Vocal Recording and Pitch Correction

[vanityaffair]

I'm beginning work recording vocals for my band, and my singer gets VERY frustrated past the 5th or 6th take of a part. That really means nothing, but my question boils down to this: How much is too much pitch correction? I want to get a proper performance, but our singer can't handle 20+ takes of the same 8 words, and time is a factor. I also do not want to rely on autotune features to band-aid what was less than a stellar performance to begin with.

[Cosmic Pig]

Holy crap. I hardly know where to begin. You must be the producer. Sputter sputter... holy crap.

Maybe I'll just leave this to the more experienced guys around here before I wear out the backspace button.

Cos.

[otek]

How much is too much cream in your coffee?

It depends on how much coffee you've got, how much cream you want, and whether or not you care about unfortunate side effects.

This question is impossible to answer, except in a very general way: "As little as possible" would be my standard answer.

Select the best phrases from your takes, and comp them on one track.

And as for Autotune, use graphic mode, much more precise.

[Tim Halligan]

Boy...what a conundrum.

First question that really needs to be asked is: is your singer really ready for the studio?
Does he know the material? Is he able to recognise the flaws in his performance? Does he care enough about the material to want to do it better, or is this just for shits 'n' giggles?

Is it you? Are you a complete tyrant/asshole in the studio? Are you asking for a performance that no-one on the planet could deliver? Are you trying to - perhaps subconciously - be "Mutt"? Are you asking for "20 takes of the same 8 words" because it really is taking that many takes, or because you think that's how many takes it should take?

I would think that asking any singer for more than a reasonable number - whatever that is - of takes is just going to wear out their voice...giving you matching issues. After they become tired, you're just heading for "flogging a dead horse" territory - quickly. Keep it fresh, move on to something else and then come back to it.

There is some fairly heavy psychology required when dealing with lead singers...be mindful of this when pushing the talkback button.


Cheers,
Tim

[malice]

I prefer "soft pitch correction" albeit full auto tuning.

Most of the time, keeping a small amount of "out of tune" parts helps keeping a more natural result.

So basicaly, I make sure important long note remain in tune and leave as much as the rest of the performance intact.

The bottom line remains: as little as possible.

good luck

malice

[studjo]

Most of the time, keeping a small amount of "out of tune" parts helps keeping a more natural result.


LOL I fear I fight with too much naturalness - even when I tune those fucking vocals.

But as everybody said - as little as possible and much as needed (I use graphic all the time - the auto function in the graphic mode works way better than the real auto mode/at least in my vintage version of autotune)


Jo

[otek]

I use graphic all the time - the auto function in the graphic mode works way better than the real auto mode/at least in my vintage version of autotune.

Uuuhh......

Clue me in if I'm missing something here, but you're talking about the "make auto" function in graphic mode, right?

Now, assuming you don't modify the waveform manually, what is the difference between that and the "real" auto mode? :?

[dikledoux]

Call me crazy, but if a singer can't get the take in 6-8 tries, then they're either just having a bad voice day (singin' through the snot) or that's as good as they're gonna sing. Either way, why beat them up about it? Comp the parts that are keepers, fix what you feel you HAVE to fix (which may be nearly nothing if you hate editing people's parts as much as I do) and move on.

If the singer's just not in good voice - put the vocals off til they are. And make sure you tell the singer your plan in a positive way - that you're putting it off because you want to catch them when they're ON because the results will be worth it. Don't just tell them they're sucking today and better luck next time. Vocals are SO psych dependent.

Now, if the singer WANTS to go at it 20+ times, great. But if they're not into it, the performance will reflect that.

dik

[studjo]

Uuuhh......

Clue me in if I'm missing something here, but you're talking about the "make auto" function in graphic mode, right?

Now, assuming you don't modify the waveform manually, what is the difference between that and the "real" auto mode? :?


the sound???
very happy
dunno why but it defenitely works different

Jo

[Pimp-X]

I think you're imagining it.

[studjo]

boys you don't have to believe me - listen for yourself and still if you can't get to the same results I'm absolutely positive what my experience is.

