TranceAddict Forums (www.tranceaddict.com/forums)
- Production Studio
-- vocal keying method?
Pages (2): [1] 2 »

vocal keying method?

Hi, I am wondering how do some of the producers do that effect where they take a vocal sample, and separate it into different parts based on what key or note is being sung, then construct a melody by rearranging these separated parts of the sample?
The way I think it is done is they load the vocal sample into something which scans it , then splits it wherever a different key is detected, then these individual sections are used like an instrument... am I right?

Well, you pretty much answered your question!

Most DAW programs allow you to manually cut the wave up into smaller parts and then copy/paste manipulate/rearrange them as you see fit.

What program are you running?

currently im just using a sound editor, and its painful and claustrophobic to manually go through a sample for different keys... the end result is fun but i am looking for an easier way than copying, pasting, adjusting the selection marker to copy again, etc... isnt there like a plugin for a program somewhere which scans a sample and tells you which keys are being played when?

Like, how about I have multiple sections selected in my vocal sample, and I assign each selection to a different key on my hypothetical MIDI keyboard (I dont own a keyboard)...Then as I record myself playing through the melody one key at a time I can alter the position of the selection markers for each time the key is being played i.e. the positions of the start and end markers for the selection is a variable of the keystroke(I dont know the proper term for it, its like how hard you press on a piano), then I time everything properly on a sequencer... Does what I just said sound like an excerpt from a sci fi novel or is there actually something out there which does just that?

Re: vocal keying method?

quote:

Originally posted by carmatic The way I think it is done is they load the vocal sample into something which scans it , then splits it wherever a different key is detected

Erm, why would you need something to do this for you, surely unless you're tone deaf you can hear when it changes note (not key, the singer will be singing in one key, its unlikely there will be a key-change/modulation in the middle of some vocals), and cut it manually. Most people would probably just chop it manually then rearrange the parts by ear to something that sounds good rather than worrying about what actual note each part is...

Now if you ARE actually completely tone-deaf, something like Melodyne should be able to detect the notes that are sung for you...

Re: Re: vocal keying method?

quote:

Originally posted by Sanguis Mortuum and cut it manually.

I'm not that tone deaf but it takes me many many many tries to get a sequence to sound in-tune, especially if the vocal was sung in an 'expressive' way or something... I can stick with whatever I've managed to get from the sample and then pitch correct it, but I'd prefer to get the correct source material in the first place... Does Melodyne highlight the sections which are on-key within a certain tolerance or something, cuz that'll be kind of what I'm looking for , to get a series of tight-sounding vocal samples which don't sound like an instrument which needs tuning because the notes are off, and without the chipmunking or golem-like effect if I had to correct the pitch...

Stuiped Q :
Can you do this with Atotune or Melodyne ?

I don't know much about autotune, but i think so. I do know melodyne and waves tune can.

quote:

Originally posted by Khayat Stuiped Q : Can you do this with Atotune or Melodyne ?

the thing is, parts of the sample I have is sung on-key, I just need to detect where these parts begin and where they end... I dont want to need to mess with changing their pitch etc... Is buying Autotune or Melodyne my only option?

Im reading now about both I really find Autotune kool but bcuz I never any of them I have to read other peoples comments and many agree that Melodyne 3 kicks Autotune

Melodyne will definately do what you want, it will still be a lot cheaper to do it by ear though...(unless of course you acquire Melodyne for free, which of course I would never condone )

doesnt the essential come free

quote:

Originally posted by Sanguis Mortuum Melodyne will definately do what you want, it will still be a lot cheaper to do it by ear though...(unless of course you acquire Melodyne for free, which of course I would never condone )

Well, it will be better if I actually get paid for doing it by ear... Cuz that's what it feels like...

What is this 'essential' thing? do you have a link to it

http://www.celemony.com/cms/index.p...ducts_essential

GSnap, which is free is more than enough for this purpose. Don't waste money on Autotune or Melodyne.

quote:

Originally posted by Eldritch GSnap, which is free is more than enough for this purpose. Don't waste money on Autotune or Melodyne.

Gsnap looks cool from what I've seen of it, but is it a vocal pitch corrector , or can it also detect correct vocal pitches?and preferably once it has detected the parts where the notes are on-key, it separates those parts for me somehow so I dont have to load the vocal sample in another program and find the sections that Gsnap says are on-key... I'm abit of a purist, I dont want to do anything other than sample the vocals directly... otherwise I might as well be re-keying the entire vocal track or something

sorry if Im asking abit much! I thought it would be simple to do something like this...

quote:

Originally posted by carmatic preferably once it has detected the parts where the notes are on-key, it separates those parts for me somehow so I dont have to load the vocal sample in another program and find the sections that Gsnap says are on-key...

So you just want a program where you can click a button and have it all done for you?

You might call that 'being a purist', I would call that 'being lazy'.

