Where to get good vocals? Also, use them in key?

[Eat Bass]

Ok, so I'm going to adress my concerns in numbers for the sake of being easy to respond. Any and all replies are appreciated

1. I've been listening to a lot of Burial lately and I noticed he uses a ton of vocals. It seems hes using mostly R&B and Gospel women singers? It's hard to tell but they're really long and crying out, whale like, drenched in reverb it really adds a ton of emotion. Where can I find some vocals like these? I tried searching 90's R&B acapellas but none were as emotional and long and crying out like a whale as Burial's.

2. Furthermore, when making a tune with such vocals, does Burial match his melodies, chords, pads, whatever to the key of the vocals? Also, does he time stretch them without effecting the pitch to ge them in the same tempo as the new track? I know he mentioned he doesn't have much musical knowledge and just does what sounds good, so I'm thinking not. But at the same time his songs sound so cohesive. I can't tell if thats due to everything being in key, him just having a good ear, his ability to arrange, or the way he creates such a smooth mix with nothing too upfront, everything kind of wishy washy with maybe one element taking the light but most just drenched in reverb.

3. So if you do think its a must to create the tune in key with the vocals, how do you go about finding what key the vocals are in? I know Logic has a tuner that will tell you the notes of things, but it doesn't do so well with a lot of sounds, only very basic ones. Also, Burial does A LOT of pitching to his vocals. Do you think he pitches in octaves to keep everything in key or is he just pitching to taste? I tend to let these concerns act as barriers and it affects my creativity if I don't know how to keep everything in key but want to, or if I don't need to keep everything in key but can't help but want it to be.

[+3]

Did you ever find that Frequency-based Mono-making VST?
Cause this is the closest I've come...

http://www.brainworx-music.de/en/plugins/bx_control_v2

[outdropt]

Could you reference a tune? I started buying vocal packs recently.. Loopmasters has some quality. I realized quickly that unmixed accapellas (for the most part) sound shit, that is accapellas ripped from a song with phase cancellation, ect .

Unless you drowned in reverb and effects. Original accapellas are always the best route.

For the in key question, i think the only way a vocal will fit, is if it is in key with the song... It will be fairly obviously that the vocal doesnt sound right against the melody ... You can always try to pitch the vocals a few semitones up or down against a saw tooth synth playing, vise versa, to make it obvious when its on key

I use melodyne to find out what key the vocals are in, and it even breaks down the melody for you... Melodyne is on point for the most part.

[Eat Bass]

Did you ever find that Frequency-based Mono-making VST?
Cause this is the closest I've come...

http://www.brainworx-music.de/en/plugins/bx_control_v2

i did find one from DDMF, but its 30 bucks. I just found a better method of frequency splitting my basses (i use fabfilter saturn with everything disabled and then i use an eq to catch anything else) and only adding subtle stereo effects to the midrange, and a lot to the high end. Depone told me adding stereo effects to the whole bass and then slapping a "mono'er" wont sound good. better off splitting the bass and affecting it accordingly, so thats what i've been doing. I just do the splitting in separate project template then load the final bass in kontakt or esx24 in the project.

Could you reference a tune? I started buying vocal packs recently.. Loopmasters has some quality. I realized quickly that unmixed accapellas (for the most part) sound shit, that is accapellas ripped from a song with phase cancellation, ect .

Unless you drowned in reverb and effects. Original accapellas are always the best route.

For the in key question, i think the only way a vocal will fit, is if it is in key with the song... It will be fairly obviously that the vocal doesnt sound right against the melody ... You can always try to pitch the vocals a few semitones up or down against a saw tooth synth playing, vise versa, to make it obvious when its on key

I use melodyne to find out what key the vocals are in, and it even breaks down the melody for you... Melodyne is on point for the most part.

is there any other way to identify what key the vocal is in? also would the vocal need to be time stretched without effecting the pitch to the tempo of the new song?

heres some examples...

http://www.youtube.com/watch?v=QS38JBh5gcw (starting at 1min)

http://www.youtube.com/watch?v=OtEBEVC0HUc

http://www.youtube.com/watch?v=LHtNPzaHO7k (starting at about 1:20)

[Sinfected]

1. Try to find something YOU like and can make fit, dont just disregard stuff cause they dont sound like burial. Finding good vocals is a real pain in the ass tho, and I would suggest finding someone who can sing and get your own recordings if you want quality stuff. But the main reason burials vocals sound the way they do is not just the vocals, but the arrangement and soundscape of the song as a whole.