Jo

[Carlo]

I think that the singer needs to hear the notes they are trying to follow...let them listen to playback 10 or 20 times, instead of singing it so much.

You can track pitch, or ya can't...that simple. Of course,performance sitiuations are different...personality rules, but in studio, they can, or they can't.

Everybody would do it if it was easy...and as far as Auto-tune...don't make me pop a cap in yo ass!

[J.G.]

Premature e-post-u-lation...

[J.G.]

Here's a lil trick I stumbled upon once, (using myself as the guinea pig too), which came about one day while singing on a tune that just wasn't in my veins; I didn't really love the style of the song nor did I appreciate the composer's direction for me to take stylisticly, or the notes, (which were written in stone), and I was dealing with not only the buddings of a cold, but with the smelly entrails of a rather heated convo. about said carved-in-stone notes and style-delivery expectations. The song too, was in a key a lil higher than I would've preferred--digressing.

So, I sang what needed to be sung, and while it turned out alright, there were some notable struggling with some of the higher notes.

I fired the voxings into Melodyne, corrected them all to a frightful 100% accuracy, and then I resang the whole thing, along to the corrected tracks.

This provided a kind of comfort zone, and allowed me to disassociate myself, (and my ego! ; ), not only from my own hyper-self-critical voicings, but from all the reasons why I didn't like the project, and lo and behold, the re-sung track was just about right on the money, pitch wise, and equally imortantly, delivery-wise.

There is sometimes a very useful, albeit a rather roundabout place for tools like pitch correction in the recording process, IMHO.

Perhaps this trick'll do the trick in your situation there, vanityaffar..?

[Brendo]

Yeah, singing along with a 100% in tune guide is something I've used in the past with some success.

[Fulcrum]

I fired the voxings into Melodyne, corrected them all to a frightful 100% accuracy, and then I resang the whole thing, along to the corrected tracks.

Thanks for that J. I'm going to try that over the next couple of days. I'm still having a fucking. Hard. Time with going sharp all over the place, and I'm desperate.

[otek]

boys you don't have to believe me - listen for yourself and still if you can't get to the same results I'm absolutely positive what my experience is.

Good for you. :)


If one would theorize for a bit, I suppose the difference would be that in Graphic Mode, using the Make Curve function actually works on a piece of audio where the pitch has already been pre-detected - Autotune, therefore, doesn't have to work in real time, and can be more precise.

Just a thought. Whatever. In so much as I even use Autotune anymore (I find Melodyne gets the job done better for the most part), I almost always use graphic mode anyway.

[Bob Olhsson]

For openers, get rid of the headphones!

[Mixerman]

I'm beginning work recording vocals for my band, and my singer gets VERY frustrated past the 5th or 6th take of a part. That really means nothing, but my question boils down to this: How much is too much pitch correction? I want to get a proper performance, but our singer can't handle 20+ takes of the same 8 words, and time is a factor. I also do not want to rely on autotune features to band-aid what was less than a stellar performance to begin with.

You have to adjust your recording techniques to the personality of the performer. If I'm working with a singer that gets frustrated easily, I approach the recording process differently, than with an artist that is a workhorse.

No matter what though, 6 takes is a lot of takes to do with an artist. Sometimes that's necessary because you discover some new approach by the third take, but in general, I stop at a max of 6 takes for a vocalist.

Some vocalists, and on some songs, it's better to record one section at a time. Work on the verses first. Then work on the choruses. This is often the case with new songs, or songs that the singer doesn't have a clear-cut approach for. In general, I'd prefer NOT to use this approach. It's usually better to get a few "performances" of the song and mix and match. But don't get bogged down by purist pursuits. Your goal is to get the best performance possible, out of an artist on THAT day.

Which brings me to my next point. Sometimes I know I don't have what I need our of the singer. I do not divulge this information at the time. I comp it the best I can, and then play it for the Artist/singer. If they show dissatisfaction with the performance, then I make suggestions and we try again. If they DON'T show dissatisfaction, I put it to bed and come back to it later. If they're on the fence, I ask them to try and beat the comp with one performance. Sometimes by the singer knowing that they have something good in the can, they let loose a little more. Sometimes they start to feel so good letting loose, that they decide to sing it 3 more times. So long as THEY are the ones that want to sing it because they think it's getting better, it's fine.

