SOUND ON SOUND

[Moroccomoose]

So this week I have been playing with a new trick I learned to help reduce....or in some cases almost eliminate...bleed.

I used it on an old recording of a rehearsal session from a band I was in a long time ago. Every track had loads of bleed making it very tricky to have any kind of control over the mix.

The technique I used was to take a track with bleed problem, duplicate it and phase invert. Send both tracks to a bus and of course they null to silence.
Then put a compressor on the phase inverted track with fast attack and high ratio, then tweak the threshold to optimise bleed reduction against pumping.

It works brilliantly on drums, much more natural than using a gate. I even managed to pull a vocal out of the mud a little bit using this technique.

Credit to Warren Huart for describing the technique on his channel.

Stu.

[CS70]

Cool one! Never been a fan of gated kicks and snares myself.

The more I mix drums, the more I tend to leave bleed in the closed mics as well, often using some EQ to shape the sound.

[RichardT]

I thought your post was going to be about something completely different.

Though in its own right it’s definitely interesting.

[The Elf]

I've used a version of this trick to eliminate guitar amp buzz - it can work well.

[Hugh Robjohns]

It's basically parallel compression (aka London compression or New York compression) with a polarity inverting twist!

Low level signals -- meaning the background spill -- wont exceed the compression threshold so wont activate the compressor. Consequently the two signal paths are identical and everything cancels out. Spill removed.

High level signals -- the wanted signals -- do activate the compressor, so get reduced in level (a lot thanks to the high ratio and hopefully low threshold). Now the two signal paths are very different, and thus cannot cancel out or reduce the wanted sound to any significant degree.

Obviously, Setting the right threshold level is critical to the effectiveness, as well as the attack and release times. And you'll want a fast acting (FET or VCA) compressor with a very wide attenuation range.

[Martin Walker]

There I was about to suggest running it under the cold tap, slapping on some Savlon, and then getting on a plaster before it starts bleeding again.

And instead I learned a new audio technique that I intend to try out - thanks Stu! :thumbup:


Martin

[Tim Gillett]

Isnt it just gating or downwards expansion by another route? What advantages does it have over those? Can we hear some AB audio examples?

Important also to compare like with like, ie: use exactly the same equivalent attack/release/ratio settings. Then we are comparing only the two methods.

[adrian_k]

I think I vaguely remember a Mike Senior SOS article from years ago on this, and some other more complex versions, I’ll see if I can find it.

[The Elf]

Isnt it just gating or downwards expansion by another route? What advantages does it have over those? Can we hear some AB audio examples?

Important also to compare like with like, ie: use exactly the same equivalent attack/release/ratio settings. Then we are comparing only the two methods.

It's slightly more benign, in that any lag in the response of the inverted signal leaves the target audio unaffected.

But yes, it's not anything new or miraculous.