2. If you have an ear for music it shouldnt be hard to find what key a vocal fits, just open a piano or something and play some notes until you find what keys sound right, and theres alot of flexibility there since vocals dont have to monotonically follow the root note of the track. About the timestretching you can just compensate it with the pitch, alot of samplers can do this automatically, kontakt can I'm sure. I think even FL's standard sampler does this to some extent, might be wrong there tho. About fitting vocals to a different tempo the easiest way imo is just trial and error, with the knowledge that in some situations it just wont work.

3. As I said, and I know its a boring answer, AND I know its been repeated alot but if you are musical you should really hear if the vocals dont fit a key, in the same way you would hear (I hope) if someone playing a guitar solo picked a note that is really out of the scale or if you tune your sub to F and play a wobble in F#. Bottom line if it sounds like it fits, it probably does. I think burial cut up vocals alot and pitch them individually to create a melody (archangel is a pretty obvious example of this).

Probably not the answer you are looking for but I dont think theres any rules to follow. In alot of non-western music they use scales that would sound "off" to our western ears, but I guess they would say the same about our counterparts.

[Eat Bass]

1. Try to find something YOU like and can make fit, dont just disregard stuff cause they dont sound like burial. Finding good vocals is a real pain in the ass tho, and I would suggest finding someone who can sing and get your own recordings if you want quality stuff. But the main reason burials vocals sound the way they do is not just the vocals, but the arrangement and soundscape of the song as a whole.

2. If you have an ear for music it shouldnt be hard to find what key a vocal fits, just open a piano or something and play some notes until you find what keys sound right, and theres alot of flexibility there since vocals dont have to monotonically follow the root note of the track. About the timestretching you can just compensate it with the pitch, alot of samplers can do this automatically, kontakt can I'm sure. I think even FL's standard sampler does this to some extent, might be wrong there tho. About fitting vocals to a different tempo the easiest way imo is just trial and error, with the knowledge that in some situations it just wont work.

3. As I said, and I know its a boring answer, AND I know its been repeated alot but if you are musical you should really hear if the vocals dont fit a key, in the same way you would hear (I hope) if someone playing a guitar solo picked a note that is really out of the scale or if you tune your sub to F and play a wobble in F#. Bottom line if it sounds like it fits, it probably does. I think burial cut up vocals alot and pitch them individually to create a melody (archangel is a pretty obvious example of this).

Probably not the answer you are looking for but I dont think theres any rules to follow. In alot of non-western music they use scales that would sound "off" to our western ears, but I guess they would say the same about our counterparts.

I mean i can tell if something sounds sort of in key with the note being played but definitely not spot on. My ears aren't trained well enough. I need to figure out someway to figure out the key of the vocals. But your kind of saying that not all vocals have to be in key. Like sometimes in Burial's tunes the vocals are the main part besides maybe some background pads playing some chords. As long as maybe one of the notes in the chord of the pad is the same as the note of the vocal it would sound good, no?

Also, more on the timestretching/temp, I know logic can do this in its sample editor if you know the tempo of the original acapella (which i find how????). But if I chop up all my own vocals wouldn't I essentially be resetting the tempo anyway, since I can just chop up the parts I want and place them in time with the drums or something? Then again Burial doesn't do anything in time. FUCK LOL

These rules/no rules are fucking with me. I need some sort of guideline until I can become fluent with this.

