How to get Enya-type vocal sounds?

[PaulG]

layering is great. but the best advice would be to find a singer like enya, talent and skill wise. then record.

man! how annoying is that! i said it first you just changed what i said and made it your own

Children, stop your bickering!

I called her and neither her or her producer sub out so that's why I asked here...

[chicken muffin]

you actually called enya on the phone

[Speakeasy]

here's a couple of things you should try...
use chorus,bandpass filters,delays,reverb...and doubling...
and vola...instant enya

[kenn]

I have NOT tried this yet, but I think it would be interesting to attempt the Enya effect by:

1) recording a few, or several vocal takes as layers, as everyone's suggested
2) mix these this down to to a fairly homogenized-sounding stereo track (eq, compression & limiting, but no reverb yet!)
3) use this stereo track to gate or vocode a similar-sounding choral patch, which is in unison or harmonies of the vocals
4) garnish all liberally with reverb, mix, and serve.

[deastman]

What makes this work for Enya is that she has multiple dry vocal layers, all of which are slightly different. You need to take your one vocal sample and use it to create multiple, different layers. For each layer, I would try some combination of the following:

-Offset the sample start time
-Slightly detune the sample
-Feed the sample into a granular effect
-Add chorus
-Add flanging
-Add phasing
-Add EQ
-Add reverb

Render the effected track out. Do this same process 4-6 times with different settings, and then layer them all together.

[rockstar_not]

I did read somewhere where they honestly did track her vox a ridiculously high number of times and layered those. I think this was in Electronic Musician or Mix a couple years ago, if I recall correctly.

Also make sure to pick up Ariesverb and use the first preset on the result.

-Scott

[PaulG]

-Feed the sample into a granular effect

I am not familiar with this type of effect. What does this do exactly?

[larm]

-Feed the sample into a granular effect

I am not familiar with this type of effect. What does this do exactly?

It takes random bits of audio "grains" and plays them back in a random fashion with different parameters (could be different distortion, delay, reverse etc.).

It's a bit like the "blur" effect in photoshop

[deastman]

-Feed the sample into a granular effect

I am not familiar with this type of effect. What does this do exactly?

It takes random bits of audio "grains" and plays them back in a random fashion with different parameters (could be different distortion, delay, reverse etc.).

It's a bit like the "blur" effect in photoshop

Correct. A 'grain' is basically a small audio fragment, ranging from a single waveform cycle to perhaps 500ms of audio. The premise of granular synthesis is to take a source sample, select a grain and play it back through an AD envelope, then select another grain and repeat. The results vary depending on the size of the grains and how they are selected, among other factors, but the basic idea is to take a single static sample and transform it into a constantly shifting sound which bears a passing resemblance to the original. Native Instruments Reaktor, for example, has tons of granular synths and effects.

Again, I'm not suggesting that what I listed is an exhaustive effects chain to be followed exactly, but I would experiment with some combination of these methods to transform a single static vocal sample into a bunch of unique, dynamic layers.

[Unkie Al]

Would this technique be possible if you only had one take of the vocal sound from a single singer? Or is it necessary to use multiple singers or at least multiple takes?

I'd go for multiple takes from the same singer. You could prob'ly fake it with a choral plug-in like Clone Ensemble, but you really do need stacked vocals to make it happen. There is a richness and randomness about it that electronics still haven't caught up with yet.

Also, if you give a listen to a couple of Enya's more successful commercial tunes like "Oronoco Flow", you can hear that she's double-tracking her vocals in octaves as well. That helps enrich the sound some more.

Hope this helps. Now sail away, sail away, sail away...

[Sepheritoh]

There are really very little you can do with effects to create the Enya vocals effect. Sha layers her vocals many times - but - Listen carefully to her vocals and you will notice that she does not sing it exactly each time. In fact, most of the times she does harminies. It would help if you know the basics of SATB choir singing, i.e. harmnies in thirds and sixes and then harmonise chords in 3rds, 5ths and 7ths. Sounds like she eventually sings every possible note on the keyboard, but to make it sound right you need to know what you are doing. She had classical training and grew up in a family of choir singers, so she knows the basics.

It was said that she sometimes double track up to 30 times, but 10 - 12 times is very common.

No choir effect will do what she does convincingly. TC Helicon had a pretty cool harmonizer, but still not as good as Enya.

[Sepheritoh]

Also, if you give a listen to a couple of Enya's more successful commercial tunes like "Oronoco Flow", you can hear that she's double-tracking her vocals in octaves as well. That helps enrich the sound some more.

Hope this helps. Now sail away, sail away, sail away...

And as your Uncle said

[zircon]

Here's an example of the effect in action, as I implemented it. 3-4 layers (split across the harmonies), plus a chorus effect, reverb/delay and EQ. Layered with synth pad. Beautiful result, IMO.

http://www.soundtempest.net/agclip.mp3

Clip is from the song "Antigravity" on the album of the same name. Performer is my girlfriend, Jillian Goldin.

[PaulG]

Here's an example of the effect in action, as I implemented it. 3-4 layers (split across the harmonies), plus a chorus effect, reverb/delay and EQ. Layered with synth pad. Beautiful result, IMO.

Hey, it sure is!

So when you are talking about 3-4 layers, do you mean she tracked each line 3-4 times? Or that you split one original track 3-4 times and processed each one differently?

[zircon]

There were 3 or 4 (can't remember which...) total mono recordings, with each recording representing one note of a chord. She sent me a mix of all of these into one mono recording (with each note appropriately balanced, volume-wise.) The FX chain I used gives it width. We could have panned the individual voices separately, but I think we wanted all the pitches to sound equally in both ears.