Moment 4 Life Censor?

[nlpictures]

Hey,
http://www.youtube.com/watch?v=_f-lnsd9Xf4#t=1m49s
I was wondering if it is possible (in Audacity) to censor a song using that method. Where like it sounds like a DJ scratch? Because it seems like an easy method to use if you dont have a instrumental.

Thanks,
Colby

[kozikowski]

How many minutes into the song is the effect?
Koz

[Trebor]

How many minutes into the song is the effect?
Koz

Here's a sample from 2:13 to 2:20 ...

Moment 4 Life (Clean) - exerpt with censorship warps.mp3

(281.25 KiB) Downloaded 433 times

is possible (in Audacity) to censor a song using that method. Where like it sounds like a DJ scratch?

Yes and no.

On the Youtube “clean” version the warp ("scratch") effects are only applied to the vocal track, not to the music.

You probably only have access to the “dirty” mix , if you apply warp effects to that it will warp the whole mix : the music and the vocal.

So yes you can edit out swearing using audacity , but no it won’t sound exactly like the Youtube track because you probably don’t have access to the vocal track on its own (without music) .

You can get a “Turntable Warping” plug-in effect for Audacity here … http://audacityteam.org/download/nyquistplugins

[Trebor]

Moment 4 Life, before-after the word 'songs'' censored with 'turntable warp' in Audacity.mp3

(648.98 KiB) Downloaded 325 times

similar effect but not quite the same as the music is also warped, not just the vocal, (see my previous post).

It may be possible to improve on this if there is a bit of the looped music without vocal somewhere on the track which could be copied and pasted-in where the word was removed.

[nlpictures]

Hi,
So I downloaded the plug-in, and I read this:

Initial step: [-36 - +36 semitones, default 0] - the resulting pitch change at the start of the selection, defined as number of semitones above or below the original pitch (1 step = 1 semitone, so 12 steps = 1 octave.)
Initial amplitude: [0 - 100 percent, default 100] - the resulting volume at the start of the selection, related to the original amplitude (so 100% = no change).
Change time: [0 - 100 percent, default 50] - the point in time in the selection (set internally to the original pitch and volume) at which the warping values change to reach and move away from this point. What this means is that depending on your start and end step and amplitude values, you can design warps that slow down then speed up; speed up then slow down; speed up to a particular pitch then remain there; slowly speed up then quickly speed up.End step: [-36 - +36 semitones, default -12] - the resulting pitch change at the end of the audio. compared to the original pitch.
End amplitude: [0 - 100 percent, default 40] - the resulting volume at the end of the selection, compared to the original volume.

How do you design a warp that slows down and speeds up? Can anyone tell me what settings to use?

[Trebor]

... Can anyone tell me what settings to use?

Initial Step 0

Initial amplitude 100 (100 => effect doesn't change volume)

Change time ? (the point in the selected portion of track at which the warp begins,
e.g. 50(%) => effect begins half way into the selected portion of the track, 75=>effect begins in last quarter of selected part of track)

End Step ? (for slow down use negative values)

End amplitude 100 (100 => effect doesn't change volume)

As I mentioned previously in this thread, unless you have access to the vocal track alone, without music,
applying the effect will warp the vocal and the music, which is not the same as the effect on the video where only the vocal is warped.

turntable warp settings.png (20.62 KiB) Viewed 7239 times

Reversing part of the swear word is easier than warping : here I've reversed "ong" of "songs" ...

before-after 'ong' of 'songs' reversed.mp3

(528.98 KiB) Downloaded 307 times

The timing is not changed by reversal, (whereas warping changes timing).
Just select part, not all*, of the swear-word, reverse it, "Reverse" is in "Effect" menu, then use "Repair", (also in "Effect" menu), on either end of the reversed bit so there is no click.

* You should only reverse part of the swearword so even if the track is played backwards (reversed) the whole swearword will not be heard.

[steve]

You could also try running this code in the Nyquist Prompt effect:

Code: Select all

(setq depth 2.5)
(defun scratch (s-in)
(fmosc 0.0 (mult depth (hzosc (/ (get-duration 0.5)))) (list s-in 0 nil) 0))
(multichan-expand #'scratch s)

You can adjust the depth (how much the speed is changed) by changing the number in the first line (currently 2.5)
This should work well on selections of about 1/4 to 1/2 second duration.

[Trebor]

You could also try running this code in the Nyquist Prompt effect:

Code: Select all

(setq depth 2.5)
(defun scratch (s-in)
(fmosc 0.0 (mult depth (hzosc (/ (get-duration 0.5)))) (list s-in 0 nil) 0))
(multichan-expand #'scratch s)

That's better than turntable warp plug-in for this censor task as it keeps the original timing after the warp, ( the speed-up must exactly cancel-out the slow-down).

The only restriction I noticed is it refused to process selections longer than two seconds, which is not a problem in this censorship application, unless there is a two second profanity I don't know about.

[steve]

The only restriction I noticed is it refused to process selections longer than two seconds,

Yes. The maximum duration that can be processed is 100000 samples.
The technical information about this is that Nyquist in Audacity has the maximum size for a "wavetable" hard coded as 100000 samples.
If I get round to making a plug-in version of this I'll have to remember to add an appropriate error message for this.

[samspots]

If the vocals are in the center track, you can use an m/s matrix splitter, apply it to the center track (left channel), and recombine.