in tone due to proximity effect.

Cheers,

Otto

[Oldone]

Truthfully, another person running the board. You need a director when cutting vocals to keep the takes on track, let you know when you need to re-sing parts. Also a person who focuses on organizing the takes on the fly marking the good from the bad. That helps with consistency which is already in place when you go to comp the vocal. Also learn about crossfades. Lastly, automation is your friend. This of course does not cover the mic, preamp, compressor used.

[mattg082]

If all else fails...Pitch and Vocalign during mixing. For me they usually always sound a little different, but usually I try to get them to sing the same thing, but focus on the energy of each take. If the singer is softer and then more aggressive that tends to throw it off unless that's what you're going for. The 3 big ones I focus on first are...Energy, Articulation and Distance of the mic. Maybe that's just those years of doing VO work embedded in my head. I don't worry about room too much because I'm using a SM7b 9 times out of ten and the vocalist is riding the mic a few inches away. If using a condenser mic, same thing I said and the comments above.

[character]

Sing it the same way each time, except better on the parts that need improved.

Seriously though - when working with other vocalists or by myself, I go for complete takes (of either the entire song or an entire verse/chorus) to fix a small part. That way, each take maintains the flow and phrasing of a complete take. Hopefully, if the singer is consistent, those nuances will match up when splicing pieces together. I find longer punch-in takes give me better bits and pieces to work with. Just recording a few words that need fixed often sounds too calculated and out of place when spliced together with another take.

The actual production quality should not vary, if you are recording these parts on the same day with the same set-up. The only way the quality would change is if the parts were re-done at a separate session and equipment was torn down and set up again in between. Even so, I've had good luck with that too because I set it up the same way. That's the benefit of having a good method to start with - the ability to recreate the same results each time if needed.

[mattg082]

Originally Posted by character

Sing it the same way each time, except better on the parts that need improved.

Seriously though - when working with other vocalists or by myself, I go for complete takes (of either the entire song or an entire verse/chorus) to fix a small part. That way, each take maintains the flow and phrasing of a complete take. Hopefully, if the singer is consistent, those nuances will match up when splicing pieces together. I find longer punch-in takes give me better bits and pieces to work with. Just recording a few words that need fixed often sounds too calculated and out of place when spliced together with another take.

I feel the same way! I like to go for complete takes and then maybe some phrasing on the punch in or on some doubles. At least 1 or 2 takes all the way through for consistency. I hate when every singer now comes in, and wants to punch in every 4 or 8 bar measure. It just always sounds a bit off before fixing. This is why I always save warm ups of the first few takes, doing the song all the way through Usually those are the best! Last singer I had refused and we ended up using Wave tune, because every punch in no matter how hard he tried kept coming in at slightly different keys and energy level. Wavetune fixed it pretty well.

[code green]

A pop filter--and telling the vocalist to get right up on it--can give him/her a guide for where to be relative to the mic that can help with consistency b/w takes.

[character]

Originally Posted by mattg082

I feel the same way! I like to go for complete takes and then maybe some phrasing on the punch in or on some doubles. At least 1 or 2 takes all the way through for consistency. I hate when every singer now comes in, and wants to punch in every 4 or 8 bar measure.

Yeah. I like for the singer to at least sing an entire verse or chorus to get the punch-in. Even if I only use a few words or syllables from the new take. I still like to get a complete take just to cut a small portion to splice into the comp track. People just tend to sound different when they only punch in the part that needs fixed. The breathing (and phrasing) doesn't necessarily match where (and how) they would breathe if singing the entire portion. So I try to get them to sing the entire portion.

For most of them, it's easier anyway because if they're fairly new to the studio, they tend to not be able jump in during the middle of a line anyway. They jump time or don't sing clearly - it's easier for them to start at the beginning.

[Kiwi]

If the arrangement is building over the course of the song, then each section can't help sounding different anyway. If you have harmonies and FX changing all the time, it should sound good and interesting.

Vocal tone changes from take to take - so ideally you want to use the first, second or third take - and if you are struggling to get it after that, give up and try another day.

On the other hand ... if your arrangement gets more complex as it builds, with more layers of vocals, it may not hurt to have the vocal tone thin out a bit in the later sections.

With pitch software or hardware, there are plenty of options with formant shifting and octaves and micro tuning to thicken up a vocal. If you like that sort of thing.

