Pitch correction software has applications from restoration and mix-rescue to outright distortion of a voice or instrument. I’ll discuss some of the more tasteful uses of these auto-tune tools (whether the original from Antares, or a variant like the free GSnap) below. But first I thought I’d highlight their misuse to illustrate the effects we usually try to avoid.
So, listen here to 10 of pop music’s most blatant auto-tune abuses:
If you’re unfamiliar with Auto-tune, and especially if you listen to much pop and rock, you might not hear it initially. When overdone, the effect yields an unnatural yodel or warble in a singer’s voice. But the sound is so commonplace in modern mainstream music that your ears may have tuned out the auto-tune!
The songs in this clip, in order, and the phrases most affected by auto-tuning to help you spot them:
Dixie Chicks – The Long Way Around – Noticeable on “parents” and “but I.”
T-Pain – I’m Sprung – Especially obvious on “homies” and “lady.”
Avril Lavigne – Complicated – Listen to “way,” “when,” “driving,” “you’re.”
Uncle Kracker – Follow Me
The whole vocal sounds strained, but especially the word “goodbye.”
Maroon 5 – She Will Be Loved – Listen for “rain” and “smile.”
Natasha Bedingfield – Love Like This – “Apart” and “life.”
Sean Kingston – Beautiful girls – “OoooOver” doesn’t sound human.
JoJo – Too Little Too Late – Appropriately, “problem” stands out.
Rascal Flatts – Life is a Highway
Every vocal, foreground and background, is treated, but “drive” in particular.
New Found Glory – Hit or Miss – “Thriller”, and every time Jordan sings “I.”
The Cher Effect
When used noticeably, an auto-tuner produces what most call “The Cher Effect“, named for her trademark sound in the song Believe*. (In essence, we named the effect like scientists naming a new disease after its first victim.) Treated this heavily, a vocal track sounds synthetic, and obviously processed.
But not all auto-tuning is so blatant. In the sample above, it’s harder to hear the pitch correction on Uncle Kracker and Avril than on T-Pain and Bedingfield.
Tasteful Uses
As with any tool, a little care can yield great results. Some simple things to keep in mind about pitch correction tools:
- Performance: Most importantly, an auto-tuner isn’t a shortcut to a perfect performance. If you can’t sing the song properly, no amount of post-processing will make it sound like you did. So when your pitch matters, and you don’t want to correct it with an effect, you’ll need to work on your performance until it’s right.
- Less is more: The fewer notes you correct, the less obvious your use of an auto tuner will be. Consider automating the plugin so it acts only when most needed.
- Graphical mode: If your pitch correction software offers a graphical mode (like Antares Auto-Tune and Melodyne,) learn how to work with it. The default “auto” modes are OK for basic corrections, but often produce noticeable yodeling.
- Backing vocals: In general, you can get away with more pitch correction on backing vocals than lead vocals.
- Outdated: Obvious vocoder-style autotuning is dated, and borders on kitschy. The synthetic warbling vocal sound marks songs as having come from a specific era, the same way gated-reverb on drums instantly places a song in the 1980’s. Remember: If you make the auto tuner obvious, people will say your song uses “the Cher effect.” Let this be a guideline.
Be sure it’s needed
Two songs have auto tuners on my mind today: Snoop’s Sensual Seduction (because of Anil Dash’s ruminations on the death of the analog vocoder,) and Natasha Bedingfield’s Love Like This, which I heard on the radio. In the former, the auto tuner is clearly a gimmick. But every time I hear Bedingfield’s song, I’m struck by the same question: Why do that to her voice?
She’s a fantastic singer, and once you’ve heard the song without the cheesy auto tuner effect, it’s hard to take the radio single seriously.
And there’s a lesson in that for home recordists, (even those of us who don’t write pop music,) which echoes the rule of mixing: If an effect significantly changes the sound of a track, especially one so important as the lead vocal, be sure that change improves the song before committing it to the mix.
See Also: The Rule of Mixing
For more home recording tips,
Subscribe to the Hometracked feed, or receive email updates.
Tags: freeplugins, mixing
238 comments
Trackback URI Comments feed for this article
I was recently exposed to Melodyne, and it appears to be better suited than Autotone for reining in the overall intonation without killing the inflections. Very intuitive if you want to only apply it to certain notes, etc. — I’ll probably be purchasing the least expensive version in the near-ish future.
Anyone else have experience with it? Anyone experienced with both?
Good post! It seems to me like the effects is intentional in some cases, while in other it’s not.
Can you imagine 20 years from now when they try to recreate the sound of today, and they sit and try to find the plug-in that best “gives that metallic sound on every third word in the chorus”.
Has it occurred to you that some of the artists use the sounds created by the autotone purposely (t pain) in their music to create a different sound than they could with their normal voice? And also the sound quality is terrible with the 128k bitrate MP3 that is provided and the autotone sounds were almost indistinguishable from the sound artifacts present in the MP3.
> Anyone else have experience with it? Anyone experienced with both?
I’m afraid not, Keith. Of the 3 biggies (Antares, Melodyne, Waves,) Antares is the only one I’ve used.
Sound On Sound did a face-off last year, though: http://www.soundonsound.com/sos/mar07/articles/at5vsmelodyne.htm
They prefer Auto-Tune for the quick’n’dirty, but Melodyne has more creative possibilities.
> try to find the plug-in that best “gives that metallic sound on every third word in the chorus”.
:-)
> Has it occurred to you that some of the artists use the sounds created
> by the autotone purposely (t pain) in their music to create a different sound
> than they could with their normal voice?
Yes. It doesn’t make the effect any more tasteful.
