Vocals - beating it to fit (and painting it to match)

Post Reply
User avatar
deusdiabolus
Posts: 36
Joined: Thu Dec 08, 2011 6:29 am
Location: Wichita. No, seriously.

Vocals - beating it to fit (and painting it to match)

Post by deusdiabolus »

I thought I would pick your brains regarding how you all work with pre-recorded a cappella vocals. Specifically...say you have an a cappella vocal that you want to work into a track that is considerably faster or slower, but you want to keep the vocal at the same pitch. What I have been doing is using a tracker machine (usually Unwieldy) to trigger the vocal syllable by syllable the way it was originally sung, using the offset command to find each word. Needless to say, this is a very time-consuming process and sometimes I still get a result that sounds artificial. I was curious as to whether there might be another method that works better, something closer to "ACIDizing" the WAV so that it auto adjusts to fit. Hey, you never know if you don't ask, right?
User avatar
Evelon
Posts: 147
Joined: Tue Nov 22, 2011 11:00 pm

Re: Vocals - beating it to fit (and painting it to match)

Post by Evelon »

There is "stretch to x ticks" column in uTracker.
Then adjust tempo to get to original pitch (if the loop is chopped/cut correctly, i.e snapped to grid).
mute
Posts: 417
Joined: Mon Nov 21, 2011 8:30 pm

Re: Vocals - beating it to fit (and painting it to match)

Post by mute »

Depends..
I chop vocals up in sections.. if the tempo is already correct, then place where needed.

I sometimes use offsets, but not usually. If im using offsets (or note delays) its usually for minor corrections.

When I need more precision, I'll render either the song or a beat/loop to wav then take it into a multitrack editor and then cut/adjust the vocals and export them.

If the vocals don't match tempo, I do the same as above but use stretching and then cut/export

When recording live vocals, I load my wav into my multitrack and then record the vocals there.. cut em up, then bring em back to buzz.

If I'm processing live vocals in Buzz (vocoder or something), I export the track or a section to wav.. load it in the multitrack, play that back and record the vocals only within buzz. Or I do playback of everything in Buzz, but record just the vocals to the multitrack editor.. that requires virtual cabling or a send/return via asio. I usually do the latter because its faster/better. Alternatively, you could just put a recording machine on just the vocal chain.
User avatar
magmavander
Posts: 688
Joined: Tue Nov 22, 2011 5:22 pm
Location: France
Contact:

Re: Vocals - beating it to fit (and painting it to match)

Post by magmavander »

There was PitchWizard as Buzz machines that was able to make pitch correction on vocal (+/- 1 octave) but it seems that it doesnt work anymore with new buzz build, it makes it crashes...
There is a free vst called VisualVox VST :

VisualVox VST is a free intonation correction VST Plugin. With it you can

- shift notes of a voice recording to perfect intonation
- fit a voice recording into rythm by stretching, cutting and moving parts of it
- fit a voice recording to another key with minimal pitch changes

- produce a perfect vibrato
- visualize how right or wrong you are singing
- create artificial polyphony from single voice recording
- draw frequency curves freely with your mouse
- let the VST plugin improve intonation automatically
- play your voice like an instrument via your Midi Keyboard


Download :
http://freemusicsoftware.org/plugins/VisualVox.zip
User avatar
mcbpete
Posts: 381
Joined: Tue Nov 22, 2011 9:45 pm

Re: Vocals - beating it to fit (and painting it to match)

Post by mcbpete »

Woah nice - thanks for the link man :o
User avatar
rajmond
Posts: 41
Joined: Tue Mar 27, 2012 7:15 pm
Location: Göteborg / Sweden
Contact:

Re: Vocals - beating it to fit (and painting it to match)

Post by rajmond »

deusdiabolus wrote:I thought I would pick your brains regarding how you all work with pre-recorded a cappella vocals. Specifically...say you have an a cappella vocal that you want to work into a track that is considerably faster or slower, but you want to keep the vocal at the same pitch. What I have been doing is using a tracker machine (usually Unwieldy) to trigger the vocal syllable by syllable the way it was originally sung, using the offset command to find each word. Needless to say, this is a very time-consuming process and sometimes I still get a result that sounds artificial. I was curious as to whether there might be another method that works better, something closer to "ACIDizing" the WAV so that it auto adjusts to fit. Hey, you never know if you don't ask, right?
Well... I've produced some three albums with buzztracker and vocals... And I must say, Usually I just record them into wavelab, then cut out the silence before and after and stuff it into the matilde-tracker...

however, sometimes when working on remixes I wanted to do stuff in a different tempo. Then melodyne is the tool for you. You can change both pitch and tempo and whatever you feel up for... http://www.celemony.com/ (however it do cost money!).
User avatar
deusdiabolus
Posts: 36
Joined: Thu Dec 08, 2011 6:29 am
Location: Wichita. No, seriously.

