You are here

Problems mixing rhythm guitars

For everything after the recording stage: hardware/software and how you use it.

Problems mixing rhythm guitars

Postby Demious » Sat Aug 08, 2020 3:27 pm

I keep having problems with phase issues with rhythm guitars and I could sure use some help figuring this out.
Ive got 2 recordings of the same guitar part. When I measure correlation, all frequencies are +1, or close to +1, as long as both tracks are panned center, but as soon as I start panning these tracks left and right, correlation drops to around, or below 0, leaving me with a completely dulled sound in mono.

I was taught that phase issues turn up when 2 of the same signals are not lined up and therefor, we dont use a copy of a track for doubling, but we record the same part twice. The waveform of both parts will be different and there shouldnt be cancelation problems.
But this is not the case. Even thought the signals are quite different, I get severe cancelation.

Doubling guitars is a common technique and I worked through several courses, from different teachers, about guitar recording and mixing, and they show you to just pan left and right and check in mono if they sound the same, but non shows, or explains a situation where the mono sound declines drastically, or how to fix this. These courses only show the perfect situation, but so far, I never encountered one that didnt give me problems.

What do I need to do to prevent/fix phase cancelation in mono, when mixing a doubled guitar part, panned left and right?
Demious
Poster
Posts: 40
Joined: Sat Jun 30, 2018 1:07 pm
Cheers!
Maarten

System: MSI Z370A-Pro, i7-8700k @ 5GHz, 32GB @ 3200MHz, Win10/x64, Studio One v5, Presonus Quantum 2626, Adams T7V

Re: Problems mixing rhythm guitars

Postby CS70 » Sat Aug 08, 2020 4:04 pm

Sounds odd. Reversed polarity in any of them? Are you absolutely positive you are not adding the tracks to themselves (for example via a reverb or delay effects)?

Filtering can happen in general but the effect so extreme as you describe seems more likely to be due to some routing mistake..
User avatar
CS70
Jedi Poster
Posts: 6450
Joined: Mon Nov 26, 2012 1:00 am
Location: Oslo, Norway
Silver Spoon - Check out our latest video and the FB page

Re: Problems mixing rhythm guitars

Postby Demious » Sat Aug 08, 2020 6:29 pm

I checked, but there are only a few connections.
I start with 2 tracks/channels, one panned left, the other right, each channel has a send to the master ambient reverb (a short, reverb, mostly early reflections, to create the sense of a room the instrument is recorded in), each panned opposite of the channel's panning, to create a sense of depth.
Then both channel are fed to a stereo sum bus, that has the processing for the complete guitar sound, lo-end compression, EQ, limiter and saturator. On this bus is the send to the main reverb (Hall reverb with a more late reflection reverb, to place all instruments in the same space).
And from there, the signal goes into the guitar main bus, that has some EQ and light compression going on.

I tested with and without processing and sends, at several positions of the signal chain, but non of the components show a significant change in correlation behavior... The only thing that shows to severely affect the phase change is panning
Demious
Poster
Posts: 40
Joined: Sat Jun 30, 2018 1:07 pm
Cheers!
Maarten

System: MSI Z370A-Pro, i7-8700k @ 5GHz, 32GB @ 3200MHz, Win10/x64, Studio One v5, Presonus Quantum 2626, Adams T7V

Re: Problems mixing rhythm guitars

Postby The Elf » Sat Aug 08, 2020 7:59 pm

I never have this problem with panned doubles. Give us ten seconds of each guitar (completely dry) from the same section of the song and let's see what you're working with.
User avatar
The Elf
Jedi Poster
Posts: 15241
Joined: Tue Aug 14, 2001 12:00 am
Location: Sheffield, UK
An Eagle for an Emperor, A Kestrel for a Knave.

Re: Problems mixing rhythm guitars

Postby Demious » Sun Aug 09, 2020 9:22 am

The Elf wrote:I never have this problem with panned doubles. Give us ten seconds of each guitar (completely dry) from the same section of the song and let's see what you're working with.

