Page 1 of 1

Phasing audio

Posted: Sat Jun 11, 2011 11:33 pm
by zerbaman
I know that this only works efficiently if the files you're using were mastered together. But my question is why it only works this way?

Re: Phasing audio

Posted: Sat Jun 11, 2011 11:40 pm
by 3za
Why what works???

Re: Phasing audio

Posted: Sat Jun 11, 2011 11:44 pm
by zerbaman
Why you can't take any copy of the original instrumental and phase the audio to the vocal version.

Re: Phasing audio

Posted: Sat Jun 11, 2011 11:48 pm
by Sine69
Are you talking about trying to remove everything except for the vocals in a song?

Re: Phasing audio

Posted: Sun Jun 12, 2011 12:00 am
by 3za
Because all the values are the same apart from the vocal, so when you invert one of them it now has the opposite values of the instrumental track. So when you sum them together, you are left with the difference (the vocal) ;-)

Re: Phasing audio

Posted: Sun Jun 12, 2011 12:01 am
by Toric
I've only ever phased audio to remove vocals to either sing to, or get better samples.

From what I understand, it's because the song is mastered in stereo, and the vocals are generally in the center, where as the rest of the audio is panned. When you phase invert a track and lay it on top of itself, it will cancel out anything in the middle, in turn, cutting everything that is panned center.

I hope this helps a bit!

-T

Re: Phasing audio

Posted: Sun Jun 12, 2011 12:02 am
by Toric
3za wrote:Because all the values are the same apart from the vocal, so when you invert one of them it now has the opposite values of the instrumental track. So when you sum them together, you are left with the difference (the vocal) ;-)

Values? This is pretty intriguing to me. Care to elaborate? :)

-T

Re: Phasing audio

Posted: Sun Jun 12, 2011 12:08 am
by efence
picture an instumental and the same track with vocal(doesnt really matter if its been stereo mixed or not)
if the instrumaental with the vocal on top were mastered/compressed/anything then when you invert the insrumental your left with the vocal and all the compression efx that were on the master

Re: Phasing audio

Posted: Sun Jun 12, 2011 12:14 am
by zerbaman
What I want is to end with an acapella of the track, using the instrumental and the full version to phase eachother out.
3za wrote:Because all the values are the same apart from the vocal, so when you invert one of them it now has the opposite values of the instrumental track. So when you sum them together, you are left with the difference (the vocal)
Yeah, but why is it that on some files it doesn't work out this way? Nothing changes, they just play as normal

Re: Phasing audio

Posted: Sun Jun 12, 2011 12:25 am
by efence
zerbaman wrote:What I want is to end with an acapella of the track, using the instrumental and the full version to phase eachother out.
3za wrote:Because all the values are the same apart from the vocal, so when you invert one of them it now has the opposite values of the instrumental track. So when you sum them together, you are left with the difference (the vocal)
Yeah, but why is it that on some files it doesn't work out this way? Nothing changes, they just play as normal
the invert needs to line up perfect bit to bit

Re: Phasing audio

Posted: Sun Jun 12, 2011 12:41 am
by Toric
efence wrote:picture an instumental and the same track with vocal(doesnt really matter if its been stereo mixed or not)
if the instrumaental with the vocal on top were mastered/compressed/anything then when you invert the insrumental your left with the vocal and all the compression efx that were on the master

This makes perfect sense.

Hmm, I guess I'm talking about something different.

So you have to have an instrumental and the track with vocals to be able to do it?

Edit: Figured out what I was missing from this thread. Didn't read something the right way. :)
efence wrote:
zerbaman wrote:What I want is to end with an acapella of the track, using the instrumental and the full version to phase eachother out.
3za wrote:Because all the values are the same apart from the vocal, so when you invert one of them it now has the opposite values of the instrumental track. So when you sum them together, you are left with the difference (the vocal)
Yeah, but why is it that on some files it doesn't work out this way? Nothing changes, they just play as normal
the invert needs to line up perfect bit to bit
Okay I understand now. Awesome! Thanks!

-T

Re: Phasing audio

Posted: Sun Jun 12, 2011 2:19 am
by 3za
Toric wrote:
3za wrote:Because all the values are the same apart from the vocal, so when you invert one of them it now has the opposite values of the instrumental track. So when you sum them together, you are left with the difference (the vocal) ;-)

Values? This is pretty intriguing to me. Care to elaborate? :)
A waveform has a positive, or negative Value at any point. If you took a sinewave, and picked a point above the zero-crossing it will have a positive value, vise versa.

