I just made a vocoder using only ableton components!

lola · Post by **lola** » Mon Dec 24, 2007 6:39 pm

Angstrom is the best

Very cool!

You should do a Video tutorial some time with all your tricks and stuff, and put it online ..
I would pay for it.

Nokatus · Post by **Nokatus** » Mon Dec 24, 2007 7:55 pm

Angstrom wrote:The sidechain input of compressor has a filter, which is used to asses the incoming vocal, so for example a band pass on the sideband at 1k will only make the compressor compress when a signal is present in the vocal at 1k.

Ah, forgot about compressor's own filter

Angstrom wrote:Like you guessed, A duplicate rack-chain in parallel, with a Utility set on phase invert. This cancels out similar signals and lets disimilar signals through. When the compressed synth goes quiet it is because A - B = 0 .

That's how I envisioned that part, yep. Two chains, one inverted, with its level differences being induced by the compressor.

I think a cancellation effect rack like that is a good thing to keep handy in other occasions, too, as you can use it anywhere as a smoothly reacting gate. Good stuff.

Nogi · Post by **Nogi** » Mon Dec 24, 2007 8:07 pm

Nokatus wrote:A great idea! Damn!
But hmmm. Have you considered trying out a sidechained gate in the manner I described above, in the place of the compressor? I wonder how that would sound. Might even work, somewhat, without the need for phase inversion tricks? Although the bands wouldn't react nearly as smoothly...

That was exactly my thought process as well. turns out using a noise gate as an envelope follower is a bit glitchy - but then again that is sometimes good. The results were certainly nowhere near as usable as Angstroms - maybe I'll put them in parallel with each other and see what amounts - interesting sounds or a quad core on fire.

@Angstrom: Well done! And, thanks for the pack.

john gordon · Post by **john gordon** » Mon Dec 24, 2007 8:25 pm

im surprised ableton never made one like that and tried selling it for 199.00.

tricil · Post by **tricil** » Mon Dec 24, 2007 9:01 pm

holy shit, thank you...

you're right about the CPU kill... i had to bring up my samples to the default 512 to get this thing right.

Grahambo! · Post by **Grahambo!** » Mon Dec 24, 2007 9:40 pm

So, bare with me for a second; newb question:

how do I impliment this vocoder in my OWN projects? Should I just save this thing as a Rack or instrument or something, and when I want to vocode something....uhh.. what do I do? Cause I'm looking at it (bare with me again, im a newb) and I dont see how it picks up the vocal audio?

I don't see that the vocals somehow mix into the synth part or vice versa...sooo how do I work this ? (other than playing his sample set over and over, which is cool as hell, dont get me wrong, but useless in my own projects).

Angstrom · Post by **Angstrom** » Mon Dec 24, 2007 9:51 pm

you can save the rack named "Vocoder" and use it wherever you want, the synth is replaceable, as is the input.

I've not checked this but I assume that (without changing anything) it will always take it's audio input from the output of Track 1.

It gets the 'vocal' part by the sidechain hidden deep in each chain so it's currently a bit of a pain to alter to point at different vocal sources. Go into each chain -> compressor sidechain and pick it. Quite tedious.

But, I have an idea to fix this, so in future there will be one place where you can choose the source audio ... but at the moment my computer is churning away installing Session Drums, so I can't make the change just yet.

Tone Deft · Post by **Tone Deft** » Mon Dec 24, 2007 11:30 pm

your last name wouldn't be MacGuyver by chance?

thanks for showing us another wicked cool trick.

Palmer Eldritch · Post by **Palmer Eldritch** » Tue Dec 25, 2007 12:12 am

Hi Folks,

Inspired from Angstrom I have build up another vocoder rack.

It has also 13 bands but instead of eq3 I use 13 autofilter plug-ins as bandpass filters for the carier signal. i use the same frequencies for the pass filters as the Doepfer A100 Modul "A129 /1/2"

http://www.doepfer.de/a100_man/a129_1_anl.pdf

Angstrom wrote: It gets the 'vocal' part by the sidechain hidden deep in each chain so it's currently a bit of a pain to alter to point at different vocal sources. Go into each chain -> compressor sidechain and pick it. Quite tedious.

To avoid this I create a second "main-chain" (for the analyse signal) and put in another compressor with monitoring the side chain signal
(it is mportant to switch the speaker symbol of this chain to off).

