Taperssection.com

Gear / Technical Help => Post-Processing, Computer / Streaming / Internet Devices & Related Activity => Topic started by: pyiteac on December 19, 2008, 02:17:14 AM

Title: M/S Decoding in Audacity?
Post by: pyiteac on December 19, 2008, 02:17:14 AM
I just picked up a VMS02 and have access to an MK8.  I want to start using the M/S technique.  I know there are VST Plug-ins for Audacity that allow M/S decoding.  What ones are the best and how difficult are they to set up and use?  Thanks for the input!
Title: Re: M/S Decoding in Audacity?
Post by: boojum on December 19, 2008, 02:35:22 AM
Not in Audacity, but the VST plugin I use is Voxengo.  it is easy to use and does an inline decode so that you can test various ratios before you do the whole file to a new M-S ratio.  The price is right, too: zero.    8)

http://www.voxengo.com/product/msed/
Title: Re: M/S Decoding in Audacity?
Post by: page on December 20, 2008, 12:44:35 AM
And if your using Audacity on a system that isn't windows (and/or can't handle VST plugins correctly), then you can do the decoding by taking your two track recording;

1) duplicating both tracks (so now 2 M and 2 S)
2) setting all 4 tracks to mono
3) panning the respective right/left pairs of M and S
4) Invert the Right S channel.

Then reduce/increase the S db levels to taste and hit mix/render when you've got something your happy with.

I run OSX and VST doesn't work in audacity, so this is how I do it.
Title: Re: M/S Decoding in Audacity?
Post by: Roving Sign on December 20, 2008, 07:51:03 AM
And if your using Audacity on a system that isn't windows (and/or can't handle VST plugins correctly), then you can do the decoding by taking your two track recording;

1) duplicating both tracks (so now 2 M and 2 S)
2) setting all 4 tracks to mono
3) panning the respective right/left pairs of M and S
4) Invert the Right S channel.

Then reduce/increase the S db levels to taste and hit mix/render when you've got something your happy with.

I run OSX and VST doesn't work in audacity, so this is how I do it.

Something about that doesnt seem right - why do you need 4 tracks?

Assuming you start with two channels..."Split Stereo Track", duplicate and invert the S channel...now you should have 3 mono channels. At this point you should "Make stereo" the your two S channels...leave the M mono...

Seems like just you would just adjust the amount of M in the mix and export (the three tracks will get summed to L/R)
Title: Re: M/S Decoding in Audacity?
Post by: page on December 20, 2008, 12:21:13 PM
Something about that doesnt seem right - why do you need 4 tracks?

Assuming you start with two channels..."Split Stereo Track", duplicate and invert the S channel...now you should have 3 mono channels. At this point you should "Make stereo" the your two S channels...leave the M mono...

Seems like just you would just adjust the amount of M in the mix and export (the three tracks will get summed to L/R)

Ultimately it's the same net effect. I've tried it both ways and my personal preference is I find it easier to keep track of everything when I have 4 channels instead of just the three. One difference is that with independent channels on the sides (non-stereo linked in audacity), you can swing the image around a little in post, rather helpful for on stage/stage lip stuff where the balance is off.
Title: Re: M/S Decoding in Audacity?
Post by: rjp on December 31, 2008, 09:41:38 PM
Audacity can use LADSPA plugins (http://audacity.sourceforge.net/download/plugins). There is a Windows installer for the LADSPA plugin package on Audacity's site, and it includes M/S conversions.

NOTE: If you're running a beta (1.3.x) version, you will need to tell the installer where your Audacity installation lives. The installer assumes you're running a stable version of Audacity.

You can convert M/S to L/R or vice versa; just make a selection and run the plugin. In 1.3.6 it shows up in the Effect -> Unsorted menu, which you will need to scroll to find the Matrix plugins.

I did some playing around with one of my LS-10 internal mic recordings, and I found that I could greatly improve the stereo image by first knocking the level back -6 dB to get some headroom, converting to M/S, boosting the side channel +6 dB, converting back to L/R, and normalizing it back to maximum.

I'm pretty certain that the LS-10's internal "mic zooming" feature works in a similar way on-the-fly, but that only works in 16-bit mode, so I've never used it.