[Hircine]

Good sources for vocals are rnb, ragga, reggae, basically any acapella you can get your hands on. To know what note the singer is on, I just stretch the sample and run it through a tuner. If it still sounds out of key, I go by ear pitching up and down until it sounds good. Do a lot of resampling. Drowning the vocal in reverb and then resampling is a good way to avoid high end artifacts when timestretching. Don't use only one acapella, chop phrases, long notes, syllables and arrange them together, lots of reverb and delays to get everything in the same space.

[Laika]

Yo Eat Bass, Logic has a plethora of professionally recorded vocal loops that are labeled with their respective key. Stretch/pitch them in flex mode, chop them up, and play them using the EXS24. Burial sounds like he chops up single syllables and stretches/processes them to fit his song.

[Eat Bass]

Good sources for vocals are rnb, ragga, reggae, basically any acapella you can get your hands on. To know what note the singer is on, I just stretch the sample and run it through a tuner. If it still sounds out of key, I go by ear pitching up and down until it sounds good. Do a lot of resampling. Drowning the vocal in reverb and then resampling is a good way to avoid high end artifacts when timestretching. Don't use only one acapella, chop phrases, long notes, syllables and arrange them together, lots of reverb and delays to get everything in the same space.

ok thanks, that helps a bit. ill try that out. im trying to get myself to produce some deeper shit. all i make currently is some buzzy annoying shit.

Yo Eat Bass, Logic has a plethora of professionally recorded vocal loops that are labeled with their respective key. Stretch/pitch them in flex mode, chop them up, and play them using the EXS24. Burial sounds like he chops up single syllables and stretches/processes them to fit his song.

oh really? ill have to take a poke around then. i was unaware of this. thanks.

[ehbes]

[quote="Eat Bass" all i make currently is some buzzy annoying shit.[/quote]
try building up pads that individually are light and air-y into something larger.

[Eat Bass]

[quote="Eat Bass" all i make currently is some buzzy annoying shit.

try building up pads that individually are light and air-y into something larger. [/quote]

ok thanks man. yeah you got the deeper more ambient stuff down. i want to get there.

[ehbes]

In maelstrom take a look at the waveforms ambient chords 1 and 2

[Eat Bass]

In maelstrom take a look at the waveforms ambient chords 1 and 2

will do. i wish i could use reason just like a vst in logic though. sometimes its a bit of a pain in the arse routing the rewire then going to the manual to find out which # in logic ='s what parameter in reason so that i can automate. i suppose i could just sample from reason tho and slap it in kontakt. matter a fact i think ill do that.

[Sinfected]

I mean i can tell if something sounds sort of in key with the note being played but definitely not spot on. My ears aren't trained well enough. I need to figure out someway to figure out the key of the vocals. But your kind of saying that not all vocals have to be in key. Like sometimes in Burial's tunes the vocals are the main part besides maybe some background pads playing some chords. As long as maybe one of the notes in the chord of the pad is the same as the note of the vocal it would sound good, no?

Also, more on the timestretching/temp, I know logic can do this in its sample editor if you know the tempo of the original acapella (which i find how????). But if I chop up all my own vocals wouldn't I essentially be resetting the tempo anyway, since I can just chop up the parts I want and place them in time with the drums or something? Then again Burial doesn't do anything in time. FUCK LOL

These rules/no rules are fucking with me. I need some sort of guideline until I can become fluent with this.

I'm having a hard time trying to explain this since i've never put any words on it, but imagine a pop song or whatever having a verse that runs in two chords. The vocals will be all around the place tonally, but in scale, they wont just sing the root note of whatever chords are played. Take this for example:

C----------------------G
Happy birthday to you
-----------------------C
Happy birthday to you
-------------------------------G F
Happy birthday to dear Eat Bass
F--------C---------G---C
Happy birthday to you

You wouldnt sing "Happy birthday to" in C and "you" in G and so on, it plays around in the scale in a way that melodically fits the chords. Therefore you usually cant find the right key for a vocal by running them through a tuner since its not the root note of the chord and changes all the time.

