Free Software!

Audacity – Noise Removal (D.I.Y. Acapella) Tutorial

System Requirements:
Windows 98, ME, 2000, XP, or later. (Sorry, Audacity is not supported on
Windows 95 or NT 4.0.)

Audacity runs best with at least 64 MB RAM and 300 MHz processor

Please note I am by no means an expert at this extraction method. Please PM me or post other relevant features to me that help “vocal extraction”. I am sure there are better, more accurate ways of doing this.

Step By Step Guide:

Once audacity is fully installed, simply open audacity. You should have a window like this:

Now simply click “file”.

Now “open”, locate your file you wish to open.

Once opened you should see its waveform like so:

Now you have the file opened you need to highlight (Click and hold the part of the wave you wish to select) the area of sound you wish to extract from the vocal e.g. Guitar riff, piano, strings and so on. Remember don’t select an area where there is some vocal or to large an area of sound as it will have too great an affect and take away too much sound.

Simply click “Get Noise Profile”. Click “effect” again now “noise removal” and “get noise profile once more”. Now select the part of the wave you wish to extract the “noise” from. Once again “noise removal”, now select using the slider how much noise to remove, preview it to preferred “noise” removal. (Remember, due to the tools great power it can take too much away leaving little vocal or sound.

Troubleshoot:

1.Why are there annoying beeps and little noises?

Now you are left with a stripped down version of the track. You may notice little “bleep” like sounds. These are remnants of the sound you extracted. These are hard to remove. I usually then play with the EQ controls or if the bleep is in an area where there is no vocal simply delete it.

2.The “noise removal” had poor effect, why is this?

You may find that the extraction had a poor affect this could be because there was loads of other background noise where the vocal was. There is no real way around this unfortunately. Try selecting more precise sounds to remove.

3.Why does your example picture have two waveforms, one below each other, mine doesn’t?

Don’t worry about this. This simply is that mine was an MP3 giving two rows of waveform. You may fine .WAV files and other files only have one row due to the compression styles.

Thanks guy ‘n’ gals for all the interest you gave for this post.
Please PM or post any techniques you have mastered or find better than mine given and I will edit the tutorial.

FAO: Val or admistrator: Please move this to the tutorial thread and make my pictures visble on the post. I used yourfilelink.com because they keep your file hosted for longer but unfortunaley is not a direct link as such.

Cheers,
Tom Williams

Yea just to clarify import raw is very rarely a perferct import option 9x outa 10 it will do the buzz sound as it does not inderstand the data.

As for software not producing perfect acapellas…It simply wont! it just happens it as strong noise removals, filters and eq’s so u just gota keep playing. At the end of the day if u at least get some sort of vocal even with a bit of a filtered sound to it its better than what you started with.
The only time your really gonna stuggle is when you try extracting a vocal from a dance trck with tons of lead layers and kick drums.
Anyway cheers for now,
And I’m glad this thread is sort of helping people!