It is essential in the psychology of recording, that an artist feels successful as often as possible. Especially the singer. If they begin to lose confidence, you're screwed. Whatever you do, don't let your singer lose confidence.

What I do, is revisit songs throughout the process. First of all, it is possible that I might be unhappy with a performance when it's first comped, and then change my mind upon coming back to it. Remember, we are subject to over stimulation, and sometimes our judgment is momentarily clouded. Secondly, this allows me to broche the subject with the singer, when they are more open to criticism, and less likely to put their guard up, get frustrated, or lose confidence.

Now, let's talk about tuning.

If you can avoid using it, do. If you DO tune them, don't do it in front of them. Some singers don't mind the tuning functions available, and prefer to use them, but I try to get them to leave during the tuning process too. Mostly because they will possibly want to tune TOO MUCH! For some singers, you risk damaging their confidence. For others, you risk their own over aggression destroying the performance.

Lastly, and in general, I do not work lines from the perspective of tuning. Work them from the perspective of performance. I'd much rather have an out of tune performance, than a perfectly in tune but uncompelling presentation. Now, with a singer that is constantly out of tune, this suggestion can be difficult. But you have a better shot of having someone sing in tune, if they are giving you attitude. And of course, you can always tune them anyway. So why not go for the performance?

Oh yeah, and if the singer is consistently out of tune, chances are their headphone mix is fucked up. I've found that some singers can't even use headphones, and in those cases, I set them up between a pair of speakers.

Enjoy,

Mixerman

[Brendo]

Uuuhh......

Clue me in if I'm missing something here, but you're talking about the "make auto" function in graphic mode, right?

Now, assuming you don't modify the waveform manually, what is the difference between that and the "real" auto mode? :?

AFAIK the make auto function makes a curve, with the same imperfections, but puts the bulk of the note in tune.

So a scoop into a note, wavering slightly sharp towards the end would stay the same, but be bumped onto the line. Like, the scoop will remain, etc.

http://womb.mixerman.net/attachment.php?attachmentid=184&stc=1&d=1163382496

[otek]

AFAIK the make auto function makes a curve, with the same imperfections, but puts the bulk of the note in tune.

Not quite.

The Make Auto function draws the tuning curve according to the parameters set in the auto section, including scale, retune and tracking parameters.

So the detection/processing algorithm is most likely identical to the auto mode, which is why I got suspicious of Studjo's comment about the perceived difference. That's why I theorized that if there is a difference, it may be due to autotune not having to do both detection and correction in one real time pass.

[Brendo]

Not quite.

The Make Auto function draws the tuning curve according to the parameters set in the auto section, including scale, retune and tracking parameters.

Doesn't the graphical section have a separate set of retune and tracking knobs?

[studjo]

Otek
I came to the same theory when I first realized this strange behavior of AT.
AFAIR you set the parameters of "make auto" in the auto mode - but that could have changed with a new version Redface (I'm still on AT3 or 4??? dunno - at least it sounds vintage and I like it :D )

Jo

[Bob Olhsson]

Oh yeah, and if the singer is consistently out of tune, chances are their headphone mix is fucked up.Sometimes just flipping the phase on the vocal mike makes all the difference in the world. Also beware that some reverb units shift the pitch flat which can really throw off people's sense of relative pitch.

[Goes211]

Excellent thread, golden info.
Thumbsup

[otek]

Doesn't the graphical section have a separate set of retune and tracking knobs?

Yes it does. They basically give instructions to the software on how to interpret the tracking curve.

I did make one mistake before: The Automatic Mode Tracking control is not involved in the Make Auto computation, since the pitch has been pre-determined already. This may actually be the difference in StudJo's case, since the Automatic Mode is dependent on the proper setting of the Tracking control.

[Mixerpuppet]

You have to adjust your recording techniques to the personality of the performer.

Enjoy,

Mixerman

I cut everything because this is the key.... adjusting your game to the player...