Sometimes a short transition to a totally different effect (cliche' radio voice or phaser or something more interesting) can act like a 'palate cleanser' for the next course.

As long as the energy level is right for each part of the song, I think tone can vary all over the place and not be noticeable.

[DavidStarz]

Do you guys ever have it where you recorded an artist (or yourself) and they don't need ANY pitch correction or something of that matter after comping/punching the takes? Like it sounded perfectly fine just the way it was? I find it extremely hard to believe how you could make all of these different takes sound the exact same in the end. Maybe there's just something I'm not getting? o.O I basically record everything I do by myself and then send it to an engineer I have to mix the song.

[Kiwi]

Depends on context and genre. Some genres can tolerate more imperfection than others. Sometimes the first take is magic and you don't want to change a thing.

Pitch problems are another issue - I would suggest you don't confuse those issues with, for example, changing vocal timbre over a session.

Tracking yourself if harder. With a client, you can tell them to just warm up will you check levels - while actually recording keeper takes. Then, watch them freeze up when you say "OK - Take 1 ...". So sometimes you can catch a client when they are relaxed and it can be better than when they are conscience they have to get it right.

Plan on getting good takes that won't need comping. Don't over-extend yourself.

Do you sing live? If not - it can help to break out of a bedroom producer mentality of non-linear editing. Live has to be linear ... get it right in one take.

[character]

Originally Posted by DavidStarz

I find it extremely hard to believe how you could make all of these different takes sound the exact same in the end.

The different takes don't literally all sound the same (in fact, the parts you're trying to fix will hopefully not sound the same although some singers make the same mistakes in the same parts over and over). But the entire take doesn't have to be identical to the previous take. It's just ideal to get the parts you comp together to match up to sound like they belong in a single take. The vocalist should be able to basically sing the song similarly more than once, and hopefully any differences or variations will be better than the parts they are replacing. You're recording the same singer; their performance should be similar and the parts should pretty easy to match unless they're doing significantly different performances on each take (or the recording set-up is changed for some reason).

If you're doing the comp edits yourself, it's good to practice and get really good with finding edit points that aren't noticeable. Sometimes two takes that are performed well enough by the vocalist to be comped together can be comped together poorly and the edit is noticeable. Likewise, sometimes two significantly different takes can be spliced together just right (at the right edit point) to sound like it was all done at once.

[Apollo Soul]

Well, for me, it starts with singing it MANY times... like just letting it roll and singing the chorus or whatever at least a few dozen times. I sing it for a while until I get it just how I want it... then let the DAW roll as I do my best to sing it exactly that same way on pass after pass after pass after pass after pass...

Then LOTS of swipe comps to get the best take.

Then play that take while I do lots of swipe comps to find the best match. If no match possible for a certain part, go back and find a different bit for the first take that can be matched.

And so on..

And so on..

Eventually, if I've got 4 or so matched very closely... then vocalign is a wonderful and easy tool to adjust their timing.

Then extensive fader riding, vocal rider, etc... for matched levels.

Then tuning if necessary... although a little goes a long way when doing clustered vocals or you get weird phasey issues.

Finally, I compress them all together in a group to wrap them in a common envelope.

Voila... after many painful hours, you have incredibly tight matched vocals suitable for slickly produced pop music.

Much easier to get a looser feel.

[hello people]

I don't see why any bits should be different from others except for if mic distance or the room have changed?

Can the singer sing the song or not?

If it's down to punching in 4 word phrases then holy shit...give it away.

[Jay Dee]

Originally Posted by hello people

I don't see why any bits should be different from others except for if mic distance or the room have changed?

Can the singer sing the song or not?

If it's down to punching in 4 word phrases then holy shit...give it away.

Word...

[shortstory]

Originally Posted by DavidStarz

Hey there Slutz. I was just wondering what you guys do when you're recording to make all the comped vocal takes sound similar so that the finished product sounds like it was all in one take? Hearing famous people's songs frustrates me sometimes since they make it sound perfect even though it's comped/punched. Do you guys have any techniques or tips on how you make all of your comped takes sound the same? Any ideas should be greatly appreciated.

I never record vocals with EQ.

