Very cool trick to deal with noisy amp

[LeSedna]

What i find interesting if it works without sucking the tone too much (it is mandatory it does but maybe very little) is that its more efficient than any noise gate since it also removes the junk during playing.

Im actually curious of why it works just for the sake of understanding.

Just so you know since its the topic : the reaper fft has a cancellation function similar to waves z or x-noise even if its not perfect. When i play for practice i use it to remove the junk even if i suck tone cause its less fatigueing for the ears. You can finetune it too. Maybe its function is sort of similar to what you are doing.

 

[egan.]

It's a different scenario but I've used this technique to get rid of click track bleed before. I haven't tried it for ground hums etc. but the principle is there for sure. The problem with noise reduction plugs is that many "over think" problems and can cause weird aliasing or adversely effect the things you actually want.

 

[espartaco]

Actually, I think some noise reduction plugins can analyse the hum and then remove it (I think it's something similar as to the trick mentioned here).

this! theres a difference between a noise reduction plug in and a noise gate. i used to work in a radio station and to clean all the audios recorded by the reporters in the street i have to use a noise reduction plug in that let you select some part of the audio region where is only the noise/hum etc so it can analyse his freq response and then the plugin filter by some proportion the frequency's where the noise is more predominant. now i dont know for sure if this is the actual process that a noise reduction do but that explanation always seems logical to me , so i was just asking because maybe that explanation is not right and the way a noise reduction plugs really work is whit the phase cancellation produced by the sum of the noise signal whit his pashe inverted 180 degrees like SimonTaddio as stated in his trick.
:Spin: sorry for the long post, i actually interested in knowing the theory behind it, just to feed the nerd in me a little

 

[LeSedna]

Yeah, the trick is as far as we know, phase cancellation can only work if you use the exact same source synchronised and if you flip its phase, like mentioned for a click. In case of noise it's difficult to get why it works because you don't record it in parallel when you record yor track. Maybe in such noises there are repetitive patterns and it's not that random and therefore the phase cancellation has some effect ? Maybe the effect is a secondary signal treatment effect I don't remember which works when you phase cancel a signal with a signal very similar in its frequency spectrum (I've had signal treatment courses a long time ago and it explained me how linear and non linear filters work, comb filtering, stuff like that but don't remember a thing about that)

 

[LeSedna]

Btw, what I mean is that there must be an explanation because there is as much reason for the trick to cancel noise as to actually ADD noise. Since it's not synced, phase flipping is just like hoping it will cancel randomly with some frequencies, but there are 50% chances it does (not o mention there are the whole 360 possibilities).

I would be interested to find the maths behind a noise cancelations plugins, I'm sure it would explain that.

 

[SimonTaddio_Qc]

Maybe in such noises there are repetitive patterns and it's not that random and therefore the phase cancellation has some effect ?

That's my theory, in any case.

 

[kernelxsanders]

i seem to be having a terribly hard time understanding how to go about this. i recorded a track playing a riff, then recorded a second track with my noise gate off, and my guitar's volume down, letting only and all the hiss be recorded, inverted the polarity and nothing happens...

is there any way of simplifying it or am i just that dumb?

 

[Jordon]

You need the guitar's volume up, not off.

 

[topsoul182]

Yeah, I've discovered this trick aswell.. it works pretty good..
It won't cancel all the noise.. but it will (more often than not) be huge improvement..
This can also be done when recording guitar or bass DI.. If it's noisy for some reason.. you can record a track with the git/bass plugged in, but without you playing anything.. Then invert that.. etc..

If I were to reamp the track I would have to record a track with the guitar volume on and then reamp that empty track and then reverse the phase in that one?

 

[OneDaySky]

Good idea man, very similar to the Bob Rock trick he used in the 80's with tracking vocals in the control room with studio monitors bleeding into the mic. Track the vocal and record a 2nd track with just the song playing through and invert that track and it cancelled the bleed.

 

[LeSedna]

Good idea man, very similar to the Bob Rock trick he used in the 80's with tracking vocals in the control room with studio monitors bleeding into the mic. Track the vocal and record a 2nd track with just the song playing through and invert that track and it cancelled the bleed.

No, this is in essence very different cause the bob rock trick is simply phase inverting the same track to remove it's bleed, it's about re-creating the exact, 100% exact same bleed, which obviously can cancel itself therefore.

In the Op, it is about using a random noise to cancel itself. Being random and not synchronized, it's different parameters that changes everything in the problem.

I think I'll try today with an ampsim to see if I can recreate that.

 

[ForHerDeadEyes]

If I were to reamp the track I would have to record a track with the guitar volume on and then reamp that empty track and then reverse the phase in that one?

Well, it should be done with the DI-track.. before you reamp!
Let's say you're goingt to record a bass DI-track infront of your computer, and it's really noisy for some reason..
You record your track..
Then you record to another track, but without playing.. Just sitting there with the bass in your arms..
Then you invert this DI track and add it to the track you played on..

This can cancel out alot of noise and is different from noisereduction, it doesn't leave any artifacts..
It could also be done a 2nd time, with the reamped track to cancel out noise from refrigarator/fans/computers or whatever.. by recording your mic'd cab without a track going through it..

The noise won't cancel out entirely, but it can be a satisfactory solution!

 

[OneDaySky]

No, this is in essence very different cause the bob rock trick is simply phase inverting the same track to remove it's bleed, it's about re-creating the exact, 100% exact same bleed, which obviously can cancel itself therefore.

In the Op, it is about using a random noise to cancel itself. Being random and not synchronized, it's different parameters that changes everything in the problem.

I think I'll try today with an ampsim to see if I can recreate that.

Yes, I wasn't saying its the same thing, but a similar idea, using the same recording source to do cancellation.

 

[HatchetDrummer]

Good idea man, very similar to the Bob Rock trick he used in the 80's with tracking vocals in the control room with studio monitors bleeding into the mic. Track the vocal and record a 2nd track with just the song playing through and invert that track and it cancelled the bleed.

Dont know where you read that? Last time I checked when you track vox in a control room with the monitors on, you polarity reverse one of the monitors so anything coming out of the speakers creates a null at the mic. Obviously it has no impact on the tone of the voice in the mic.

Hence the reason why high end monitor controllers have L or R phase reversal buttons. My Dangerous Monitor does, as does my console on all 3 control room speaker outputs...

The O.P.'s idea does sound cool though, amp noise in generally a solid state noise tone ( little to no fluctuation in freq ) it should actually work quite well.

Dave

 

[Brett - K A L I S I A]

That's what I thought too, with the addition that the audio signal coming from the speakers has to be mono, the microphone at the very same distance from both speakers, and the room has to be pretty good sounding That being said, the idea of recording the sound of the speakers without the singer seems viable as well, as long as the singer didn't ask for volume changes on the speakers during tracking and the microphone didn't move.

 

[OneDaySky]

Dont know where you read that?

'The mixing Engineer's Handook By Bobby Owsinski (Published 1999)
Thats where i read it!

 

[SimonTaddio_Qc]

^
Nice, I'll definitely check out this book

 

[rfahey86]

I've tried using a hiss track with the phase inverted and never had any luck. I don't know if this helps but if you record in front of you computer, the computer monitor is picked up in through the pickups and creates a very low hiss. if you turn away from the monitor it usually fixes that. IMHO

 

[chrisrivalry]

what if you add eq or compression to the track, won't that affect the audio's waveform therefore creating problems with cancellation at certain frequencies?