Warping Vocals

[Raveshaper]

As some of you might remember, I posted about something I called "Vinyl" a while back.
It is a simple codec file that binds song tempo to changes in pitchbend. The result is that when you pitch samples and record the output as you play it back, the sample changes pitch but stays at a constant speed. This makes it possible to not simply transpose a sample to a new note, but to achieve portamento-like note glides and other dynamic behavior.

At the time, I didn't really have a decent demonstration of what this simple tool could potentially achieve. You had to listen very closely to the original "tech demo" to hear anything odd happening.

So, to illustrate how this works, I used Calvin Harris's acapella of Ellie Goulding singing I Need Your Love on SoundCloud and processed it using Vinyl.

Here's the audio:


Here's what's happening:
1. Original audio for reference.
2. Vocal flattened by Vinyl (manual process). All notes corrected to root key (C 3).
3. Flattened vocal doubled, one transposed -12, the other transposed +7. For reference of linear transpose limitations.
4. Vinyl warped version of the flattened vocal. For reference of what vinyl warping can achieve.

Edit: While this may not sound the best, pay attention to how the audio still matches the same points in time even though it has been warped. This is impossible to achieve by just using the pitch wheel on a sampler. It works on any sample, not just vocals, increasing the range of creative possibilities for any given sound source.

This would be a hell of a song challenge ingredient, but I would have to do a write-up on how to apply it first.
I haven't had luck explaining myself in the past, but who knows.

[JNeffLind]

Hey man. Is this the thing we'd talked about via e-mail. I'm curious, but can't get the player to play in the link you sent.

[Raveshaper]

I think I told you about this one. So far, I'm 0-2 on getting this to play for anyone else. I guess Google is shutting down their sharing options along with Google plus.

I suppose I can mirror it on SoundCloud and embed it here. Sorry it won't play directly.

[Raveshaper]

Here's the SoundCloud mirror of the tech demo:

[thefixr]

Just listened to this and I am seriously impressed. I could probably do a lot with this!

[CR68]

sounds great! how can we use this? is it inside reason or standalone? would be great to get more information!
thx in advance

[JNeffLind]

Here's the SoundCloud mirror of the tech demo:

This is very cool man. Sorry I forgot to check back on this thread originally. Let me say now, nice work! I'm with a previous poster in wondering how to incorporate this. Are you making a refill or RE? If so, I'd be interested in the price was right.

I think you're onto something here man. Tons of potential to do cool stuff.

[Ronin]

very cool!

[Puckboy2000]

This is Jedi magic!!! Please post a video. I will buy it if it's an RE or an easy to use Combinator effect. Excellent work! I NEED to be able to do this for my songs!!

[Raveshaper]

My schedule is pretty loaded but I will work hard on polishing this one for release.

Thanks for the support guys!

[Raveshaper]

A quick update on the progress:

I have expanded the original architecture from +/- 24 semitones to +/- 59 semitones. This is a hard limit imposed by the maximum tempo range and the exponential function used to adjust sample playback.

In order to use the full range of tempo, the initial song project tempo must be no more than 32 bpm. This is only the default tempo prior to importing the audio you want to process using Vinyl. After you capture the processed output you can export and use it in projects with your desired tempo.

But this means everything from ~1.000 bpm to ~999.999 bpm is fair game. Pitch can be shifted almost out of the audible spectrum without losing sync with tempo. And of course, any effects that use time signatures, like tap delays, LFOs, etc. will give unexpected results depending on settings just as before.

I will need to restructure the code a bit more. I'm adding a pitch extraction process that converts semitones to pitch values, complete with an optional mode for inverting pitch to allow for automated flattening of a sample.

More soon.

[freeQlow]

Very Cool.
Code?
Are you trying to save a template of automation?
Is this in Neptune? Audio?
I'd like to learn more as well sir, take your time!

[Raveshaper]

This is a way of manipulating audio to achieve extreme pitch shifting and advanced variable speed re-sampling of any sound source.

The results that you will be able to get out of this will make Neptune look like a child's toy that needs its batteries changed.

Some of the sounds you end up with can not be created in Reason through any other known method.

This works inside native Reason, no REs required, no stand-alone app running in tandem.
Requires MIDI hardware to function.

Beyond that, I'm too busy working on it to go into lengthy details. I will be making a video or two about it when it's fully tested and ready.
It suffices to say that it will revolutionize the way music is made in Reason, full stop.

[Puckboy2000]

This is a way of manipulating audio to achieve extreme pitch shifting and advanced variable speed re-sampling of any sound source.

The results that you will be able to get out of this will make Neptune look like a child's toy that needs its batteries changed.

Some of the sounds you end up with can not be created in Reason through any other known method.

This works inside native Reason, no REs required, no stand-alone app running in tandem.
Requires MIDI hardware to function.

Beyond that, I'm too busy working on it to go into lengthy details. I will be making a video or two about it when it's fully tested and ready.
It suffices to say that it will revolutionize the way music is made in Reason, full stop.

I can't wait! Will the pitch shifting be done via mod wheel???
Like I said before, this is Jedi Magic!!!!!!

[Raveshaper]

Pitch wheel support will exist, but there is currently no known way of allowing a single pitch wheel to send enough data to drive the entire range of pitch. This means the full pitch range will make use of pitch bend on midi channels 1, 2, and 3. You'll just have to be able to edit the pitch wheel to use different midi channels if you want to manually manipulate the entire ten octave sweep. Or, you can edit endless rotary encoders to send pitch bend data on hardware like Maschine.

[Raveshaper]

I'm still developing, this is by no means an official release, sadly.

I just wanted to tease the Aphex Twin-level of sound design you can achieve on audio by running it through this process.

First 8 bars are standard, but then we travel all the way up and down the spectrum to see what happens.
And the answer is: amazing.



Edit: the reason you hear stepping in the pitch changes is because this is being driven by sequenced notes. Yes, you will be able to sequence notes and use them to process your track and/or flatten your samples.

[JNeffLind]

Beyond that, I'm too busy working on it to go into lengthy details. I will be making a video or two about it when it's fully tested and ready.
It suffices to say that it will revolutionize the way music is made in Reason, full stop.

I like your confidence. Reminds me of someone I know (ahem). Be warned though, the internet is full of haters looking for confident people to pick at. Proceed with caution. I think this looks awesome and I believe it'll be a game changer for a lot of people. Take your time and get it right. I (and many others) will be waiting (im?)patiently.

[Raveshaper]

I'm not worried. I'll put the work in and get the best result possible. I plan on folding this process into how I compose. I can see how I won't be able to do without it.