I have a friend who has spent alot of time tring to cut a record in 4 or 5 different studios at a significant expense. One night we recorded 3 songs and the playback was pretty good but he wasn't satisfied. So we spent the rest of session talking about everything Mixerman says in his post. We spent a little more time working the issue. His main problem is his mother was an opera singer and so his ability is quite impressive but he's so loud he can't hear himself and goes flat at the end of each phrase... We tried various headphone techniques, and nothing was really working... So we started working performance issues from a stand point of him trying to debug it himself... Finally he quit trying to sing as loud as he could and tried to optimize the volume to gain a consistancy in pitch...

He finally bought a little home recording setup to work on things to his satisfaction, then he's drop the money for a vanity record.

I offered the OttoTuner route and it was rejected... (It's nice to to someone interested in improving rather than cheating :)

Anyhow... I tried Mixie's advice repeated here except If I remember the orginal post from years gone by that Mixie also included Diplomacy in conjunction with Psychology... (early recpit thread)


None of this helps the cases where the vocalist might be tone deaf... (for those running small shops with no label work)

Not even Autotune...

[otek]

Whenever I talk about Autotune, Melodyne etc. in forum threads, I always do it with the tacit understanding that everything, absolutely everything has already been done by the artist/producer/engineer to ensure the optimum vocal performance.

It's like talking about EQ - I don't even start on EQ until the mics are placed in their optimum positions. Ever.

The way I see it, if the performance is not there, autotune cannot and will not fix anything.

[Mixerpuppet]

Whenever I talk about Autotune, Melodyne etc. in forum threads, I always do it with the tacit understanding that everything, absolutely everything has already been done by the artist/producer/engineer to ensure the optimum vocal performance.

It's like talking about EQ - I don't even start on EQ until the mics are placed in their optimum positions. Ever.

The way I see it, if the performance is not there, autotune cannot and will not fix anything.



Great Caveat Otek... Thumbsup

It isn't always obvious with such a wide number of ways people get into the recording path...

Some folks may break out the toolbox before anything is even broke...

I had a friend who was going in for his 3rd album and wanted my amp and me to help get "my tone"... The engineer put the amp in a closet, cranked everything up, threw a 57 in the middle, and then before the first power chord was finished started with the radical eq... twisting like he was trying to pull the knob off the console... it was mad....

I took a closer look at the console from behind the engineer and the console was never zero'd from the previous client.... in fact nothing was cleared.... EQ's comps, efx....

I spent at least a month at my first rec job zeroing, pulling patch cables, dusting, cleaning etc...

It does take a little time to learn... pratice and implement.

The album sucked in the end, a cameo of the engineer doing a out of place guitar solo, mastering was crap etc...

Maybe it was the ADATs Uh oh

[Brendo]

Yes it does. They basically give instructions to the software on how to interpret the tracking curve.

Y'know, somehow I knew that.

So in graphical mode, you draw in the target notes.. and then set the graph mode's retune and tracking to set the uh, attack and threshold of the tuner?

[Calvin]

It is essential in the psychology of recording, that an artist feels successful as often as possible. Especially the singer. If they begin to lose confidence, you're screwed. Whatever you do, don't let your singer lose confidence.

Mixerman

Lots of great advice here, including the tidbit quoted above from Mixerman. I worked on a few sessions (as a guitarist) years ago for a "producer" whose main technique to try to get performances from the artist was to continually point out "flaws". He did this in a fairly mean-spirited manner. The result was always stiff, safe, boring, scared performances. And that's for the reasonably talented artists. For those artists with less experience/talent, the performances could really deteriorate over the course of a bunch of takes until nothing was usable (which of course meant that the "producer" felt he had to keep trying...). This approach did not tend to foster warm and fuzzy feelings among the session's participants. :Mad:

Calvin

[Mixerman]

Sometimes just flipping the phase on the vocal mike makes all the difference in the world.

Interesting. I've never tried that.

Thanks, Bob!

Mixerman

[Carlo]

Sometimes just flipping the phase on the vocal mike makes all the difference in the world. Also beware that some reverb units shift the pitch flat which can really throw off people's sense of relative pitch.