Though in T-Pain’s case specifically, it’s understandable why he’d want to change the sound of his normal voice:
http://www.youtube.com/watch?v=YCDWLaFc9u4
> And also the sound quality is terrible with the 128k bitrate MP3
That’s a fair criticism. I replaced the MP3 with one encoded @192K
They prefer Auto-Tune for the quick’n’dirty, but Melodyne has more creative possibilities.
I like creative possibilities. :)
I first noticed the abuse of this effect when I heard Kid Rock’s “Only God Knows Why”…the entire song has been heavily processed with the effect. I think that is where it all went downhill :)
After that it seemed nearly everyone was doing it. It stands out so much and seems to remove all originality from every song it’s used in. Kind of gives the impression that the artist is trying to capitalize on an over-used gimmick instead of coming up with something original on their own.
how do you know the effect wasn’t intentional in a lot of these songs? the auto-tune vocals are pretty popular at the moment..the cher effect comes in 10 year cycles..
I guess I can see why these practices occur as I have been part of the target market given we have a third of the recordings listed above in the house. I took up the guitar as a hobby a couple years ago. That got me doing a lot of reading and experimentation around recording (have learned a lot on this site). Now that I’ve scratched the surface of what goes on under the hood it makes me realize how oblivious I was to what I have listened to previously. Seems that music consumers in the mass market just take what they hear at face value given they do not have the insight into how commercial recordings are made to listen critically. I assume the recording industry must take this into account in their decision making?
This article’s a little old, but relavant. Neko Case goes off on auto-tune; http://www.pitchforkmedia.com/article/feature/31252-interview-neko-case
The relavant stuff is about 2/3 of the way down the page.
I disagree with your implication of the use on Life is a Highway – that may not be the case. I’ve seen Rascal Flatts in concert several times and that’s how the vocals sound without the use of correction software. Gary LeVox has great tone and pitch.
I won’t disagree that it’s used way more than it should be though.
> @Beau: gives the impression that the artist is trying to capitalize on an over-used gimmick
Which we could almost forgive if it sounded halfway decent. But it sounds LIKE CHER!! :-)
> @VintageP: I assume the recording industry must take this into account in their decision making?
I’m sure you’re right about that. Whatever else can be said of Bedingcourt, Kingston, and T-Pain, they sell a lot of music.
> @Jeff: Neko Case goes off on auto-tune;
Sweet, thanks Jeff.
Keith, I have used both Auto-Tune and Melodyne (uno, then the new version’s cre8). I much prefer Melodyne. Especially the cre8. It definitely sounds more natural and the graphical mode allows for better control. It still does require tweaking, though; sometimes it will for no apparent reason shift a couple of notes down an octave when you use the automatic mode. However you can just move them back up in the graphical mode.
I have found that it makes for a pretty good sound if you mix the tuned version and the original version of the vocals together. I’m not a very good vocalist so Melodyne’s been a song-saver for me.
-bill
> it makes for a pretty good sound if you mix the tuned version and the original version of the vocals together.
Heh, that’s one of Paul “SoundOnSound” White’s favorite tricks! It has to be the easiest way to generate a realistic-sounding vocal double. Combined with a tape-flutter effect, you might as well BE John Lennon :-)
dude, you’re off here. i agree with other folks like chimera. yes it is tasteless, but that is the point. that’s pop.
that is an “in” effect that people are intentionally putting into the music. it is a desired “techno” effect for the lack of a better phrase. you’ll find people that can recognize and like it. that’s what happens when you’ve grown up with healthy rotations of funk, techno, Zapp and Roger Troutman (RIP) going into your earhole.
these pop music people know it will be an outdated sound soon, but they don’t care. by that time folks are sick of, they’ll using some other stupid shitty sounding effect
Maroon 5 “She Will Be Loved” is abuse, and probably intentional. The auto-tune is so much more ridiculous on the chorus than on the verses that you figure it had to be a producer decision, to help create a more commercial sound for the single.
Jenn, why do you think you heard them live without autotune? They don’t sing live any more than Britney or NSYNC do, they are a boy band. When you see them live, you hear them recorded, just like any other pop act.
And yes, they really did completely destroy a fantastic song. Its a shame Tom let them do it.
Cher’s vocal on “Believe” was not auto-tuned, it was vocoded with a Digitech Talker. This was done deliberately to create that odd sound on her voice, with her permission. If you google it carefully you’ll find an article about it by the engineer who did it.
Joe, autotune can be done in realtime, so a band can both sing live and use autotune.
> by that time folks are sick of, they’ll using some other stupid shitty sounding effect
Heh, no doubt! (But FWIW I’d say the same thing about “compression abuse” or “EQ abuse” or any other kind of unnecessary mix tomfoolery. I wrote the article to educate, rather than to criticize, is what I’m sayin’)
> Its a shame Tom let them do it.
Hear hear!! (Though with statutory licensing, he may not have had a choice!)
> Cher’s vocal on “Believe” was not auto-tuned, it was vocoded with a Digitech Talker.
That’s a common mis-misconception.
I linked to the article you referenced (click the text “The Cher Effect”) but here it is again for completeness:
http://www.soundonsound.com/sos/feb99/articles/tracks661.htm
Notice the big yellow addendum at the top? (“Stop!! Historical footnote”)
In other words, when they claimed it was a Digitech Talker, they lied!!!
“I’m Sprung” definitely does this intentionally.
almost all of these are intentional. and in some cases they take a single note, and bend it into several notes. that’s not autotuning, that’s adding an effect.
What Derek said: producers use vocoder to make lyrics sound ‘chunkier’ like Cher. Auto tune is for pitch shifting to fix blown notes. You dont seem to get the difference.
More Comments: 1 · 2 · 3 · … · 10 · Next ›