Vocoder basics here!

axxim · Post by **axxim** » Tue Nov 03, 2009 1:22 pm

Hi again,

for all those interested in vocoders or new to them, may find some tips and simple basics here:

http://www.axxim.de/korg_r3/

there you will find also a section with vocoder programs for the R3.

TXBDan · Post by **TXBDan** » Wed Nov 04, 2009 1:15 am

Very cool. Thanks for that.

Is it possible to use the R3 as an autotuner? Ie, apply a minimal effect? Or does it not have enough bands to get close to accurate enough for a 'real' voice?

axxim · Post by **axxim** » Wed Nov 04, 2009 10:08 am

Hmm,

I don't know exactly what an autotuner is but I think you mean this voice correction that weas used by Cheer where the voice gets in a perfect tuned loop.. If I am wrong please tell me what it is (a link maybe?)

If I am right then the answer is no because the autotuner affects the voice signal. In the vocoder it is the opposite: the voice is what affects the carrier sound. The tune of the vocoded output is the tune of the carrier that is electronically exact if wanted. To get an autotune effect with a near to the original voice maybe a Fourier-vocoder with 512 bands or more will do it. But here I have to pass because my first reason to have a vocoder for me is because I don't like my voice.

But you can do a mix of your voice and the vocoder. Just recently forum member Sunday Radio (http://www.korgforums.com/forum/phpBB2/ ... hp?t=40028) created a nice song that you should take a hear at.

virtualmidei · Post by **virtualmidei** » Wed Nov 04, 2009 10:47 pm

hey man i need help with my vocoder R3

when i record the vocoder already

and put to listen it on audio is listening my normal voice also mixed with the vocoder

So the point is that i want only the vocoder sounding and not my real voice

u now what i mean??

axxim · Post by **axxim** » Wed Nov 04, 2009 11:56 pm

Hi virtualmidei,

no I don't know exactly what you mean... If you want you can ask me in spanish (I see you are in Mexico).

What do you hear when you use headphones? vocoder only? Maybe you have raised the direct level or something so. I would recommend you to go to my homepage anounced here and download the vocoder sounds, then test if it sounds like vocoder without your voice in your headphones. If so then you should record it in cubase and it should sound the same as with the headphones.. if not then you have something wrong with your cabling

Spanish: Si no entiendes bien lo que te escribo, describeme el problema en español

Saludos

virtualmidei · Post by **virtualmidei** » Thu Nov 05, 2009 5:21 am

hehehe que bueno mucho mejor

ok mira el problema es que cuando grabo dentro de cubase en el audio track

suena el vocoder pero .... siento que suena mucho mi voz real

pero bueno el punto es como se debe de grabar dentro de cubase el vocoder?

apagando los speakers y grabando con headphones o se puede encima del track con los speakers prendidos esta es mi gran duda espero me entiendas ahora

axxim · Post by **axxim** » Thu Nov 05, 2009 10:41 pm

Sorry to all others, but I will switch to spanish for this replay!

Hola VM,

es posible que tengas otro microfono (por ejemplo de una webcam) conectado y activado?

El micrófono solo debe estar conectado al sintetizador. El programa de vocoder que uses puede ser que tenga algo de señal directa, la cual no sientes por los auriculares por que tu mismo estas en ese momento cantando. Prueba con el programa RobotR que esta en mi pagina web, o cualquiera de los otros que se encuentran alli (para vocoder por supuesto)

Saludos

TXBDan · Post by **TXBDan** » Fri Nov 06, 2009 1:04 am

Thanks, axxim.

Basically i can't sing very well so i was wondering what kind of true-to-life fidelity i could achieve using the R3's vocoder. Of course it wouldnt be perfect or even very good, but i'm curious.

For example i heard that that stupid Owl City Fireflies song thats on the radio all the time is the guy using a vocoder. It sounds pretty clean to me. Is this possible with the R3 or is using software with a lot more bands?

xmlguy · Post by **xmlguy** » Fri Nov 06, 2009 1:46 am

TXBDan, vocoders are not intended to have true-to-life fidelity. In fact, they're exactly the opposite. A vocoder replaces your voice with a synth voice. I don't hear any vocoding on the Owl City track. I hear doubling, transducer, maybe some autotune. What part do you think is vocoded?

There is a very common misunderstanding where people hear some vocal effect such as pitch correction (AutoTune, T-Pain/Cher effect) then hear or read someone else calling it a vocoder, even the artist himself- except that they're all wrong. The same thing happens with talkboxes, harmonizers, doublers, and transducer effects that have nothing to do with vocoding.

On fireflies, there's definitely a transducer effect (ie. megaphone-oldtimey mic sound). The vocals are probably doubled the old fashioned way, singing in harmony to yourself on a multitrack an octave below the melody. A harmonizer or doubler could be used, but since this is a produced track, there's a lot of different ways to achieve that doubling effect. I think double tracking is more likely because the timbre of the voice is a bit rougher on the down octave harmony, not just pitch shifted.

TXBDan · Post by **TXBDan** » Fri Nov 06, 2009 1:57 am

i see, interesting. thanks for the info!

xmlguy · Post by **xmlguy** » Fri Nov 06, 2009 2:28 am

You're welcome, TXBDan. By the way, the arpeggio at the beginning of Fireflies is an interesting challenge for people to figure out on the R3. The tricky part is the legato portamento bit that occur between the basic staccato notes.

axxim · Post by **axxim** » Fri Nov 06, 2009 9:10 am

Hi,

seems to me that I have to search and hear that "Owl City Fireflies" song to get an idea of both, the voice and the arpeggio...

2disbetter · Post by **2disbetter** » Fri Nov 06, 2009 10:04 am

xmlguy, thanks for the break down, very informative.

xmlguy · Post by **xmlguy** » Fri Nov 06, 2009 11:17 am

Glad to help, 2disbetter. Vocoding seems like such an easy thing, just sing into the mic and hit the keys, but I've found that it takes quite a bit of experimentation and practice to get consistently good results, particularly if you want the audience to be able to easily understand the words. Sometimes your voice needs to sound quite unnatural and very odd if you hear it as a source before the vocode occurs, while the result after the vocode sounds great.

I like to think of it as using a synth to replace your vocal chords, somewhat like what a talkbox does, but the tube used in a talkbox has its own distinct characteristics such as a narrow frequency range, while a vocoder carrier can have a wider frequency range and has the formant shift function that lets you move the range up or down, outside the normal vocal range. The R3 vocoder also splits the bands panned left and right alternately for a unique stereo effect, and each band has its own level control that you can use to alter the overall frequency profile.