Right on Bob...even though it will make you think you sound better, reverb almost always drags my pitch down, and like you said earlier, pull an ear out of yer 'phones, and it can really help!

[Bob Olhsson]

...even though it will make you think you sound better, reverb almost always drags my pitch down,..Real plates and chambers don't do that.

[malice]

Right on Bob...even though it will make you think you sound better, reverb almost always drags my pitch down, and like you said earlier, pull an ear out of yer 'phones, and it can really help!


I'm a bit adament to track vocal without reverb for that reason.

Sometimes it is an issue with some singer, but when they hear them selve in the CR, they sudddenly pay attention to what I tell them.

malice

[r.baby]

I fired the voxings into Melodyne, corrected them all to a frightful 100% accuracy, and then I resang the whole thing, along to the corrected tracks.


Wow. I NEVER thought of that. Good trick, thanks!!

[J.G.]

You're welcome.

: J

[vanityaffair]

thank you guys sooo much, it's hard to try and wing half of my recording process based on my own intuition, awesome to hear from the guys whose knowledge i can glean, albeit one thread post at a time. if this new locale for our boards will allow me to upload audio, i'll be sure to post a sample when it's done. if not, i'll find some way to link it.

[nobby]

Oh yeah, and if the singer is consistently out of tune, chances are their headphone mix is fucked up. I've found that some singers can't even use headphones, and in those cases, I set them up between a pair of speakers.

Enjoy,

Mixerman

I felt that some here were assuming the original poster had a great cue mix.

I had a problem with this a while back. I kept playing back takes where my pitch had gone off, and I could sing it in tune with ease in my car. I don't have oughta toon or melodyne or anything like that, so I was forced to actually address the problem and not just the symptom. (And call me old school, but I like to be competent at my craft.)

I worked on my cue mix, pulled one headphone away from my ear and sang with a hard (glass) surface in front of me, and voila, problem solved.

[otek]

I worked on my cue mix, pulled one headphone away from my ear and sang with a hard (glass) surface in front of me, and voila, problem solved.

The only problem I could possibly see with that is, if you can hear the reflections, so can the mic.

Caution is recommended around reflective surfaces as always.

[jimmy v]

I'm still having a fucking. Hard. Time with going sharp all over the place, and I'm desperate.

In my limited experience this animal can be tamed.Generally this is a fear issue.Fearing that I wont be able to reach the note I push rite past it.I just take afew seconds do do some "square breathing" and problem solved,I'm relaxed and hit the note.When working with singers I tend to spend some time teaching relaxation tecnics.Taking a few min to relax and breath can work wonders.Singing uses a lot of air.Take some time to put some back.

[Brendo]

I'm just today discovering how much faster I am at tuning vocals using Melodyne, and how much more natural they're coming out - mainly because I'm using the free version that comes with Alsihad, so you can't run the pitch correction macros over it, and you can't edit the pitch curves, so i'm using much broader strokes.

[Comte de St Germain]

Which brings me to my next point. Sometimes I know I don't have what I need our of the singer. I do not divulge this information at the time. I comp it the best I can, and then play it for the Artist/singer. If they show dissatisfaction with the performance, then I make suggestions and we try again. If they DON'T show dissatisfaction, I put it to bed and come back to it later. If they're on the fence, I ask them to try and beat the comp with one performance. Sometimes by the singer knowing that they have something good in the can, they let loose a little more. Sometimes they start to feel so good letting loose, that they decide to sing it 3 more times. So long as THEY are the ones that want to sing it because they think it's getting better, it's fine.

It is essential in the psychology of recording, that an artist feels successful as often as possible. Especially the singer. If they begin to lose confidence, you're screwed. Whatever you do, don't let your singer lose confidence.

What I do, is revisit songs throughout the process. First of all, it is possible that I might be unhappy with a performance when it's first comped, and then change my mind upon coming back to it. Remember, we are subject to over stimulation, and sometimes our judgment is momentarily clouded. Secondly, this allows me to broche the subject with the singer, when they are more open to criticism, and less likely to put their guard up, get frustrated, or lose confidence.