If you was to make all the positive points negative points, vise versa. When you sum them together, they would cancel out.
E.g. 1 + -1 + = 0

If there was a difference, you would be left what what the difference.
E.g. 2 + -1 = 1

That 1 you got is part of the vocal :)

Re: Phasing audio

Posted: Sun Jun 12, 2011 2:34 am
by zerbaman
efence wrote:
zerbaman wrote:What I want is to end with an acapella of the track, using the instrumental and the full version to phase eachother out.
3za wrote:Because all the values are the same apart from the vocal, so when you invert one of them it now has the opposite values of the instrumental track. So when you sum them together, you are left with the difference (the vocal)
Yeah, but why is it that on some files it doesn't work out this way? Nothing changes, they just play as normal
the invert needs to line up perfect bit to bit
Self explanatory isn't it?

Re: Phasing audio

Posted: Sun Jun 12, 2011 2:41 am
by Toric
3za wrote:
Toric wrote:
3za wrote:Because all the values are the same apart from the vocal, so when you invert one of them it now has the opposite values of the instrumental track. So when you sum them together, you are left with the difference (the vocal) ;-)

Values? This is pretty intriguing to me. Care to elaborate? :)
A waveform has a positive, or negative Value at any point. If you took a sinewave, and picked a point above the zero-crossing it will have a positive value, vise versa.

If you was to make all you the positive points negative points, vise versa. When you sum them together, they would cancel out.
E.g. 1 + -1 + = 0

If there was a difference, you would be left what what the difference.
E.g. 2 + -1 = 1

That 1 you got is part of the vocal :)
Stuff Eric knows, most likely. Thanks for the explanation! Why do vocals have a greater value? Just curious and I think may be relevant to the thread?

Edit: is it because you are using an instrumental and then an instrumental + vocal track and that's what makes the vocal track have a higher number? Or does it just cancel everything out but the vocals?

-T

Re: Phasing audio

Posted: Sun Jun 12, 2011 2:46 am
by 3za
Toric wrote:Why do vocals have a greater value? Just curious and I think may be relevant to the thread?
That was just in my example, they will have negative values too.

I was just talking about the value of a waveform at just one point, in your audio file their are billions, depending on your sample rate, and lengtth of the file.

Re: Phasing audio

Posted: Sun Jun 12, 2011 3:20 am
by Toric
3za wrote:
Toric wrote:Why do vocals have a greater value? Just curious and I think may be relevant to the thread?
That was just in my example, they will have negative values too.

I was just talking about the value of a waveform at just one point, in your audio file their are billions, depending on your sample rate, and lengtth of the file.

Yea. I think I've gotcha. It's really hard to communicate this stuff the way you want to via text.

Which is why I have music theory textbooks.

Thanks!

Also, thanks Zerbaman, for lending me your thread for a second! :)

-T

Re: Phasing audio

Posted: Mon Jun 13, 2011 8:56 am
by Depone
Forget the vocals supposed number, that doesnt make sense. Irs just straight up phase inversion. Ty it yourself with 2 copies of the same audio. Put them both in your daw, invert one and you are left with silence. If one had a vocal or another element, it would stay as theres nothing to 'counter phase' it, and your left with just that part

Re: Phasing audio

Posted: Mon Jun 13, 2011 2:35 pm
by 3za
Depone wrote:Forget the vocals supposed number, that doesnt make sense.
I am sure everything I posted in this thread makes sense.

Re: Phasing audio

Posted: Mon Jun 13, 2011 2:52 pm
by -[2]DAY_-
Toric wrote: Why do vocals have a greater value?
:lol: :lol: :cornlol: :lol: :lol:

naw, its not literally an arithmetic problem.. if you invert phase of the instrumental and layer with the full track, only the instrumental part of the waveform will cancel out, leaving the vocal left behind by itself (hopefully).


lol i'm sure plenty people have tried explaining it. Doesn't even work too well usually, afaik

Re: Phasing audio

Posted: Tue Jun 14, 2011 2:28 am
by Toric
-[2]DAY_- wrote:
Toric wrote: Why do vocals have a greater value?
:lol: :lol: :cornlol: :lol: :lol:

naw, its not literally an arithmetic problem.. if you invert phase of the instrumental and layer with the full track, only the instrumental part of the waveform will cancel out, leaving the vocal left behind by itself (hopefully).


lol i'm sure plenty people have tried explaining it. Doesn't even work too well usually, afaik
Yea! I know. I was a bit blitzed out of my mind when I was on here. I was actually looking for some info and found this thread (completely irrelivant) and I confused myself by taking things literally and not having a better DAW & Electronic music vocabulary.

:P
:corndance: :corndance: :corndance: