Tips & Techniques:Noisy Guitar amp..

I learned this one from Tom Dowd,

We had the right sound for a gtr track, could not make this noise go away,
Tom had me record the noise from the amp on to a track from top to bottom and then we ahead and recorded the gtr track on another track.. when we were done He had me flip the phase on the noise track and bounce it with the gtr track at unity gain to a new track... the Noise pretty much cancelled it self out..He was a genious!

Regards and Happy 008.

[structuredloud]

Strictly speaking noise won't cancel out with a phase invert because it is random. But the sample and invert technique (mentioned in the post) is one of a few techniques that modern "denoise" processors.

[elmono]

Originally Posted by structuredloud

Strictly speaking noise won't cancel out with a phase invert because it is random. But the sample and invert technique (mentioned in the post) is one of a few techniques that modern "denoise" processors.

Interesting, It is amazing what how much time You can save now with all these plug-ins.. Tommy Dowd was probably doing this in the 70's

regards,

C.

[namsabnek]

Got all excited yesterday. recorded three minutes of noise at the same level as the guitar i was tracking. flipped the phase on the "noise" track and nothing happened. Absolutely no noise reduction. Could it be because the noise has variables in the sound which don't coincide exactly with the noise on the guitar track?

[CrankyRayHanky]

No luck here either- I just seemed to be adding more noise

[EstateMatt]

noise is entirely random. I'm not sure who this is that posted, but he's either lying or is missing something. Sounds kind of audio-engineering-school-ish to me. if this worked, it would be built into all analog electronics.

[dana18237]

yeah this trick hasn't worked for me either. i recorded about 20 seconds of noise. then on another track i recorded 20 seconds of playing guitar. flipped the phase on the noise track and summed them. no noise reduction at all. i noticed someone post something about "sample and phase." what exactly does that mean?

[shuld]

Tried the same thing!!! Nada!!!!

[echoclerk]

If the Noise on the Guitar amp is true "noise" then this will not work. noise has no phase so it can't cancell out.

You may be able to get rid of rumbles or pitched hums. if you can match up the frequencies / waveforms.

[rmorris]

Originally Posted by echoclerk

If the Noise on the Guitar amp is true "noise" then this will not work. noise has no phase so it can't cancell out.

You may be able to get rid of rumbles or pitched hums. if you can match up the frequencies / waveforms.

tha's right as far as I can see. Random noise ( Gaussian / White/ pink / hiss ) is random and is as likely to add constructively as destructively.
But a noise like a 50Hz hum is repetitive and will cancel out if added in anti-phase.
Since you can't really know when you begin recording wrt the phase of the hum I'd suggest that whether playing the noise track in or out of phase gives a better result is random and the ideal phase change may be anywhere between zero and 180 degrees.
Some analogue kit ( eg up market mic pre / DI s) wil let you set this variable phase change as it's usually intended to match up DI and mic signals.
Or, of course, you can do it in a DAW.