Red#1: Usually where the PASSION starts to come back.

Red#2: Above all else.




So why not go for the performance?


That's what tracking vocals is all about. It's easy to get it right, getting a real performance is much harder. IMO this is where many projects go bad; not enough time to revisit vocals which is sometimes solved by forcing "new" scratch tracks before overdubs. With other projects sometimes the scracth IS the one, or at least the ones all others should be pushed by.

[Mixerman]

Thanks for that J. I'm going to try that over the next couple of days. I'm still having a fucking. Hard. Time with going sharp all over the place, and I'm desperate.

Have you tried singing using speakers instead of headphones?

Mixerman

[Mixerman]

The only problem I could possibly see with that is, if you can hear the reflections, so can the mic.

Caution is recommended around reflective surfaces as always.

True this you speak of Master Otek. But you know what? Personally, I'd rather have a weird sounding vocal that's kick ass because the singer is confident (because he's feeling in tune) than some of the other alternatives available to us.

Hell, listen to the last Johnny Cash record. Terrible vocal recordings, but great fucking vocals.

Mixerman

[otek]

Personally, I'd rather have a weird sounding vocal that's kick ass because the singer is confident (because he's feeling in tune) than some of the other alternatives available to us.

Oh yes, I agree 100%.

A while ago I had the singer in the control room, with an M88 handheld, big monitors blasting. He was dancing around, doing this quasi-Massai-Warrior dance routine wearing nothing but his shorts.

Yes we had leakage. Yes the vocal performance was awesome. No, the dance routine and wardrobe weren't my idea. :lol:

[malice]

Have you tried singing using speakers instead of headphones?

Mixerman

That's a great trick.

Done properly, it's often leacking less than with singers asking for too much volume in the cans.

Tuning improve by a night and day factor with speakers.

malice

[seagate]

That's a great trick.

Done properly, it's often leacking less than with singers asking for too much volume in the cans.

Tuning improve by a night and day factor with speakers.

malice


How do you stop the speakers from leaking into the mike?

[malice]

How do you stop the speakers from leaking into the mike?


Positioning.

You must form a perfect equilateral triangle between the two speakers and the mike. both speakers must be at the exact same distance from the capsule.

Then you feed a mono signal of the backing in both speakers and you switch the phase of one of the speaker.

You don't have to put the backing to loud. The singer will hear himself perfectly.

malice

[J.G.]

Cheers, Malice,

I too was wondering how this is achieved without leakage.

: J

[Brendo]

Oh yes, I agree 100%.

A while ago I had the singer in the control room, with an M88 handheld, pretending to be a kick drum.

[seagate]

Positioning.


Thank you!

:)

[otek]

Quote:
Originally Posted by otek
Oh yes, I agree 100%.

A while ago I had the singer in the control room, with an M88 handheld, pretending to be a kick drum


With an M88? Nah, he favors the Didrik DeGeer mics for that sort of thing.

[Bob Olhsson]

Leakage from a speaker makes terrific "glue." It just means you can't change the arrangement without changing the vocal.

A handheld M-88 is a serious secret weapon for vocals.

[J.G.]

It just means you can't change the arrangement without changing the vocal.

From a writer/singer's perspective, me LIIIIIIIIIIIIIIIIIIIKE that bonus! :Twisted:

[Mixerman]

Ah, leakage.

Well this is what separates the boys from the men in this business, and it separates those who don't mind their cock blowing in the wind a little bit.

If you're the kind of person that can't make concrete arrangement decisions as you go along, using speakers instead of headphones is not a viable option. Your arrangement HAS to be together and well-tested with scratch vocals before you do this. OR you HAVE to use only the parts you're absolutely sure are going to make the final record.

Now, on rock tunes, leakage is rarely a problem. For starters, the limiter tends to push out all but the vocal when the singer is singing. It's when the singer stops singing, and the limiter releases that the leakage comes up. One trick for this, is to use a slowish release on the limiter. This will give you some time between when the singer stops singing and when the leakage is allowed to creep back in. This way when you're comping, you can remove a high percentage of the leakage.