[chris carter]

Honestly, the way you get it to sound the same is to have the singer sing the same way and maintain their position on the microphone. Some singers are better at this than others. As well, the producer has to be on their ps and qs making sure the takes are consistent - that's part of their JOB.

Personally, I'm not the biggest fan of comping. I would only do it if I had to. I would prefer to just punch - that way you know what you have. And if you are a good producer, you don't need comping. While comping CAN be a good technique, far far far too many people do it instead of committing because they just don't know what to commit to. And then you have all these takes that all have the same exact screw up on every one... ugh. So if you comp, you'd better to the comps before the artist leaves or you can be in for a nasty suprise.

On a side note, a good trick for making sure doubles are locked in solidly (so you don't have to resort to vocalign and stuff) is to pan the original left and the live take right (this is on the control room speakers or producer's headphones, NOT in the artist's headphones). Then listen for summation. You can distinguish all kinds of differences between takes that way from intonation to timing to tone. Then, as the producer, you can tell the artist EXACTLY what they need to do on the next pass to match better. Old trick. Still works like freakin' magic. No vocalign necessary!

[Kiwi]

Originally Posted by hello people

I don't see why any bits should be different from others except for if mic distance or the room have changed?

Can the singer sing the song or not?

If it's down to punching in 4 word phrases then holy shit...give it away.

Really? Do you record vocals much? All singers I know of their vocal tone changes over the course of a session ... if you want soft, sweet stuff, get it early. If you want harder, edgier stuff, let them break in their voice for a bit and catch it just before they lose their voice ...

[hello people]

Oh yeah...just like John Lennon when he sand Twist & Shout at the end of a long day's night

[Ron Vogel]

I don't know if this has already been posted....but if you use the same singer it will sound the same unless you go crazy on the coloration coming in.

[Kiwi]

Originally Posted by Ron Vogel

I don't know if this has already been posted....but if you use the same singer it will sound the same unless you go crazy on the coloration coming in.

Not only has it previously been posted, but it's also previously been exposed as being a load of bollocks.

Do you record vocals much? Have you never noticed that a singer's vocal tone changes over the course of a session? Really?

[joeq]

Originally Posted by DavidStarz

Do you guys ever have it where you recorded an artist (or yourself) and they don't need ANY pitch correction or something of that matter after comping/punching the takes? Like it sounded perfectly fine just the way it was?

all the time

it so depends on the singer and also the style of music
good singers in the jazz/blues tradition, for example, tend to have excellent chops, great pitch, AND they are working in a medium that allows for some variation

pop stuff often needs to be tighter
the singers are often less seasoned
and of course the 'style' is trending towards ridiculous amounts of tuning

Hearing famous people's songs frustrates me sometimes since they make it sound perfect even though it's comped/punched.

Consistency is the mark of a great musician. Even in our trashy reality-show modern world, fame often still has talent behind it. Most of the credit has to go to the singer for singing it the same, and to the producer for keeping them on track, and less to their engineer's tricks for recording and comping.

I find it extremely hard to believe how you could make all of these different takes sound the exact same in the end. Maybe there's just something I'm not getting?

.

as character said, it's easy - just sing it exactly the same! What is your background singing-wise? Taking lessons? Practicing every day? Gigging every week?


Do what you have to do to maintain levels, distance to mic, location in the room, write it all down. As the Engineer, what more can you do? After that, it is up to you as the Singer and you as the Producer. And as mentioned above, producing yourself is less than ideal.

I basically record everything I do by myself and then send it to an engineer I have to mix the song

And do you think it sounds choppy? What about the editing itself? Are changes in vocal tone disturbing you? Or could it also be stuff like double-breaths, odd phrasing and mis-selection of edit points that are breaking the flow? Are you picking the takes with the best feel - or the ones with the fewest mistakes? That kind of thing.

Sure there are variations in any performance, but the continuity of the music should propel the listener 'through' the edits, and towards hearing it as an intended whole. If the singer himself is propelling the vocal line through the music, this should happen. Perhaps if he is focusing too much on some technical aspect when punching - and not enough on forward motion....

For myself, I am constantly surprised at how many edits "work" when I think they ought not to, rather than the other way around.

[hello people]

Originally Posted by Kiwi

Not only has it previously been posted, but it's also previously been exposed as being a load of bollocks.

