Is the mid-side decoder feature in DoM a unique function?

[SY.Cha]

Thank you for creating great software like DoM.

I'm currently working on creating DCP for an indie film festival and have a question.

Is the mid-side decoder feature in DoM a unique function? (There seem to be plugins for Mid-Side, but it's not clear if there are any that function identically to DoM's Mid-Side decoder.)

I've read the manual and understand the operation of the mid-side decoder, where the C channel is created using the L and R channels.

Is there a way to perform this process in other editing tools or programs (such as FCPX, DaVinci, etc.) to create a Center channel track? Are there any plugins for this?

Alternatively, is there a way to use DoM only for the mid-side decoder and extract the audio track?

I would greatly appreciate any ideas or suggestions regarding this issue.

[carl]

To be honest the mid-side decoder is a bit of a mystery. I added it a long time ago, with no idea that it would still be there 10 years later, and recently somebody pointed out that it doesn't do what I always imagined it did.

I think my intention was to take a stereo signal and extract the common parts from it (i.e. (L+R)/2, kind of the "mid" part of a mid-side signal) and then put that in the centre speaker, with the intention to improve the sound compared to having just L and R with a "phantom" centre.

I never did any proper listening tests myself (in a cinema) though some people do seem to use and like it. It's certainly not any revolutionary, unique idea, and may even make things worse!

I think if you wanted to do something like it in some other editor/DAW the simple way would be to sum left and right, and reduce the gain by 3dB.

[Carsten]

I think it's working nice enough in it's current version, but guess it needs a tweak. I can't remember wether I filed a bug on Mantis. I think Mid-Side decoders use a very basic processing that works for most typical material - they are intended to direkt voice/dialog to the center speaker, and in typical mixes, that will work without more advanced processing options.

The flaw currently seems to be that there is a 180 degree phase shift between left and right after processing. Which in my opinion is not necessary for the decoder to work. I once did a basic analysis in audacity to prove that.

Maybe there is also some confusion with the M/S stereo coding system? https://en.wikipedia.org/wiki/Joint_encoding


- Carsten