Old 01-11-2019, 02:17 AM   #1
Masi
Human being with feelings
 
Join Date: Sep 2015
Posts: 690
Default correlation vs width

I've seen plugins metering both correlation and width. Is there any difference between them other than the scale thy are using?

This is how I read their respective scales.

Width:
0% to 100%, 0% = mono, values above 100% mean a signal that is out of phase

Correlation:
0 to 1.0, 0 = mono, negative values show a signal that is out of phase

Masi
Masi is offline   Reply With Quote
Old 01-12-2019, 03:58 PM   #2
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 3,409
Default

Quote:
Originally Posted by Masi View Post
I've seen plugins metering both correlation and width. Is there any difference between them other than the scale thy are using?
Not that I'm aware of. I interpret them the same way you do.
clepsydrae is offline   Reply With Quote
Old 01-13-2019, 02:07 AM   #3
ErBird
Human being with feelings
 
Join Date: Jan 2017
Location: Los Angeles
Posts: 1,161
Default

Quote:
Originally Posted by Masi View Post
Correlation:
0 = mono
1 = mono, no? Everything else is correct.
ErBird is offline   Reply With Quote
Old 01-13-2019, 07:42 AM   #4
Colox
Human being with feelings
 
Join Date: Feb 2012
Location: Sweden
Posts: 1,206
Default

Phase is a subject that can't be explained in a few sentences, and the question you're asking won't get you all the way there.

But essentially, I should think both the correlation meter and the "width" meter that you're referring to, are using roughly the same technical process to derive the value they display. Roughly.

The difference in the meters you're referring to seems to be that one of them are oversimplifying things to the point like sounding like a kids toy; using the word "width" and using a scale of 0-100.

The phase relation between two streams of audio separated by pan affects a heck of alot more than simply perception of width. And the best way to relate to matters of phase is to learn how to use it adequately, and from that use tools that shows you values and curves in the right contexts. Correlation is a context which takes into account the current phase angle difference between two waveforms.

The perception of width requires many things to co-exist and work together in certain ways. A certain phase relation is just one of those factors, but there are many others wich are required. It's the field of psychoacoustics too.

But the meter you're talking about, that one that says "width 0-100", shows only phase difference between two channels (I bet it does) yet equates this value as the same thing as pure perception of width. That way of interpreting things, is misleading.

Long and detailed response to a simple question. Hope it clarifies some things.
__________________
There are only two kinds of people in the world: those who entertain, and those who are absurd.
- Britney Spears

Last edited by Colox; 01-13-2019 at 06:51 PM.
Colox is offline   Reply With Quote
Old 01-13-2019, 10:08 AM   #5
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

I have never been able to figure out what that kind of meter is even for. You'll hear the width in stereo, and you'll hear any phase issues if you sum to mono. That's exactly all that actually matters.

Honestly, though, if you just stay away from processes that deliberately mess with phase for hyperreal stereoness, you'll never have to worry about. OTOH, if you're trying to do psychoacoustic tricks for people on headphones then you just accept the fact that it won't translate properly to any other playback system and very well could mess it all up and go with that.
ashcat_lt is offline   Reply With Quote
Old 01-13-2019, 11:10 AM   #6
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 3,409
Default

Quote:
Originally Posted by ErBird View Post
1 = mono, no? Everything else is correct.
Oops, yeah: 1 = mono/in-phase, 0 = no correlation, -1 = antiphase.

Quote:
Originally Posted by ashcat_lt View Post
I have never been able to figure out what that kind of meter is even for.
Agreed that they aren't very useful, except for catching antiphase stuff like flipped or poorly-positioned mics, etc., which sometimes isn't immediately noticeable.

@Colox -- agreed that "width" is the more psychoacoustic and non-technical of the two.
clepsydrae is offline   Reply With Quote
Old 01-13-2019, 12:48 PM   #7
ErBird
Human being with feelings
 
Join Date: Jan 2017
Location: Los Angeles
Posts: 1,161
Default

I wanted to visualize this, and I'm a bit math-obsessed, so I made a little visual aid using the Desmos graphing tool. You may find it useful.

https://www.desmos.com/calculator/xwmd72rzys




Last edited by ErBird; 01-13-2019 at 02:30 PM.
ErBird is offline   Reply With Quote
Old 01-13-2019, 12:55 PM   #8
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 3,409
Default

Quote:
Originally Posted by ErBird View Post
I wanted to visualize this, and I'm a bit math-obsessed, so I made a little visual aid using the Desmos graphing tool.

Desmos is awesome. :-)
clepsydrae is offline   Reply With Quote
Old 01-13-2019, 01:09 PM   #9
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Quote:
Originally Posted by ErBird View Post
I wanted to visualize this, and I'm a bit math-obsessed, so I made a little visual aid using the Desmos graphing tool. You may find it useful.

<snip>
Could you elaborate a little on this? What does it tell us?