Do you record vocals much? Have you never noticed that a singer's vocal tone changes over the course of a session? Really?

Well why is the vocal session going so long? You can only do one song at a time right?

What are you doing? Recording one word snippets then breaking for lunch, then doing the next 6 words, word by word in the afternoon session?

[BillSimpkins]

I've always worked with the singer to get consistent sounding takes.

[doorknocker]

I don't see the problem. I had a country singer in this week and he did 12 songs with two takes each in a row.
We were done within 5 hours and the vocal tone hardly changed.

I really would go for whole takes whenever possible and work relatively fast. Just have the singer do another take instead of fretting over details.

The problem with heavily comped vocals is not tone but lack of conviction/ flow on the singer's part. In other words: If you got several takes that were sung all the way through you can still comp the best bits and it will sound organic.

[BHickey]

What exactly are the problems that you are having? I find that often the answer for to your question is not equipment related, but preparation.

If its a tone thing, trying to get all the takes during the same session should help. If you are doing takes over multiple days, taking careful note of all equipment used and the exact settings on the equipment (including mic placement in the room) can help (taking pictures of the gear will often suffice for this).

If its a pitch or timing thing, it comes down to rehearsing, picking the takes more carefully, and making your edits more carefully. Persoanlly, I will make an edit that lets a bad note or pronunciation through if it is the best option that doesn't distract and take the listener out of the song.

[kelvyn]

Lots of interesting comments...

You have a great singer where you don't need to comp so much.
Have a singer that comes prepared for the session.
Have a singer whose voice doesn't get frazzled and can sing 20 songs a day.
Have singer who is mentally stable and can deal with 'Red Recording Light Syndrome'.
Have a singer who understands that words have meaning and inflection is as important as technique and intonation.
But what if you don't have a singer with all these fantastic attributes?

I think what the OP is asking for is a solution to his problem and not the different idealized situations being referred to here.

I work with a lot of singers who can't sing but have something in their voices that is appealing but needs a lot of work. I actually have the same situation as the OP right now where a singer whose vocal abilities are fair but her technique is not... She is a very talented songwriter with an original way of expressing herself but we have to drop words in sometimes... no big deal as the end result is what counts (like an author rewriting some sentences over and over) but the inconsistencies in timbre are difficult to smooth out. All I can say is... trial and error and a huge amount of patience and guidance on behalf of the engineer/producer. Singers are fragile entities.
Good luck

[Apollo Soul]

Originally Posted by doorknocker

I don't see the problem. I had a country singer in this week and he did 12 songs with two takes each in a row.
We were done within 5 hours and the vocal tone hardly changed.

I really would go for whole takes whenever possible and work relatively fast. Just have the singer do another take instead of fretting over details.

The problem with heavily comped vocals is not tone but lack of conviction/ flow on the singer's part. In other words: If you got several takes that were sung all the way through you can still comp the best bits and it will sound organic.

This bit really depends on the style. I loop the same section and do somewhere between 40 and 50 takes, but the reason is to get the tightest matching takes so they can be stacked 4 deep in parts.

Doing pop production where you bounce back and forth into phrases with 4 or more perfectly matched takes is a whole different animal than doing the traditional one person sings a song the whole way through deal.

[audiotech]

I sometimes have some problems with talent when they have to come back to the studio for a re-cut or to replace a line or two. Although I usually can get the quality of their voice very similar, if not a perfect match, it's usually their vocal energy level that causes the most difficulties.

Dennis

[joeq]

Originally Posted by kelvyn

I think what the OP is asking for is a solution to his problem and not the different idealized situations being referred to here.

But these idealized situations are all pointing at the correct solution. The solution to the problem really does lie in improving the singing and producing elements. Not the engineering elements.

There's only so much the engineer can do in tracking, editing or mixing techniques. Get out a tape measure and keep the distance to the mic the same? Sure, but the singer has to have that flow, that forward momentum. He needs to sing consistent to sound consistent. Then the problem takes care of itself.


The OP is asking how famous artists achieve their consistency. One thing is that they are not alone when they sing. Most have a producer in the CR directing the performance and deciding what needs to be done again, punched, or comped so the singer can just sing. IME, most singers come up a notch when they have someone with good ears taking some of that responsibility off their shoulders. And if that someone is actually a real producer they can come up several notches.