Malice posted about the traingulation of the singer and the speakers. I used to be very exact about getting the mic centered slightly behind the cones in the null zone between out-of-phase speakers. But these days, I don't even bother, because even when leakage is there, it's not really a problem. In many cases, headphone leakage can be far more annoying and problematic than speaker leakage.

To me, the positives of using speakers instead of headphones far outweigh the negatives. For starters, I have a direct and personal line of communication with those in the control room with me. Encouragement and discussion are far more effective when delivered face-to-face. Secondly, I can immediately play back a performance through the speakers in the control room, which is considerably more revealing to the singer than listening in headphones.

These days, seasoned and touring singers prefer their "in-ears." But that's because they have typically sung hundreds of shows with them. With these singers, I let them use their "in-ears," because it's their comfort zone. And some singers are more comfortable alone in the studio, with the producer away from them. You have to evaluate the situation, who you're dealing with, and make decisions based on these factors. And sometimes taking a singer out of his comfort-zone is a good thing, but if you do that too early, you'll just frustrate the singer. You almost have to wait for the singer to frustrate himself before you pull her out of the comfort-zone.

And one last peiece of advice, and this goes for both headphones and speakers. Avoid percussion in the singers mix unless it's an absolutely critiical part of the song an rhythmic bed. For instance, the shakers on "Sympathy for the Devil" would need to be in the mix because they are essential to the feel of the song. But most other percussion parts you should pull. I rarely record additional perrcussion parts before the song has been sung. The problem with additional percussion is the leakage can and will be insidious in nature. You can't get rid of it. And if you decide later that shaker part mucks up the chorus, you might be stuck with it because of the leakage on the singers track.

In some ways, leaving the additional percussion to last is counter to the idea of providing the singer with an inspiring track, because we all know that that tambourine in the chorus has provided lift, right? Possibly. And if the tambo part is going to be critical to the inspiration of the track then I'd say break the rule. But in general, percussion is a future headache best avoided.

Enjoy,

Mixerman

[Comte de St Germain]

Yeah, I tend to ditch all the "extras" too.

I have noticed that a good bass track blend is essential to hearing pitch and nailing it. The bass is the foundation, the guitars/keys the walls and the vocals the roof.

Also consonants typically coincide with the snare and either need to be spot on or swing like a monkey in a brothel.

[Bob Olhsson]

The bass can indeed make or break a vocal performance!

The first time I used speakers was recording the Rare Earth single "I Just Want to Celebrate." Pete, the lead singer, was sounding really frustrated so I hooked up my brand new Shure SM-53 that I'd been wanting to try out and just handed it to him the next time he came into the control room for a listen. Then I cranked up the Altec 604s, Pete put the mike up to his mouth and you can hear what happened on the record. Everybody got so excited that they wanted to do the backgrounds in the control room around the same mike. Then we stacked another pass!

The final result wasn't pristine sound but everybody felt it really captured the band's live feel and the record turned out to be their biggest hit.

I recently learned that Elvis Presley almost always recorded with a hand held mike and absolutely refused to wear headphones. According to Wally Heider, Sinatra recorded almost everything post-Capitol with a hand held 57.

[otek]

I recently learned that Elvis Presley almost always recorded with a hand held mike and absolutely refused to wear headphones. According to Wally Heider, Sinatra recorded almost everything post-Capitol with a hand held 57.

Bono is supposedly another headphone-hater.

[volthause]

If I may ask a dumb question:

How do you avoid feedback with the vocal mic in the control room with the mains cranked? I would have to assume that you have a fair amount of the performer monitoring back through the mains, right?

[Bob Olhsson]

I had a bunch of chamber verb on the vocal to help him hear relative pitch (don't try this with most digital verb) We were at the edge of feedback.

[HalfBlood]

Positioning.

You must form a perfect equilateral triangle between the two speakers and the mike. both speakers must be at the exact same distance from the capsule.

Then you feed a mono signal of the backing in both speakers and you switch the phase of one of the speaker.

