Jump to content

Here's a Cool DAW Tip. Got Any Others?


Anderton

Recommended Posts

  • Members

I was adding background vocals to a song, but the harmony was reaching the upper limit of my range. As pitch shifting is pretty good these days, I did a premix of the song and pitched it down two semitones. This was easy to sing along with, so the immediate problem was solved.

But here's the cool part...when I transposed the vocals up two semitones to fit with the real track, the vocal quality was subtly different due to the formant shift...almost like female vocals, but not quite; more like angelic, which is ideal for background vocals. The quality of the pitch transposition was such that the voices sound fine, no glitching or anything, just...different. This wasn't the kind of formant shift you get through resampling, it was more subtle.

Just for kicks I tossed the vocals into V-Vocal to adjust the formants downward, so I could see if it was possible to retrieve my natural sound. I could, but I liked the formant-shifted version better.

I used to do this all the time with tape - slow it down a few percent, sing along, then speed up to give my vocals more of a "pop" sound. Two semitones is a fair amount of shift, but with today's decent pitch transposition algorithms, it's very doable. thumb.gif

Link to comment
Share on other sites

  • Members

 

Quote Originally Posted by A. Einstein

View Post

for the demos i don't even bother to record vocal harmonies, I make that with celemony from one single vocal line

 

Yes, you are going the other way, and use celemony to create harmonies, this guy say he just uses one of the shifted voices as the single vocal line, because he think it sounds better than his own voice, perhaps it is because it has been auto corrected and in better pitch.
Link to comment
Share on other sites

  • Members
Quote Originally Posted by electrochrisso View Post
Yes, you are going the other way, and use celemony to create harmonies, this guy say he just uses one of the shifted voices as the single vocal line, because he think it sounds better than his own voice, perhaps it is because it has been auto corrected and in better pitch.
Guy Craig can waste as much time as he want's in pre-production, he is not on my salary list.

And anyway, I don't produce the song "Rum and Coca-Cola" with melodyne with The Andrews Sisters, nor does David Jungle bell that way:


Link to comment
Share on other sites

  • Members

 

Quote Originally Posted by A. Einstein

View Post

Guy Craig can waste as much time as he want's in pre-production, he is not on my salary list.

 

It takes less time to pitch a premix down, sing, then pitch the vocals up than it does to find a suitable singer and hire them!
Link to comment
Share on other sites

  • Members
Quote Originally Posted by Anderton View Post
when I transposed the vocals up two semitones to fit with the real track, the vocal quality was subtly different due to the formant shift...

[...]

I used to do this all the time with tape - slow it down a few percent, sing along, then speed up to give my vocals more of a "pop" sound. Two semitones is a fair amount of shift, but with today's decent pitch transposition algorithms, it's very doable. thumb.gif
Have you tried any other instruments with this technique?

Something that pops to mind would be a distorted guitar rhythm part. Play it around standard and drop it two semitones. I wonder if one would get better clarity doing it this way.
Link to comment
Share on other sites

  • Moderators

Take the lead vocal in the chorus. Duplicate and pitch shift up an octave. It sounds totally ridiculous. Add a nice distortion to get a warm saturation. Get a 1/4 note delay into plate thing happening to really get the note to sound warm and just bloom outward. It's an octave, it'll cut. Keep it warm. Mix it back slightly behind the multitude of other vocals that come in at the chorus.

Link to comment
Share on other sites

  • Moderators

Track the lead vocal several times. VocalAlign and Melodyne them to match up to the lead. Some you shift -10, +10, -20, +20 etc. Some you shift the formant in Melodyne too. These are all doubles pitch wise of the lead vocal. Now pan them all out and mix them back so you think you're really only hearing the lead vocal.

the idea is you're trying to not get that doubled sound. no flamming, no wild card pitch variations. It all tracks the lead vocal except for the intentional, incremental changes through pitch, 10, 20, 30, 40 ms delay, formant, etc. But all different takes. This blends but brings the variations beside pitch and timing to the forefront.