Then the sidechain inputs of all the compressors at the end of the 13 sub-chains of the first main-chain can be switched to > "post fx" of the second main-chain (means behind this compressor of chain 2).

After reloading the rack I only have to switch the sidechain of this one compressor to the source of the analyse signal (voices, drumloops, etc.)

chain 1 > chain 1 > Auto filter 1 > chain 1 sc compressor 1 <-------------<|
................................................chain 2 phase revers utility 1
.............chain 2 > Auto filter 2 > chain 1 sc compressor 2 <-------------<|
................................................chain 2 phase revers utility 2
.
.
.............chain 13 > Auto filter 13 > chain 1 sc compressor 13 <---------<|
.................................................chain 2 phase revers utility 13
chain 2 > Vinyl Distortion > -sc monitor- compressor >--------------------^

Track w/ modulation signal --^

The Vinyl Distortion is to generate some klicks because it seems that the sidechain monitoring only work if there is a signal at the regular input of compressor (I think to avoid cpu consumption of the compressor without a signal to work with)

On my system (specs below) this rack use approximately 50% of CPU with a simple 1 OSC Operator as carrier signal (512 samples audio-interface latency)

cheers (and thank you Angstrom) eldritch

Angstrom · Post by **Angstrom** » Tue Dec 25, 2007 12:19 am

Palmer Eldritch wrote:
Angstrom wrote: It gets the 'vocal' part by the sidechain hidden deep in each chain so it's currently a bit of a pain to alter to point at different vocal sources. Go into each chain -> compressor sidechain and pick it. Quite tedious.
To avoid this I create a second "main-chain" (for the analyse signal) and put in another compressor with monitoring the side chain signal
(it is mportant to switch the speaker symbol of this chain to off).

Then the sidechain inputs of all the compressors at the end of the 13 sub-chains of the first main-chain can be switched to > "post fx" of the second main-chain (means behind this compressor of chain 2).

yes, that's exactly how I planned to do it, sorry I got sidetracked by playing with session drums, my box arrived this morning so I've been filling my harddrive with drums.

I'll update the pack at some point with the modified version, with an 'input' chain to control the master modulator selection.

Oh, I should say that I used EQ3 instead of Autofilter because EQ3 is a a 48db per octave filter, while autofilter is only 24db per octave. Steep curves are good for these sort of applications I think. Of course Autofilter is a better match for the inline compressor filter, but I figured steepness in one place was better than matching, but too shallow, filters.

john gordon · Post by **john gordon** » Tue Dec 25, 2007 12:32 am

brilliant!!!!! good job.

Angstrom · Post by **Angstrom** » Tue Dec 25, 2007 2:04 am

I updated the rack with the special routing system, now it is much easier.

in the updated Vocoder look at the chain called 'INPUT' (igone the one called 'output')
In the sidechain of the compressor you see in that track you must choose your vocal source (or whatever you want to use).

that's it. The vocoder bands are now getting audio from the source and as long as you have a synth or some audio on the same track as the vocoder you should hear some vocoding.

download the (updated) Vocoder rack (128k)

I also noticed that adding a macro to this enables you to quickly get a better sound from the source. I actually turned the input down a little and it suddenly got a lot lot better !

Still, I know that 40% is a lot of CPU, but if you really need a free modular vocoder, you can't complain too much

BAKOY_MUSIC · Post by **BAKOY_MUSIC** » Tue Dec 25, 2007 3:12 am

if we can get in your site that is

hanging out here !

Angstrom · Post by **Angstrom** » Tue Dec 25, 2007 3:47 am

I can get a connection quite easily, there doesn't seem to be an issue with the server for me. My transfer limit is nowhere near being reached.

the logs say that 63 people have downloaded the Live pack and 29 people have downloaded the rack. So it must be working for some people.

I'm not sure why you are having a problem.

Grahambo! · Post by **Grahambo!** » Tue Dec 25, 2007 5:24 am

Newb stuff again, sorry. I downloaded the updated rack. Cool. Bu how do I put a synth in it? I don't understand where to go from here. I can't put a synth instrument in it, according to Live 7. Maybe because it thinks it's an audio track? Live just tells me i can't insert an instrument because only Audio FX can go in there... Hm. Sorry for the Newb junk questions...

EDIT: NM. Add instrument first. then Vocoder. Got it.