Stereo to mono plugin wanted

[t.find]

Its worth a shot: http://destroyfx.smartelectronix.com/extras/

DFX monomaker

[Ah_Dziz]

The side portion of a m/s audio stream is by definition the part which will cancel when summed. Panning tools aren't "throwing out" the side portion of the signal when summing to mono. It disappears on it's own.

[docdued]

I know what you mean, but the standard definition is indeed mid = L+R. What you would like to have is the "common signal" (let's call it c) that L and R share. So we'd have L = c+l and R= c+r, with l and r being the residual signals that would only be present in L and R, respectively. But there's no way to resolve this set of equations, there's just one variable too much... This amounts to the fact that (at least to my knowledge) it's impossible to precisely isolate the "center signal" in the stereo field. Which is kind of weird, because "in our head" we can do that (e.g. we hear that there's only the drums and the bass sitting exactly in the center of a stereo field recording).

[s_t]

...it's impossible to precisely isolate the "center signal" in the stereo field. Which is kind of weird, because "in our head" we can do that (e.g. we hear that there's only the drums and the bass sitting exactly in the center of a stereo field recording).

Long ago there was a plugin that achieved this. If I remember correctly, it used FFT processing and suffered from some artifacts because of that, but it worked.
Unfortunately I can't remember what was the name of the plugin and who made it.
cheers

PS: Maybe our brains use FFT

[Harry_HH]

I know what you mean, but the standard definition is indeed mid = L+R.)

Can someone give me a link to a "white paper" (or any theoretical paper where this is clearly defined). I found in the net a lot of different articles and manuals, but none of these defined the relation between L/R and M/S
clearly (mathematically) enough, to my mind. The basic difficulty to convert stereo to mono AND preserve all the stereo information (human brains perceive) is the fact that opposite phases which can be perceived in the stereo picture cancel each other in the mono. However, my target and my original question in this thread was: how to convert stereo file to mono and preserve all the information.
gavriloP has given one practical advice (use the MSED plugin), this seems to work but didn't convince me totally - haven't tested yet the t.find's DFX monomaker plugin. I know this may be splitting the hair but it's not totally out of relevance because stereo-to-mono conversions are done all the time without thinking the end result at all. H.

[docdued]

...it's impossible to precisely isolate the "center signal" in the stereo field. Which is kind of weird, because "in our head" we can do that (e.g. we hear that there's only the drums and the bass sitting exactly in the center of a stereo field recording).

Long ago there was a plugin that achieved this. If I remember correctly, it used FFT processing and suffered from some artifacts because of that, but it worked.
Unfortunately I can't remember what was the name of the plugin and who made it.
cheers

PS: Maybe our brains use FFT

Alright, you could probably do it via spectral subtraction indeed... but not via mere addition or subtraction of the two stereo signals. Might be worth a shot though...

[helium]

I've developed audio related stuff for quite some time and I can assure you that the standard way for MS-coding is M = L+R, S = L-R (ignoring possible constant factors).

You probably wont find a paper specifically about MS encoding because its way too simple. But maybe you are happy enough with Wikipedia or maybe the FLAC audio compression definition just mentioning it becasue I just read it).

[Dean Aka Nekro]

Looks ROPEY to me, I use some of the processors/encoder/decoder plugins mentioned already along with alot of the brainworx plugin which made microphone setups with a mid/side setup so damned easy to encode/decode in one instead of fiddling with multiple channels alot on decent to plain crummy desks myself, Never seen this one before as said seems ROPEY but I guess you have nothing to lose by trying it out:

http://help.stereotool.com/presets.shtml
Homepage: http://www.stereotool.com/

This one is the one that made life alot easier:
http://www.plugin-alliance.com/en/newsr ... -talk.html (Not M/S but still intresting reading FWIW)

http://www.brainworx-music.de/en/plugins/bx_control_v2 (version 1) plus bx_solo. The new stereo upmix one seems intresting:
http://www.plugin-alliance.com/en/plugi ... maker.html (gonna try it when I have time to demo properly), Good for information here also:

http://www.plugin-alliance.com/en/newsr ... -this.html

Harry_HH some papers on sum/difference:

http://www.safesoundaudio.co.uk/PDFs/M- ... -Paper.pdf
http://www.soundonsound.com/sos/nov10/a ... essing.htm
http://www.soundonsound.com/sos/mar11/a ... h-0311.htm
> Intresting one here: http://www.phmusic.co.uk/pharri5833/m-what.htm <
http://courses.washington.edu/psych333/ ... alysis.pdf
http://www.paia.com/ProdArticles/msmicwrk.htm
http://www.aes.org/e-lib/browse.cfm?elib=5246
http://www.df.unipi.it/~fiig/moehrs/pub ... munich.pdf
http://www.google.co.uk/url?sa=t&rct=j& ... dEwqqybdfA
http://www.lectrosonics.com/White-Paper ... ixing.html
http://www.academia.edu/693259/An_Inves ... plications
http://eandt.theiet.org/news/2012/jun/b ... rdings.cfm (Actually googling 'Alan Blumlein' might be worth a shot )
http://www.tvtechnology.com/audio/0098/ ... fun/212158
http://www.tvtechnology.com/audio/0098/ ... reo/211144
http://www.grimmaudio.com/whitepapers/speakers.pdf
http://www.paia.com/ProdArticles/msmicwrk.htm
http://designingsound.org/2011/08/tim-n ... recording/
http://www.proaudiodesignforum.com/foru ... &start=193

Hope you find some meat to get stuck into in that lot (Searched for 'Sum Difference, Mid Side and Lateral Vertical Audio Processing/Microhpone Techniques FWIW)

Dean

[BertKoor]

The theory is rather simple.