Re: Vocals - beating it to fit (and painting it to match)

Post by deusdiabolus »

Evelon wrote:There is "stretch to x ticks" column in uTracker.
Then adjust tempo to get to original pitch (if the loop is chopped/cut correctly, i.e snapped to grid).
Adjust the main tempo or adjust it somewhere else?
User avatar
Joachip
Posts: 352
Joined: Mon Nov 21, 2011 7:21 pm
Contact:

Re: Vocals - beating it to fit (and painting it to match)

Post by Joachip »

I would say, making large changes in tempo to vocals is only possible in rare cases. Usually timestretching or manually chopping up stuff gives unnatural results, if you change the speed more than a certain limit (which, of course, can be an effect in itself.)

The more I have to change the tempo of the vocals, the more I usually split them into small parts. If I really want full control, I might even have to split them into separate syllables, which I then place manually. This takes a long time.

And then there's another approach: If your vocals are in 95 bpm and your track is 150 bpm, simply increasing the speed of the vocals to 100 bpm might do the trick, as 100 vs 150 bpm will give you some interesting rhythm patterns. It - sort of - fits! ;)
User avatar
deusdiabolus
Posts: 36
Joined: Thu Dec 08, 2011 6:29 am
Location: Wichita. No, seriously.

Re: Vocals - beating it to fit (and painting it to match)

Post by deusdiabolus »

I appreciate all the input you've been giving me. What I have mainly been using this technique for is doing remixes of other tracks. I tried doing the time-stretch thing with Utrk, but apparently I'm doing it wrong or it's not designed to work in that fashion (I am trying to keep the original pitch of the vocal and just change the tempo so it will fit). I haven't tried the VisualVox VST yet (just downloaded it) but I will give it a go.

For what it's worth, using my method (cut vocal into parts, then using a tracker to repeatedly trigger the sample where each word or syllable should be and then using the offset command to "nail" each word/syllable in place), I have found that it's usually best to loop with your beat or another rhythmic instrument in the background when doing offsets, because simply looping in the pattern editor doesn't always let you hear the correct flow of the lyrics against the beat.
User avatar
HerrFornit
Posts: 435
Joined: Sat Feb 25, 2017 12:27 pm
Location: Dortmund
Contact:

Re: Vocals - beating it to fit (and painting it to match)

Post by HerrFornit »

Hi, perhaps you already know this.

the link for VisualVox VST does not work any more. Here is an other one:
http://downloads.fyxm.net/download-now- ... 12707.html


And a manual:
VisualVox VST Manual
Jonathan Schmid-Burgk

1) After loading VisualVox VST as an insert effect on the track which contains the source audio material you have to play it back once from end to end so VisualVox can scan it.
2) In the graphical editor of VisualVox you see the scan on the left side. By clicking the transfer button the whole scan is transferred to the output window on the right side. There you can manipulate single notes.
3) An easy option is to let VisualVox improve intonation automatically by choosing "tune all notes" from the "MANIPULATION" drop down. Before that you should click on the keys between the two main windows to select which of them to be allowed for correction. (Tip: use right click to select major/minor)
4) All changes made are saved with your project file in host applications like Cubase. When loading your project again however, you have to let VisualVox rescan the source track once again without any interruptions.
5) Manipulation of notes in the right window is done like this:
1. By a left click you select a note. The eight sliders below will show its parameter settings.
2. With the left mouse button you can shift notes vertically without affecting their time position.
3. With the right button you can move notes in time domain.
4. With the slider "Feine-tune" you can tune a single note freely.
5. With the slider "Raw contours" you can fit a note to its basic tune without destroying its fine shivering, which can be important for the natural character of a note.
6. The "Fine contours" slider does the opposite: With it you can flatten the voice without affecting raw contures. If you remove both kinds of contures 100% you get an artificial sounding robot voice.
7. If you move the little black lines at the ends of a note, the effect depends upon the mouse button you use:
• With the left button you can time stretch notes to fit them into the rythm.
• With the right button you can shorten notes.
8. The three vibrato sliders are used to produce a "perfect" vibrato. The vibrato attack slider affects the time it takes until the vibrato reaches its full amplitude.
9. To delete or duplicate the selected note (the latter to create poliphony for example) please use the drop down "MANIPULATION".
10. On the right side above the output window you will find three mouse cursor tools:
• The star is the tool you use for all manipulations described above.
• With the "X" you can disect notes.
• The little line stands for a pen. With it you can draw frequency curves with the mouse pointer. With the right mouse button they can easily be erased again.
6) Playback is initiated automatically and synchronously when the host plays back the project. In addition to that, you can use the little play and stop buttons below the output window to play back the current frame of audio material.


Here is another program, kind of melodyne and harmonizer: Vielklang2
https://products.zplane.de/downloads/vielklang
It's a demoversion, which is usable with restrictions. But you can record samples in buzz as wav files.
Post Reply