[linux-audio-user] Should Stereo to MS then MS to Stereo be a no-op (identity)

[Steve Fosdick]

I just tried taking a stereo track in ardour that was recorded with a X/Y pair and add two pugins to the track: a Stereo to Mid-Side(MS) matrix and a Mid-Side to Stereo matrix.  The idea was that the MS to Stereo matrix gives me a nice simple control to control the width of the stereo image.

What I discovered, to my surprise, was that with the width set to 2.0, the widest setting, the stereo image was still narrower than with the pair of plugins bypassed.

Can anyone think why this should be?

Steve