Gear / Technical Help > Ask The Tapers

How can I prevent recordings with vocals in one channel and band in the other?

(1/12) > >>

Sometimes when stealth taping (with cards clipped to shirt collar), I end up with very "unbalanced" recordings.  For example, with drums/guitar/bass heavily favoring the left channel and vocals almost entirely in the right channel.  I assume this is the result of the sound reaching my left mic before the right one, and with the vocals (in this case coming off the right channel PA monitor) reaching my right mic first.  This results in a mess that I cannot correct.  If all instruments favored a single channel, that's easy enough to correct by delaying the favored channel by few samples, thereby centering everything in the middle of the sound field. The room in question has the stage set up along the long wall of the rectangular shaped room.  The left and right channel PA monitors are therefore some distance apart.  The room could really use a center fill channel, but I'm not holding my breath that will ever happen. 

Sometimes in this room, I make a nice, well balanced recording, but other times, I end up with an unbalanced mess like this.  Is there anything I can do to prevent this?  I try to stand in the dead center of the room, as far back as possible.  Should I try to orient my card mics so they're both facing straight ahead, or should I try to toe them outward in a more ORTF-like configuration? 

Any other suggestions?  I don't suppose there is anything I can do in post??? 

I'm trying to picture the geometry here: You're centered in the room, but the vocals are reaching you from the right PA well before the left?

Less directional mics or a more parallel arrangement of directional mics might help.

The Izotope RX remix algorithms are getting better all the time, but are only in the advanced version and so rather costly.  They aim to isolate vocals and specific instruments from a mix so that you can do whatever you like with each of them, such as re-panning their stereo image placement. I've not used them and am not sure if their quality is up to this task yet, but there is good potential here if the latest routines are able to avoid producing objectionable artifacts.

You might try a simpler stereo image rebalancing tool that splits content into a few frequency ranges and provides controls to modify the stereo width, center balance and panning of each range independent of each other.  If you are able to isolate the primary vocal mid and upper midrange material sufficiently in one control band you might be able to shift it to a more centered position.  That will effect all content in that frequency range though so tread lightly, its easy to make even more of a mess of things with those kinds of tools.

If nothing else works and the lopsided imaging is too odd and objectionable as it is, you might try something a bit more radical- Use the channel with the strong vocals as Mid channel and the other as Side channel into a Mid/Side decode matrix.  Adjust the balance between the two channels to suit. This will center the vocal (along with everything else that is dominant in that channel) and spread the stuff which is dominant in the other channel out to the sides somewhat amorphously.  It will destroy whatever discrete Left/Right aspects exist and and replace them with a sort of pseudo stereo effect of solid center with more spacious sides.  You get more of a sense of depth-stereo than left/right-stereo, but that can remedy the radical panning problems.  It has worked for me in the past to get something listenable out of overly lopsided recordings.

Lastly, decide if it might be acceptable to a listener to have the vocal all the way to one side.  That's not typical but can work sometimes.  It can be more difficult to accept as the recordist than a random listener since you were at the event when the recording was made and have a mental image of how things were laid out at the live situation and a conception of how its "supposed to sound".  Try to listen to it like a stranger.  That can help when assessing the Mid/Side trick thing too.  Don't compare too much with your concept of how it's supposed to sound, but find something that is simply enjoyable and avoids drawing the listener's attention to the original problem.


Thank you for the thorough response!

Conicidentally, I just tried using RX7 to isolate the vocal in the right channel and paste it into the left. The result is actually a big improvement.

However, I'm intrigued by your pseudo mid side suggestion.  What is the best workflow to accomplish that in Audition 3.0 and/or RX7?

Less directional mics is a no go. The cards are barely enough to suppress the screaming of the usual drunken idiots that frequent these rock clubs...

You can either do it manually in your DAW software, or with a free plug-in.  Some DAWs provide a useful shortcuts to do it easily.. Its been a while since I used Audition so I can't recall specifics there. 

To do it manually in the DAW's mixer without any shortcuts:
1) Break the stereo file into two separate mono tracks.
2) The channel with more vocals will be the Mid channel. Pan or assign that channel to center in the mixer (or route it to both Left and Right channels of the mix bus, which is the same thing).
3) Copy the other channel to a third mixer channel. Pan one copy of that hard Left.
5) Invert polarity on the other copy and pan it hard right.
6) Link the hard-panned channel faders together so that you can adjust their level in unison.  This pair of channels represents the Side signal.
7) Bring up the fader on the Mid channel- you'll get a mono, centered representation of the channel with more vox in it.
8 ) Bring up both hard-panned channels together and by the same amount (easy to do in an even way if they are linked, but that's not strictly necessary).  The more of this you add to the mono center, the wider and more diffuse sounding it will get.

You can further adjust things by un-linking two Side channel faders and setting their levels independently of each other.  This will make it wider on one side than the other, and also increase the level of the wide content on that side simultaneously. 

[Edit] You don't have to keep the Mid panned to center, you can pan it slightly to one side or the other as needed.  This can be helpful in combination with independent adjustment of the Side channel levels to get a good image and Left and Right energy balance.

Doing it manually like this in the DAW mixer will help you wrap your head around what is going on, making it a good way to go.  You can also do it using something like the Voxengo MSED plugin (free download) without the copying and extra mixer channels, but its not as easy to visualize what is going on that way.


[0] Message Index

[#] Next page

Go to full version