I guess a way around it would be to chop the individual phrases and run them through a tuner and just treat it like a lead stab, but that would be somewhat redundant.

Finding the tempo for a ripped acapella should be easy enough if you just take the original song and run it through a tempo detector, and I'm sure you can figure out mathematically how many percents you need to timestretch and stuff to get it in another specific tempo, but again, its really redundant. You can still cut the acapella up into phrases and move them however you want even after timestretching, the timestretching will still change the speed in which the phrases are sung, so theres infinite room for experimentation aswell.

And as with everything else concerning a subjective "no rules" kind of thing like music, experience is always the best guideline. You should get some random acapella and just mess around with, hiphop acapella for example since tonally they are as fun as the day is black, that you know is 90bpm for arguments sake, and making it fit 100bpm to begin with. Once you got the tempo down you can experiment with some chords/pitch of the acapella etc. You dont have to make a full blown project out of it but its really good practice just to get grips with what you can and cant do with vocal samples. Old remix packs with vocals are a good source to practice with aswell. I probably have alot of old stuff I could dig up that you could use for practice, not alot worth using for a real project tho.

[zerbaman]

datpiff.com has plenty of acapellas.

[Eat Bass]

I mean i can tell if something sounds sort of in key with the note being played but definitely not spot on. My ears aren't trained well enough. I need to figure out someway to figure out the key of the vocals. But your kind of saying that not all vocals have to be in key. Like sometimes in Burial's tunes the vocals are the main part besides maybe some background pads playing some chords. As long as maybe one of the notes in the chord of the pad is the same as the note of the vocal it would sound good, no?

Also, more on the timestretching/temp, I know logic can do this in its sample editor if you know the tempo of the original acapella (which i find how????). But if I chop up all my own vocals wouldn't I essentially be resetting the tempo anyway, since I can just chop up the parts I want and place them in time with the drums or something? Then again Burial doesn't do anything in time. FUCK LOL

These rules/no rules are fucking with me. I need some sort of guideline until I can become fluent with this.

I'm having a hard time trying to explain this since i've never put any words on it, but imagine a pop song or whatever having a verse that runs in two chords. The vocals will be all around the place tonally, but in scale, they wont just sing the root note of whatever chords are played. Take this for example:

C----------------------G
Happy birthday to you
-----------------------C
Happy birthday to you
-------------------------------G F
Happy birthday to dear Eat Bass
F--------C---------G---C
Happy birthday to you

You wouldnt sing "Happy birthday to" in C and "you" in G and so on, it plays around in the scale in a way that melodically fits the chords. Therefore you usually cant find the right key for a vocal by running them through a tuner since its not the root note of the chord and changes all the time.

I guess a way around it would be to chop the individual phrases and run them through a tuner and just treat it like a lead stab, but that would be somewhat redundant.

Finding the tempo for a ripped acapella should be easy enough if you just take the original song and run it through a tempo detector, and I'm sure you can figure out mathematically how many percents you need to timestretch and stuff to get it in another specific tempo, but again, its really redundant. You can still cut the acapella up into phrases and move them however you want even after timestretching, the timestretching will still change the speed in which the phrases are sung, so theres infinite room for experimentation aswell.

And as with everything else concerning a subjective "no rules" kind of thing like music, experience is always the best guideline. You should get some random acapella and just mess around with, hiphop acapella for example since tonally they are as fun as the day is black, that you know is 90bpm for arguments sake, and making it fit 100bpm to begin with. Once you got the tempo down you can experiment with some chords/pitch of the acapella etc. You dont have to make a full blown project out of it but its really good practice just to get grips with what you can and cant do with vocal samples. Old remix packs with vocals are a good source to practice with aswell. I probably have alot of old stuff I could dig up that you could use for practice, not alot worth using for a real project tho.

ok i get what your saying now. that makes sense. so really just make sure the vocals are in the same scale as the song and dont have any afflicting notes playing simultaneously. in essence, try to stay in scale and do what sounds right. fair enough...