Just yesterday for some reason I looked into Span's correlation meter and tried to understand it. Googling I found Izhaki's (and I assume this is Roey Izhaki) reply here https://dsp.stackexchange.com/questi...al-phase-meter interesting, but I'm not sure I got it all.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 01-13-2019, 01:34 PM   #10
ErBird
Human being with feelings
 
Join Date: Jan 2017
Location: Los Angeles
Posts: 1,161
Default

Quote:
Originally Posted by clepsydrae View Post
Desmos is awesome. :-)
Indeed. It's incredibly useful. It's awesome for messing with waveshapers.

Quote:
Originally Posted by Fabian View Post
Could you elaborate a little on this? What does it tell us?
Sure. So assuming a simple sine wave input to both channels, you're able to adjust the relative phase angle of the right channel, in degrees, from 0=mono to 180=out-of-phase. Change this value by dragging the slider below p=0. Purple is simply the sum of the left and right channels, reduced in amplitude by half, which is customary when merging channels. At zero phase angle there is no cancellation and the correlation bar is at +1. At p=90 the merged signal is reduced by -3 dB and the correlation bar is at zero. This is the maximum allowable phase difference between channels without the channels being considered out of phase to some degree. Between p=90 and p=180 the channels cancel out more and more until cancelling fully when p=180 and correlation = -1. The same concept applies to complex stereo signals, it just wouldn't be easy to visualize.

I also don't care for the "width" type meters.
ErBird is offline   Reply With Quote
Old 01-13-2019, 01:55 PM   #11
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Quote:
Originally Posted by ErBird View Post
Indeed. It's incredibly useful. It's awesome for messing with waveshapers.



Sure. So assuming a simple sine wave input to both channels, you're able to adjust the relative phase angle of the right channel, in degrees, from 0=mono to 180=out-of-phase. Change this value by dragging the slider below p=0. Purple is simply the sum of the left and right channels, reduced in amplitude by half, which is customary when merging channels. At zero phase angle there is no cancellation and the correlation bar is at +1. At p=90 the merged signal is reduced by -3 dB and the correlation bar is at zero. This is the maximum allowable phase difference between channels without the channels being considered out of phase to some degree. Between p=90 and p=180 the channels cancel out more and more until cancelling fully when p=180 and correlation = -1. The same concept applies to complex stereo signals, it just wouldn't be easy to visualize.

I also don't care for the "width" type meters.
Ah! It was an interactive plot. Nice! Really cool. Thanks.
Now I'll play around with that, see if I can relate it to the stackexchange post I linked to.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 01-14-2019, 03:08 PM   #12
ErBird
Human being with feelings
 
Join Date: Jan 2017
Location: Los Angeles
Posts: 1,161
Default

Quote:
Originally Posted by Fabian View Post
Ah! It was an interactive plot. Nice! Really cool. Thanks.
Now I'll play around with that, see if I can relate it to the stackexchange post I linked to.
Thanks. Updated: https://www.desmos.com/calculator/6xwmy3cwba



Last edited by ErBird; 01-15-2019 at 02:57 PM.
ErBird is offline   Reply With Quote
Old 01-14-2019, 03:42 PM   #13
bezusheist
Human being with feelings
 
bezusheist's Avatar
 
Join Date: Nov 2010
Location: Mullet
Posts: 829
Default

i don't know how useful it could actually be, but you can output correlation from a JSFX and then view it on a scope or even render the file and view that way...
here's an example of one i hacked together...(MSED is there for reference)
__________________
I like turtles
bezusheist is offline   Reply With Quote
Old 01-15-2019, 11:21 AM   #14
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Quote:
Originally Posted by bezusheist View Post
i don't know how useful it could actually be, but you can output correlation from a JSFX and then view it on a scope or even render the file and view that way...
here's an example of one i hacked together...(MSED is there for reference)
And the example...? Seems you forgot to include it.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 01-15-2019, 01:10 PM   #15
bezusheist
Human being with feelings
 
bezusheist's Avatar
 
Join Date: Nov 2010
Location: Mullet
Posts: 829
Default

that IS the example...:P

i took the equation posted here > https://dsp.stackexchange.com/questi...relation-meter
...and hacked together a JS with my limited knowledge.
i attached it, as is, no warranty, etc...
(i'm not sure if it's "correct", but it "syncs" with MSED)

edit: output is channel 3, you can change it by changing “spl2”.
also forgot to assign output pins, so you have to do it manually.
Attached Files
File Type: txt corrmeter.txt (387 Bytes, 118 views)
__________________
I like turtles

Last edited by bezusheist; 01-15-2019 at 02:15 PM.
bezusheist is offline   Reply With Quote
Old 01-15-2019, 01:55 PM   #16
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Quote:
Originally Posted by bezusheist View Post
that IS the example...:P

i took the equation posted here > https://dsp.stackexchange.com/questi...relation-meter
...and hacked together a JS with my limited knowledge.
i attached it, as is, no warranty, etc...
(i'm not sure if it's "correct", but it "syncs" with MSED)
Looks good.
Thanks.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 01-15-2019, 01:55 PM   #17
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Quote:
Originally Posted by ErBird View Post
Even better
Thanks.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 01-20-2019, 03:39 AM   #18
Masi
Human being with feelings
 
Join Date: Sep 2015
Posts: 690
Default

Thank you guys for the detailed explanation of correlation.

Masi
Masi is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 11:39 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.