Subtract left from right (or vice versa) and that is the side. Subtract that from the summed left & right, and what remains is the shared middle.

The matrix below shows the results of this calculation for a signal that gets panned from left to mid to right, and finally both anti-phase situations:

Code: Select all

  L    R  | Sum  Side  Mid
 1.0  0.0 | 1.0  -1.0  0.0
 0.5  0.5 | 1.0   0.0 +1.0
 0.0  1.0 | 1.0  +1.0  0.0
 0.5 -0.5 | 0.0  -1.0 +1.0
-0.5  0.5 | 0.0  +1.0 -1.0

[docdued]

No, it's not. Mid = L+R, Side = R-L. Then Mid-Side=2*L, twice the left channel. There are one or two sign errors in your matrix, btw.

[Harry_HH]

well, at least this discussion (see e.g. the 4 latest posts)
shows that the nature of "stereo" or "mono" or "L/R vs. M/S"
is far from clear, for many of us. And this is very basic
audio processing happenging every day in thousands of studios/DAWs'. H.

[hvz]

Here's how to do it - even if you feed it audio in anti-phase it will keep all of it:

- Download the Stereo Tool VST plugin from http://www.stereotool.com/
- Click on RESET to turn all the processing off
- Open the 'Stereo Image' window, enable "ON" on top and "Stereo Image / Flat Stereo" at the bottom. Then move the sliders 'phase' and 'width' both to 0.

(There used to be a preset for this, but it was removed to make room for other presets a long time ago).

'Phase' removes stereo phase, but keeps channel separation.
'Width' removes channel separation but keeps phase info

Using either of them separately can give some artifacts, but if you use both there are nearly none - the only situation where I found some mild artifacts is when the phasing was so bad that a normal (L+R) conversion to mono was completely unlistenable.

More info: http://help.stereotool.com/stereo_image ... ator.shtml

[BertKoor]

Mid = L+R

Definitely incorrect, that's just the sum of both. A hard-panned left or right signal does not belong in the Mid channel, it's in the Side only.

Then Mid-Side=2*L, twice the left channel. There are one or two sign errors in your matrix, btw.

Thanks for noticing this. But you didn't bother to provide the correct formulas.
The correct way to get the mid channel, is to take absolute values of both sum and side, subtract side from the sum, and then multiply it with the sign of the sum (that is: -1 when negative, otherwise +1)

Now calculated with Excel, as above plus all possible combinations.

Code: Select all

  L    R   |  Sum   Side   Mid
 1,0  0,0  |  1,0   -1,0   0,0
 0,5  0,5  |  1,0    0,0   1,0
 0,0  1,0  |  1,0    1,0   0,0
 0,5 -0,5  |  0,0   -1,0  -1,0
-0,5  0,5  |  0,0    1,0  -1,0
           |  
-1,0 -1,0  | -2,0    0,0  -2,0
-0,5 -1,0  | -1,5   -0,5  -1,0
 0,0 -1,0  | -1,0   -1,0   0,0
 0,5 -1,0  | -0,5   -1,5   1,0
 1,0 -1,0  |  0,0   -2,0  -2,0
-1,0 -0,5  | -1,5    0,5  -1,0
-0,5 -0,5  | -1,0    0,0  -1,0
 0,0 -0,5  | -0,5   -0,5   0,0
 0,5 -0,5  |  0,0   -1,0  -1,0
 1,0 -0,5  |  0,5   -1,5  -1,0
-1,0  0,0  | -1,0    1,0   0,0
-0,5  0,0  | -0,5    0,5   0,0
 0,0  0,0  |  0,0    0,0   0,0
 0,5  0,0  |  0,5   -0,5   0,0
 1,0  0,0  |  1,0   -1,0   0,0
-1,0  0,5  | -0,5    1,5   1,0
-0,5  0,5  |  0,0    1,0  -1,0
 0,0  0,5  |  0,5    0,5   0,0
 0,5  0,5  |  1,0    0,0   1,0
 1,0  0,5  |  1,5   -0,5   1,0
-1,0  1,0  |  0,0    2,0  -2,0
-0,5  1,0  |  0,5    1,5  -1,0 
 0,0  1,0  |  1,0    1,0   0,0
 0,5  1,0  |  1,5    0,5   1,0
 1,0  1,0  |  2,0    0,0   2,0

You see that some scaling (divide by 2) is needed to stay within the +1 / -1 range.

[docdued]

That's gonna distort quite a bit, because of the jumps in sign when the sum crosses zero. Really, I don't think there's a valid approach that doesn't use FFT...

[helium]

@BertKoor: Your idea is nonsense, sorry.

Here are two sine waves left and right with a slight delay between them (shown in red and blue in the top row) and your "mid" signal (shown in black on the bottom row):


As you can see this doesn't look like a sine any more. Your way of mixing will cause heavy distortions.