i actually probably will chop one phrase and tune it and treat it like a stab, only on certain parts though. i think this would be cool.

now for the tempo thing. is it fully necessary to change the tempo of the vocals when chopping up bits and pieces like that? i know logic can change the tempo of anything without changing the pitch and it does it pretty fluently with no artifacts as long as its within a reasonable tempo. again, im just not sure how necessary this is? i will mostly be taking a few words of a vocal and looping it and affecting it, then i will also be chopping individual words and even vowels of words and arranging them the bast i can to try and convey some emotion and melody. im just not sure i see the point of matching the tempo when i will be chopping and and placing each piece where i want it. sorry i know this question is getting redundant. i just want to have a definite plan of action if pure experimentation isn't yielding the best results.

[outdropt]

Wasn't able to get on after work last night so i am a little late on this...
So that first tune....

He no doubt has multiple audio tracks going on, it sounds very complicated....

The background ambiance vocals are most likely the original vocals and cut up syllables made into a melody drenched in reverb,. Like 100% wet with a medium tail + compressor, bounced down and time stretched. Probably some filter movement too.

Around 1min... sounds like syllables pitched around to make a melody .... and he uses this pretty consistently

Also sounds like he is using a send with heavy reverb and adding to it at the very end of a vocal clip to carry it into the next section of vocals... Like at 4.42...

Some sections he adds distortion to the vocals to harsh up the top end but very lightly and not very often.


For your questions in regards to Time stretching.... There are certain settings in most DAWs to change the way the time stretching effects the vocals. I am going to have to use what i know (ableton) as an example.

There are different settings that preserve different information (pitch, tone, timbre, formants, ect) when you time stretch.. It depends on whats important for that clip... For beats you dont need to preserve the pitch to much, while for vocals you do... So i would check for the setting in your DAW that is ment for vocals...

In ableton i use Complex...

"Complex preserves the formants (spectral peaks, filtering, essential... this allows you to pitch up a vocal and not have it sound like a chipmunk)."

I would try to find a setting like this in your DAW

Edit~~~ just read your last question.... When your chopping like that forget about the tempo of your track, at this point double/half the tempo just to get some ideas but dont worry about it not working with your song, in regards to the vocals because your most likely going to be repositioning them on a grid that is in tempo... I mean if its a 4 bar loop it would probably sound best if a word begins on every or every other bar.. But it depends on the feel of your track and the melody behind it.

[Eat Bass]

Wasn't able to get on after work last night so i am a little late on this...
So that first tune....

He no doubt has multiple audio tracks going on, it sounds very complicated....

The background ambiance vocals are most likely the original vocals and cut up syllables made into a melody drenched in reverb,. Like 100% wet with a medium tail + compressor, bounced down and time stretched. Probably some filter movement too.

Around 1min... sounds like syllables pitched around to make a melody .... and he uses this pretty consistently

Also sounds like he is using a send with heavy reverb and adding to it at the very end of a vocal clip to carry it into the next section of vocals... Like at 4.42...

Some sections he adds distortion to the vocals to harsh up the top end but very lightly and not very often.


For your questions in regards to Time stretching.... There are certain settings in most DAWs to change the way the time stretching effects the vocals. I am going to have to use what i know (ableton) as an example.

There are different settings that preserve different information (pitch, tone, timbre, formants, ect) when you time stretch.. It depends on whats important for that clip... For beats you dont need to preserve the pitch to much, while for vocals you do... So i would check for the setting in your DAW that is ment for vocals...

In ableton i use Complex...

"Complex preserves the formants (spectral peaks, filtering, essential... this allows you to pitch up a vocal and not have it sound like a chipmunk)."

I would try to find a setting like this in your DAW

Edit~~~ just read your last question.... When your chopping like that forget about the tempo of your track, at this point double/half the tempo just to get some ideas but dont worry about it not working with your song, in regards to the vocals because your most likely going to be repositioning them on a grid that is in tempo... I mean if its a 4 bar loop it would probably sound best if a word begins on every or every other bar.. But it depends on the feel of your track and the melody behind it.

yeah logic has plenty of options to preserve the vocals when time stretching. im going to have to watch another video on it though as i havent used it in a while and im still fairly new to logic.

thanks for the responses though guys this is making me a bit more comfortable and at least giving me some ideas to work with.