I'm a little unclear about what your after, but I assume yor talking about the stutter edits like in song like "Year Zero" by andy moor.
If this is the case, then neither autotune or melodyne will do what you want, but you might need them anyway.
What you probably want to do, is split the sample up into short segments, then stick it in a sampler like kontakt, and map each key (which you have to work out yourself, because I've used both melodyne and auto tune and neither have anything to do with it) to a certain slice of the sample.
where the autotune might come in handy, is in cleaning up the sample, because you want it to be as close to pitch as possible to get a good effect, this only applies if you recorded it yourself.

yes, let me put it this way

I want to load the vocal sample into something , which when I press a key for a note, will play the section of the sample which is sung at that note

quote:

Originally posted by carmatic yes, let me put it this way I want to load the vocal sample into something , which when I press a key for a note, will play the section of the sample which is sung at that note

Yeah ok, the issue is, theres no program that will detect the original input note, then warp it to the note your playing. There are programs which will detect the input pitch and make a midi fromit (melodyne) or others which can play a single sample all over the keyboard (a sampler) but no program will do both. So your going t have to work out the notes that are being sung yourself, in order to tell the sampler what the original note is.
You can do it two ways:

1/ load a series of samples of different notes and then play them. You still need to work out what notes were originally being sung yourself.
2/ load a single sample into the sampler, tell the sampler what the original key was and then allow the sampler to pitch shift the single sample all over the keyboard. This will result in poor quality.

This is hard to explain without showing you, but I hope its a bit clearer.

The way I do it is in the editor in live, this gives you control over placement and pitch using the transpose knob. I don't like programming it all into samplers much, but maybe give it a try if the editor method is irritating you.

Melodyne is designed for this sort of thing (it doesn't actually rearrange anything, unless you tell it to, it just re-pitches what's already there, and does a damn good job of it).

There are other pitch-correction plugins/programs like Autotune, but, not surprisingly, they just correct intonation problems. They're not so great for actually changing the entire rhythm and melody.

Kit: Melodyne literally DOES stick it onto a piano roll for you and let you change the notes, just like you'd change MIDI notes in a sequencer. It's a little scary how little effort is required.

quote:

Originally posted by DigiNut Melodyne is designed for this sort of thing (it doesn't actually rearrange anything, unless you tell it to, it just re-pitches what's already there, and does a damn good job of it).

so you CAN make melodyne rearrange a sample to get a new melody? i.e. it finds the different sections containing different notes being sung then maps them to your keyboard

quote:

It's a little scary how little effort is required.

well, thats my style!

quote:

Originally posted by carmatic it finds the different sections containing different notes being sung then maps them to your keyboard

i think not. it's a pitch correction program. but it can show you when and what note is sung and you can cut it out and put it in your sampler mapping it to the corresponding key. you can do it also in your audio editor select a range which you think is one key. loop it, use your ears or spectrum analyser to determine the key. export that put a name on it with a key in a filename. do this for all parts and import it to a sampler and map to a corresponding key.

but thats really bullshit you're after. this is not how it's done. you just import the whole vocal to a sampler. map it to your keyboard so that each key you press and hold plays a sample to the end of file or as long as you hold the key down. but different keys trigger the sample at different places. and you just jam until you find something cool.

If you just want to "rearrange" the notes, then import it into a sampler. Melodyne is overkill for something this simple, and it's a hell of a lot more expensive.

I have to reiterate for the people who weren't listening, however, that Melodyne is not simple pitch correction like Autotune. I don't know how to make this any clearer than I already have, so here's a screencap from their site:





Make sense? You can import an audio track and change the pitch and duration of any notes, through a piano roll UI similar to the MIDI/Piano-Roll UI in every other sequencer. Pitch correction only fixes minor tuning problems from an instrumentalist or vocalist.

Sorry, maybe I misinterpreted, I know that melodyne can rearrange the melody, but I thought he wanted to be able to play the melody on a keyboard to get stutter edits or something similar. I probably misunderstood, carmatic, can you post an example of what you want to do please, I'm really lost now and I'd hate to see you spend money on autotune or melodyne when a sampler was what you needed...

A a side note, autotune can be put in to manual mode to re arrange notes, but melodyne is better designed for this I find I can't get more than 2 semitones without some really serious artifacts creeping in too, melodyne or autotune...

i dont know how long youve been making music but if you cant take an audio sample and cut it up to match the pitches of a melody in a song, then you need to learn how to use your sequencer as well as train your ear. your trying to take an easy way out and in doing so your making things 100 times harder. you dont need a special program to do what you want. just cut and move the audio sample. the best way to train your ear is to go to a piano, synth, whatever and then play a random note. humm that note or sing it out loud. keep doing this until its easy. then listen to a song and hum to the key of the song. you should then be able to hear where the certain vocal phrases will match with the melody. the only way to do this right is to train your ear. othewise you will never be able to make good tonal music.

Powered by: vBulletin
Copyright © 2000-2021, Jelsoft Enterprises Ltd.