You don't have to put the backing to loud. The singer will hear himself perfectly.

malice


Was just reading through some older threads and thsi sruck me as a VERY valuable piece of info.

I'm wondering... If you recorded the signal from one speaker in mono and then with the speaker in the same position recorded the vocal takes and THEN flipped the phase inside the DAW combining the pre-recorded track of the music with the vocal takes, then phase cancellation would be close to perfect... am I wrong?

Or is there an issue concerning computer lagging?

Tomorrow I'm trying that in any case.

[HalfBlood]

It works almost perfectly!

I placed a mono mixdown in the DAW and recorded it again through the mike, at pretty high volume too.

Then without moving the mike, re-recorded at the same volume whilst singing veeery low volume.

Then I played them back together and flipped the phase on 1 and voilà!

Even with the mike just next to the speaker, speaker volume high and singing softly, the only thing that bled through a bit was the drums.. I guess it's difficult for them to cancel out completely being fast transients. Also some very high freq FX bleed through.

But the bass and the mids were almost completely gone, and I'm thinking that when I use this trick for a singer who has a BIG problem with headphones, I can cut out the drums and hi-fx and get him to sing.

This is so simple, I don't why I never thought of it.

Also I HATE singing with headphones, although I've always done it that way. For me this is like being told I'm free at last.

WHOOHOO!

[otek]

We always use speakers for cues when recording choirs.

Two passes with the choir makes for a bigger sound, and when flipping the polarity on the second take, the music cancels out to a very high degree.

As you noticed yourself, you can knock some high end off of the cues and send less overheads and other high frequency instruments.

If all you need is one take of vocals, you don't need to sing on the second pass - simply play the song back once more.

What Malice suggested is faster and more efficient for lead vocals, however. With choirs it doesn't work because of the multiple mics (in my case, mostly three or four).


otek

[HalfBlood]

Hey, I feel like it's Christmas..:grin: Thanks for that thread guys! This is pretty important for me as I find it kills my performance when I sing at home. Firstly because I know anyone hearing will just hear my voice :Uh oh: and also I need to move to get into the groove.

I'd always dismissed the idea, and yet I could have arrived and thought of it. That is the negative side of things, but who cares.




As you noticed yourself, you can knock some high end off of the cues and send less overheads and other high frequency instruments.

This has to do with hi freqs being faster and therefore less easy to coincide and cancel out right?


What Malice suggested is faster and more efficient for lead vocals, however. With choirs it doesn't work because of the multiple mics (in my case, mostly three or four).



Why is it faster and more efficient? I would have to take time to position the speakers whereas with this method I don't. is this because computer innacuracies which can occur between the two takes are made irrelevant as the counterfase is simultaneous to recording?

[HalfBlood]

If all you need is one take of vocals, you don't need to sing on the second pass - simply play the song back once more.



hang on, I just noticed this phrase.. what do you mean?

Say I wanted 2 vocals. I would need a playback only track in any case for phase cancellation, otherwise I would be stuck with the vocals at a fixed volume, and all piled on the final vox track. Or did you mean something else?

[otek]

This has to do with hi freqs being faster and therefore less easy to coincide and cancel out right?

It has to do with them being smaller wave forms, yes. A low frequency wave form has more "tolerance" due to its size.

Why is it faster and more efficient? I would have to take time to position the speakers whereas with this method I don't.

If you don't have your own place where the installation could be more or less permanent, then yes.

You have to factor in the amount of time needed to make the "dummy runs" of each vocal take. If you are cutting, say, an album's worth of Beach Boys covers :D , you will recoup the extra time it takes to set up the speakers in no time.


otek

[meLoCo_go]

It has to do with them being smaller wave forms, yes. A low frequency wave form has more "tolerance" due to its size.


Exactly. Imagine that your face moves closer to the mic by say 3 cm - so if you had a wave with wavelength of 6 cm (~5000 Hz) reflecting from your face it will be out of phase (phase face :lol: ) at the pick up point. But for 1 m wavelength it is almost no difference. Smaller waves allow more precise measurement of distance, there's a good reason bats are using ultrasound and not infrasound to navigate.