Link to comment
Share on other sites

  • Members
Quote Originally Posted by BlueGreene View Post
Have you tried any other instruments with this technique?

Something that pops to mind would be a distorted guitar rhythm part. Play it around standard and drop it two semitones. I wonder if one would get better clarity doing it this way.
I've recorded harmonica parts by pitching (a copy of) the whole song down to a key I had a harp for and then pitching the harp part back up to the original. Worked great. Except that I'm really a lame harmonicist.
Link to comment
Share on other sites

  • Members

A variation on how to to Craig's varispeed...

I set the master daw's sound card to external sync, patch in an old Adat XT as master clock (no tape in it), start play or rec on the daw, reach over to the XT and click the pitch buttons up or down.... daws instantly varispeed for whatever you want to do for overdubs.

Just like the tape machines but with a larger range of varispeed limits.

Link to comment
Share on other sites

  • Moderators

Here's another. Hard pitch correct bass guitar. Sometimes in Melodyne I'll straightline the bass guitar. This give an authentic sound from the fact that it's a bass guitar after all, but the solid as hell pitch can sound great. Sounds way evil.

Link to comment
Share on other sites

  • Moderators

How about...

When reamping, use a automatable send. That way you can actually use the send fader like a guitar knob going in to the cranked stage amp. You know? Why send a static level? You crank the amp then turn down the guitar for the verses, the solo comes and you dime out the guitar volume knob. It is a great way to bring life to reamping, and do some impossible tricks like guitar knowb crescendo while playing in into the bridge, etc. Look Ma! Three hands!

Link to comment
Share on other sites

  • Members
Quote Originally Posted by Lee Knight View Post
How about...

When reamping, use a automatable send. That way you can actually use the send fader like a guitar knob going in to the cranked stage amp. You know? Why send a static level? You crank the amp then turn down the guitar for the verses, the solo comes and you dime out the guitar volume knob. It is a great way to bring life to reamping, and do some impossible tricks like guitar knowb crescendo while playing in into the bridge, etc. Look Ma! Three hands!
I have a lot of problems with gain staging while reamping. When I'm playing, I change my right hand attack depending on how the rest of the rig is responding, and that's a very dynamic, expressive process. When I reamp it just feels dead because unless the rig is pretty much what I had going in my headphones while I was recording, the part loses that dynamic response.
Link to comment
Share on other sites

  • Members

The coolest trick on all DAW is to read the particular manual, and then work with the possibilities the DAW offers.

Then each time during work and having an idea, then you just do it. The possibilities are virtually endless how you can treat music in a DAW.

Link to comment
Share on other sites

  • Members

 

Here's a Cool DAW Tip. Got Any Others?

 

 

That's pretty cool. I still do the pitch thing with analog, but you can't go as far due to the chipmunk effect from the faster vibrato.

 

One of my favorite tricks, or I guess just what I normally practice is to run all my analog channels (16 separate ins and outs total) through an analog console and do all the EQ and processing outside the box. The DAW thinks it

Link to comment
Share on other sites

  • Members

I worked with a band where back vocal wasn't too good. I used Poly Melo to tune all the vocal parts, then extracted Midi file, put it in Cubase, added some Omnisphere vocal patches under the back vocal. You would never tell there was something wrong.

 

One of the first trick was when I tuned just a saxophone ( which was slightly flat) part in mastered track. That was fun!

Link to comment
Share on other sites

  • Members

here's a li'l trick I figgered out with V-Vocal:

 

Rather than use V-VOCAL's automatic pitch-shift function, take the time to tweak your notes manually. And on each note's pitchline envelope, keep the gnarly-looking attack, and keep whatever wavering occurs near the attack of the note, however rough (e.g. momentary pitch overshoot). But hard-correct each note from mid-envelope on out to the final release. Keep your crappy attack, but have the pitch be flawless on the tail. It's a subtler way to pitch-shift vocals without nasty "car-whine" crunchiness becoming obvious.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...