Yeah, that! I traced the source of the problem, its the guitar itself. The problem is already in the unprocessed DI signal itself. The signal is recorded with the bridge pickup, so I guess the problem is in that pickup. It are EMG 707 pickups and this guitar is only a few months old and the bridge pickup doesnt show this problem, so Id think there must be something wrong with the bridge pickup. The question now is 'what?'.
The connection between de coils is internal, there are just 2 wires, a black and a white and from both pickups, the white is used for signal, the black goes to ground, so just reversing black and white wont fix it if the coils of one pickup are out of phase with eachother.
And still it seems strange, I dont hear a phasy, or thin sound from this pickup, so Im a bit confused about whats going on here and how to proceed.
Demious
Poster
Posts: 40
Joined: Sat Jun 30, 2018 1:07 pm
Cheers!
Maarten

System: MSI Z370A-Pro, i7-8700k @ 5GHz, 32GB @ 3200MHz, Win10/x64, Studio One v5, Presonus Quantum 2626, Adams T7V

Re: Problems mixing rhythm guitars

Postby Sam Inglis » Sun Aug 09, 2020 10:20 am

Demious wrote:Ive got 2 recordings of the same guitar part. When I measure correlation, all frequencies are +1, or close to +1, as long as both tracks are panned center, but as soon as I start panning these tracks left and right, correlation drops to around, or below 0, leaving me with a completely dulled sound in mono.

Unless I've misunderstood something, this is entirely correct. If you pan both parts centrally and then measure correlation of the resulting signal, you should see +1, because you are now metering a mono signal. If you pan the parts hard left and right, correlation should be roughly zero as they are entirely separate recordings.

It is also completely normal for double-tracked, hard panned guitars to take a hit when the mix is auditioned in mono. It's one of those compromises that you choose to live with when you use this technique.
Sam Inglis
Moderator
Posts: 2688
Joined: Fri Dec 15, 2000 1:00 am

Re: Problems mixing rhythm guitars

Postby The Elf » Sun Aug 09, 2020 1:05 pm

Demious wrote:
The Elf wrote:I never have this problem with panned doubles. Give us ten seconds of each guitar (completely dry) from the same section of the song and let's see what you're working with.
Yeah, that! I traced the source of the problem, its the guitar itself...
If you've double-tracked the guitar I don't see how any of this would make any difference. Yes, it might affect the sound of each take, but panning each takeL/R would work just as well, regardless of how good/bad the source sounds.

I don't think we have the full story here.
User avatar
The Elf
Jedi Poster
Posts: 15241
Joined: Tue Aug 14, 2001 12:00 am
Location: Sheffield, UK
An Eagle for an Emperor, A Kestrel for a Knave.

Re: Problems mixing rhythm guitars

Postby Demious » Sun Aug 09, 2020 8:08 pm

Im still testing... It wasnt the guitar either. I recorded net parts, but these all do the same, from different guitars, no plugins in the chain, no sends, no side chains... just an unprocessed DI. But I found something, though.
As soon as theres a signal going into a channel, from either a recorded track, or from an input, even a silent signal (a silent, gated guitar) and I pan just a little, correlation drops 45 or 90 degrees. All frequencies on a straight, horizontal line. When theres actual sound on the channel, the meters start bouncing, but are often very close, or below 0. Doesnt even have to be 2 channels, one mono source already does this.
And it doesnt have to be hard panned, one click of the mouse wheel out of center is enough.

Any ideas of what I can do to close in on the problem even more?

The program Im using is Studio One 5.

Sam Inglis wrote: It is also completely normal for double-tracked, hard panned guitars to take a hit when the mix is auditioned in mono. It's one of those compromises that you choose to live with when you use this technique.