[Sinfected]

now for the tempo thing. is it fully necessary to change the tempo of the vocals when chopping up bits and pieces like that? i know logic can change the tempo of anything without changing the pitch and it does it pretty fluently with no artifacts as long as its within a reasonable tempo. again, im just not sure how necessary this is? i will mostly be taking a few words of a vocal and looping it and affecting it, then i will also be chopping individual words and even vowels of words and arranging them the bast i can to try and convey some emotion and melody. im just not sure i see the point of matching the tempo when i will be chopping and and placing each piece where i want it. sorry i know this question is getting redundant. i just want to have a definite plan of action if pure experimentation isn't yielding the best results.

Its not getting redundant, theres no such thing as a bad question.

Its a necessary as you want it to be. In the end vocals are just another soundwave like any other to the DAW, so you can do whatever you want with it. If you want to keep the vocal as it is, without any chopping its important to get it in the right tempo for obvious reasons. But if you are going to chop up vocals you can resample individual phrases and articulations, then take them even further, and you can timestretch/pitch wherever you want in the process. This opens up alot of possibilitys since you can treat the vocals any way you want. I think alot of people get hung up on resampling and think its some magic trick for fat basses, but its mainly just a workflow enhancement for everything. Once you have one phrase cut up you can save 5 versions in different pitches, and 5 more timestretched different, morph them together with a vocoder, layer it, distort it, reverse it, whatever, you get the idea. If you have kontakt you can set up an instrument with it too so you have tons of different versions on different midi notes in the same instruments to make things even easier.

Its a personal judgement you have to make specific to the vocal that you are going to use. Just dont get too hung up on whats "right" and treat it like you would any other audio, with the golden rule: when it sounds good, its good.

[Eat Bass]

now for the tempo thing. is it fully necessary to change the tempo of the vocals when chopping up bits and pieces like that? i know logic can change the tempo of anything without changing the pitch and it does it pretty fluently with no artifacts as long as its within a reasonable tempo. again, im just not sure how necessary this is? i will mostly be taking a few words of a vocal and looping it and affecting it, then i will also be chopping individual words and even vowels of words and arranging them the bast i can to try and convey some emotion and melody. im just not sure i see the point of matching the tempo when i will be chopping and and placing each piece where i want it. sorry i know this question is getting redundant. i just want to have a definite plan of action if pure experimentation isn't yielding the best results.

Its not getting redundant, theres no such thing as a bad question.

Its a necessary as you want it to be. In the end vocals are just another soundwave like any other to the DAW, so you can do whatever you want with it. If you want to keep the vocal as it is, without any chopping its important to get it in the right tempo for obvious reasons. But if you are going to chop up vocals you can resample individual phrases and articulations, then take them even further, and you can timestretch/pitch wherever you want in the process. This opens up alot of possibilitys since you can treat the vocals any way you want. I think alot of people get hung up on resampling and think its some magic trick for fat basses, but its mainly just a workflow enhancement for everything. Once you have one phrase cut up you can save 5 versions in different pitches, and 5 more timestretched different, morph them together with a vocoder, layer it, distort it, reverse it, whatever, you get the idea. If you have kontakt you can set up an instrument with it too so you have tons of different versions on different midi notes in the same instruments to make things even easier.

Its a personal judgement you have to make specific to the vocal that you are going to use. Just dont get too hung up on whats "right" and treat it like you would any other audio, with the golden rule: when it sounds good, its good.

much appreciated bro. i do have kontakt 4. i bought komplete 7 for 2 hundred with razor!!! couldnt pass that deal. but yeah i dont know kontakt at all so im going to have to learn to do what you said. any links to tutorials on using kontakt in the method you described? thanks again, cant say it enough.