Possible to automate MS mic decoding in Audacity?

Hi all,

New poster so hoping in right section and thanks for having me.

Doing some AV work at the moment and, cutting a long story short, syncing audio recorded on Zoom H6 with camera audio by loading both in to Audacity and fine tuning from my cue on both audio and video made on the day (me saying ‘camera speed’ as the Zoom is often already on).

I’m using an MS (mid-side) mic on the H6 and, once decoded, sounds great. Audacity is the perfect tool for decoding the MS tracks, and herein lies my question.

With an MS recording you get basically a stereo track, but to get the full ‘3D audio picture’ you need to decode the stereo track using these steps:

1/ Split stereo track;
2/ Copy the second track;
3/ Invert the copied track;
4/ Change the second track to ‘left’ (and the copied track is right);
5/ Change the first track to ‘mono’;
6/ Make tracks two and three ‘Make stereo track’.

This leaves you with three tracks which I’m then bumping out to a stereo flac. The decoding of each track is not a lengthy process, but repetitive and boring after you’ve done a dozen or so (and I have about a hundred short audio recordings or more) which leads to my question(s):

  • Is there any way I can automate these steps in Audacity?
  • Is it possible to write a script (not that I’ve got the nous to do so but I’ve got the nous to research how) which will do this for me?
  • Is there a tool more suitable to automate this process than Audacity?

Tnx in advance for any help you can give and I’ll finish off with this. I started using the wonderful Audacity about nine years ago and it has been a good friend of mine ever since. Love it and thanks untold to the folks who develop(ed) and maintain this audio Swiss Army knife. Excellent work!

(PS: Before anyone suggests it, I’m aware Zoom does have a decoder that does this in a jiffy from what I can see, but I only have Win on my laptop which is not setup for this and my studio desktop box does run Windows I seriously prefer not to use Windows for this as it would involve a lot of booting and rebooting from Win to Linux and this interrupts workflow and I am in a bit of a rush as it is (while it’s a hassle manually decoding every take, it is faster than rebooting and doable).

Then you run out with the clapboard:

“Camera Mark!” [bang]

Audacity’s batch processing tools is called Chains.

http://manual.audacityteam.org/o/man/chains_for_batch_processing_and_effects_automation.html

Koz

Try the “Channel Mixer” plug-in: http://wiki.audacityteam.org/wiki/Nyquist_Effect_Plug-ins#Channel_Mixer

LOL. If only, that would be a luxury, but bit hard when I have an action cam in one hand and the Zoom H6 in the other! Basically, the Zoom is rolling, hit record on the cam with my thumb, hold them both together and ‘camera speed’. I wish there was I could get a visual cue in there. I suppose I could do something with a foot. :-k

Thanks for the tips/clues/leads. I’ll look into shortly (about to get to work on the rushes on that box as not in the field today).

PS: Forgot to mention that I am using Xubuntu-core 15.10 (very new hardware or would be an LTS). Also, maybe dumb question, but is it possible to edit a post once I’ve posted it? Maybe my blindness to some things obvious, but I don’t see an ‘Edit’ button …

Brilliant and thanks to you both. Chains look great and will delve into that further in the near future, but for now, the Channels plug-in has a ‘MS Decode’ default! Just installed the plug-in and tried it out. Worked perfectly, but probably too perfectly. When I do this manually I end up with a stereo mono track (the front mic) and a stereo track (the side mic l/r). Three tracks.

The volumes can then be tweaked if necessary and export the final result to FLAC stereo. The Channels default ‘MS decode’ goes a step further than this and leaves me with the bumped out stereo track only which is probably exactly the same result as I would get if I exported decoded and then exported the three tracks to a stereo track.

This is no big deal as I think I’ve only needed to tweak one take out of about twenty, so if it is necessary in future, I’ll do that specific take manually. In the meantime, you’ve saved me some time and I thank you. Channels and ‘MS Decode’ was pretty much was I was looking for, and a lot simpler than what I expected to find. :slight_smile:

PS: Is there a way I can see exactly what the default MS Decode is doing, the steps it is taking to come up with the final stereo track? Just curious … :-k

I think after the forum lets you post without an elf helping you. Your posts are moderated now, until the software thinks you’re probably not going to try and sell us Male Enhancing Products.

You’re not, are you?

We would rather you don’t go back and change earlier posts. Given the forum workflow, nobody will see the changes and it will just cause link confusion later.

Koz

Thanks for that, Koz. I’m actually a mod on the Ubuntu Forums so no, I won’t be attempting to sell any elongation implements (roaring trade on the other forum already, LOL!)/

If I think of a relevant bit of info after I’ve posted, i’ll simply plop another post with that gem contained. :slight_smile:

Thanks again.

For instance, I just noticed a terrible and misleading statement in a previous post of mine. Must have been excited. Let me clarify.

When I decode the MS recording manually I end up with one MONO track (the front mic) and a stereo track (sides).

FTR. :wink:

I think I partly answered one of my earlier questions regarding what Channel-mixer.ny is doing. From the file:

(15 (list 50 50 50 -50)) ; Mid-Side Decode
	  (T (list left-mix-L left-mix-R right-mix-L right-mix-R))))

Not that I’ve properly got my head around that line.

I grabbed an MS file, decoded manually and saved, repeated that proces except using channel-mixer.ny and the two sound quite different. Mine somehow has more ‘depth’. Hard to describe. The channel-mixer.ny is thinner and quieter (I tried normalising but ended up amplifying). While you can hear sound all around in both, it seems as if the side-mic is turned down with the plug-in and more focus is on the front mic.

Might be tweakable with the plug-in GUI, but for mine, balancing front and side should really happen post decoding rather than as part of it. These are fieldwork recordings and I don’t need to touch the balance of the two once out of the H6 or before, during or after decoding at all, except on the very rare occasion, so any changes to it are not good.

Think I’ll be sticking to the manual approach after all, but I’ll keep playing around and thanks for the tips. I’d never even seen the plug-ins before and that gives me a whole lot more to explore. Chains next, but that could be beginning of next month the way things look from here. Audacity just got even better! :slight_smile:

If I recall correctly, if you run the effect using the “Debug” button, then the exact settings used are displayed in a “debug window” that pops up when the processing is complete. That will give you a starting point for manually tweaking the mix. Note that you can increase the preview length in Audacity’s preferences: Audacity Manual

thanks