Ive noticed, but this is more then a compromise, I hardly have any sound of the guitar left, it completely drowns in the mix.
What is the best way to move the rhythm guitar to the sides, doesnt have to be hard panned, to create more space for bass and vocal in the center, in such way that its actually mono-compatible?
Demious
Poster
Posts: 40
Joined: Sat Jun 30, 2018 1:07 pm
Cheers!
Maarten

System: MSI Z370A-Pro, i7-8700k @ 5GHz, 32GB @ 3200MHz, Win10/x64, Studio One v5, Presonus Quantum 2626, Adams T7V

Re: Problems mixing rhythm guitars

Postby CS70 » Sun Aug 09, 2020 8:36 pm

Demious wrote:Im still testing... It wasnt the guitar either. I recorded net parts, but these all do the same, from different guitars, no plugins in the chain, no sends, no side chains... just an unprocessed DI. But I found something, though.
As soon as theres a signal going into a channel, from either a recorded track, or from an input, even a silent signal (a silent, gated guitar) and I pan just a little, correlation drops 45 or 90 degrees. All frequencies on a straight, horizontal line. When theres actual sound on the channel, the meters start bouncing, but are often very close, or below 0. Doesnt even have to be 2 channels, one mono source already does this.
And it doesnt have to be hard panned, one click of the mouse wheel out of center is enough.

Any ideas of what I can do to close in on the problem even more?

The program Im using is Studio One 5.

Even more odd. Correlation usually is between -1 and 1, I guess you're talking of a phase meter?

Let's put down the facts: you think you have two tracks only, with nothing on, both going to a master bus, on which you have a correlation and phase meter. No additional buses, no reverbs, no delays, nothing. Right? And panning one of the two tracks gives you correlation issues.

I still think you're inadvertently routing the same signal twice in some way.

But, I don't know Studio One, but in other DAWs something similar might be due to a corrupt project, where some channel/bus information exists but his hidden from the guy.

Have you tried loading the two tracks into an entirely new project?
User avatar
CS70
Jedi Poster
Posts: 6450
Joined: Mon Nov 26, 2012 1:00 am
Location: Oslo, Norway
Silver Spoon - Check out our latest video and the FB page

Re: Problems mixing rhythm guitars

Postby The Elf » Sun Aug 09, 2020 9:32 pm

As above, there has to be something odd happening in your routing. As to what... :?
User avatar
The Elf
Jedi Poster
Posts: 15241
Joined: Tue Aug 14, 2001 12:00 am
Location: Sheffield, UK
An Eagle for an Emperor, A Kestrel for a Knave.

Postby Demious » Tue Aug 25, 2020 7:38 pm

Thanks for all the info, guys
Demious
Poster
Posts: 40
Joined: Sat Jun 30, 2018 1:07 pm
Cheers!
Maarten

System: MSI Z370A-Pro, i7-8700k @ 5GHz, 32GB @ 3200MHz, Win10/x64, Studio One v5, Presonus Quantum 2626, Adams T7V

Re: Problems mixing rhythm guitars

Postby Hugh Robjohns » Tue Aug 25, 2020 8:13 pm

Demious wrote:Ive got 2 recordings of the same guitar part. When I measure correlation, all frequencies are +1, or close to +1, as long as both tracks are panned center...

Of course! Correlation is essentially measuring the differences between the two channels. Anything panned to the centre is feeding both channels equally, so both channels are inherently the same and thus the display will read +1.

...but as soon as I start panning these tracks left and right, correlation drops to around, or below 0, leaving me with a completely dulled sound in mono.

A reading of 0 means the two channels are uncorrelated -- you have different stuff on the two channels... Which you obviously do! This is the definition of 'perfect stereo' and it doesn't inherently mean the mono sound should be dull -- although the sound quality will normally be slightly different between the stereo and summed mono signals. Significant dulling normally only occurs when the correlation meter goes well below the 0 mark towards -1.

If it sounds dull when summed to mono (or when both tracks are panned centrally, which is the same thing!) then there is onbiously some degree of cancellation going on.

So taking the 0-ish correlation meter reading and the dull sound when summed to mono, the implication is that the two guitar parts are different -- in frequency content, level and/or timing -- but they are not different enough!

Even thought the signals are quite different, I get severe cancelation.

They are clearly not as different as you think they are, or each sound is contributing to both channels in some way. That could be because they are not fully panned or there's something awry in your routing.

However, the strongest candidate from what youve said is your opposite panned reverb. If the reverb effects contain some dry signal, or have very strong early reflections, that could very easily be the cause of your problems.
User avatar
Hugh Robjohns
Moderator
Posts: 28938
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound