Post by jazznoise on Feb 10, 2016 14:20:33 GMT -6
I basically can't take credit for this, I stumbled across it in an old TapeOp forum thread years ago when an engineer as lazy as I am claimed to have found a way to avoid having to do headphone mixes. We'll call this type of monitoring "naked" because that's about as sexy as a conversation on monitoring can get.
So, first are the reasons not to do it:
Now that I've thoroughly put you off, it's very simple. Set your mic up in the room with the singer, get your sounds. Turn up the mix until the singer feels they're projecting comfortably for the loudest parts, and then hit the red button. You know the one. Get them to do their takes until you feel you've covered all the ground - it's usually pretty fast, I find, as this is a very natural way to sing.
Now simply mute the vocal and play the track back and record the mic with the singer not singing. Everyone stays where they are, nothing moves. I'll mention this in advance for some people who I know will immediately get bored and start wiggling about like an idiot. No gain changes, no monitor level changes. Now what we have is two tracks - singer + room noise, and room noise. The two noise signals should be nearly identical, so flip the phase of the second track and the two signals will cancel. Simply bounce those to a single track or nest them until your client goes home and you're done.
"Hey Chris I still have noise!"
Well it never cancels perfectly, and here's why: The singers acoustic shadow may change if he/she moves. Your acoustic shadow will change it, if you're in the way (actually the worst culprit, you're best off moving out of the way and having a long count in) and any processing or gain changes will reduce it as distortion from the speakers, mic and pre change. This means compression increases the noise, of course, but EQ is fine as long as you print the EQ to both. The good part is I find the bleed sounds pretty good - high pass above the singer and it's usually a very evenly bandpassed signal.
Now that you've done it, you can go crazy. Try setting up room mics and recording some ambience with the singer, same thing applies. And you can do it for most instruments too.
So, first are the reasons not to do it:
- Your singer wants artificial verb
- Your singer can't stand still
- You can't leave well alone (basically any processing or change of levels will make this less effective)
Now that I've thoroughly put you off, it's very simple. Set your mic up in the room with the singer, get your sounds. Turn up the mix until the singer feels they're projecting comfortably for the loudest parts, and then hit the red button. You know the one. Get them to do their takes until you feel you've covered all the ground - it's usually pretty fast, I find, as this is a very natural way to sing.
Now simply mute the vocal and play the track back and record the mic with the singer not singing. Everyone stays where they are, nothing moves. I'll mention this in advance for some people who I know will immediately get bored and start wiggling about like an idiot. No gain changes, no monitor level changes. Now what we have is two tracks - singer + room noise, and room noise. The two noise signals should be nearly identical, so flip the phase of the second track and the two signals will cancel. Simply bounce those to a single track or nest them until your client goes home and you're done.
"Hey Chris I still have noise!"
Well it never cancels perfectly, and here's why: The singers acoustic shadow may change if he/she moves. Your acoustic shadow will change it, if you're in the way (actually the worst culprit, you're best off moving out of the way and having a long count in) and any processing or gain changes will reduce it as distortion from the speakers, mic and pre change. This means compression increases the noise, of course, but EQ is fine as long as you print the EQ to both. The good part is I find the bleed sounds pretty good - high pass above the singer and it's usually a very evenly bandpassed signal.
Now that you've done it, you can go crazy. Try setting up room mics and recording some ambience with the singer, same thing applies. And you can do it for most instruments too.