Doubling Guitars

[Mixerman]

I don't see how liking symmetry is ignoring strategies though, mixerman. I can't stand hearing some frequency area too heavy on either side in headphones, it makes me feel like one of my ears is weighted down, or being stabbed or something like that.

Using panned doubled guitar doesn't do this to me when they both have roughly the same tone, and if there's two different tones both doubled, then I can't stand one tone in one ear and one in the other I think it sounds more 'full' if they crisscross instead of a doubled guitar in my left ear and a doubled different tone in my right ear. I find the frequency difference really thrown off, even if they are all playing the same part. Tone A left and tone A right, tone B left and tone B right works to thicken it out, but tone A x2 L and tone B x2 R just sounds lopsided.

The only time I can deal with hard panning is with exactly the same tone on both sides, or for some kind of special effect part of a song, but riding the whole song just makes it weighted sounding as different frequencies will sound heavy on one side or the other. Its a "strategy" to avoid doing things experimentally when you know your ears don't like the result.

This is one of several reasons why you shouldn't judge mixes in headphones.

There is an absolute abundance of material out there that uses asymmetry, hard panning of unique parts (percussion comes to mind), and guitars with different tones on each side. Your problem with this in headphones is a personal hangup. It's not based on the history of mixing within the stereo field, not by a long shot. The panning plane of space is just as important to use fully as the frequency plane of space. There's really very little difference between not utilizing the full width of the mix and putting an LPF and an HPF filter on your mix in order to chop off the top and bottom frequencies of the mix respectively. Either way, you're purposely limiting the full breadth of your mix space. Depending on the track, this might be a good mix decision. But to make this decision repeatedly because you can't get over your own personal problem with lacking symmetry is only going to hold you back from improving your mixes.

Make no mistake, dbl tracked guitars of the same player and rig isn't nearly as wide as two distinct tones and parts panned wide. Again, there are times when a pure dbl tracked gtr is a good treatment, as it evokes a certain feeling from the track. But not purely as a matter of course. You relegate any music you're involved in to one particular sound, which in turn tends to evoke one particular feeling, and makes it more difficult for listeners to distinguish between songs.

Enjoy,

Mixerman

 

[RAMI]

I think what is being suggested is that within the symmetry of a mix, you can occasionally throw in asymmetrical elements just to break things up...to change the focus.

Yes, i agree. When I talk about having a symmetry fetish, I'm exaggerating (a little). But I'm also refering more to my "bed", usually drums, bass and rythm guitars. After that, I'm all for having the guitar solo tilted a little left, or the group of back-ground vocals sitting as a group a little right, or whatever.

I'm sure nobody meant that as soon as something comes in a little right, something else has to cone in a little left, etc....

 

[Greg_L]

I didn't reads those long posts, but I suspect this is a case of someone taking typed words a little too literally. Blame the internet.

 

[Dogbreath]

I didn't reads those long posts, but I suspect this is a case of someone taking typed words a little too literally. Blame the internet.

Actually some pretty damned good points in there.
Gonna put some of these tidbits to use,
Thanks MixDude

 

[Fletcher]

If you really want a great texture and a wide image from dbl tracked guitars, it's even better to swap out the player on the dbl. Of course, the effectiveness of this is dependent on the quality (and existence) of the second guitar player.

Due respect mate - if you change the rig you're making the sound "wider" - if you change the player you're adding a second part rather than achieving a "double"... much like doing the same part in an 'open tuning' isn't a double, but another part that is VERY different from the first. There are elements of a double track where using a different guitar, different amp, different recording signal path will net you a degree of wider results - but the same player is necessary to play the way it was played the first time in order to achieve a true "double".

There is an absolute abundance of material out there that uses asymmetry, hard panning of unique parts (percussion comes to mind), and guitars with different tones on each side.

All kinds of parts - I worked with one artist who had a different guy produce every song on the record [interesting experiment] - one of the songs could have worked really well as a keyboard-centric song with a bit of acoustic guitar or as a guitar-centric song with small additions of piano and B-3 [interesting quandary].

As the song was built it was built to go either way, with the parts a bit sparser than if the song have been built if the decision was to go in one direction or the other. The final net result ended up being two almost different arrangements that were entirely asymmetrical - the guitar centric song on the left, the keyboard centric song on the right... put them both together and it was seriously lush without anything stepping on anything else -- but for all intents and purposes two entirely different arrangements [where either could have stood on their own] that shared common vocals and rhythm section.

If you're doing this for fun [vs. profit / as a career] its a damn interesting learning exercise!!!... no matter the genre it is still something you can achieve and it will teach you a hell of a lot about production and arranging as you get better and better with the technique.

Peace.

 

[Mixerman]

Due respect mate - if you change the rig you're making the sound "wider" - if you change the player you're adding a second part rather than achieving a "double"... much like doing the same part in an 'open tuning' isn't a double, but another part that is VERY different from the first. There are elements of a double track where using a different guitar, different amp, different recording signal path will net you a degree of wider results - but the same player is necessary to play the way it was played the first time in order to achieve a true "double".

We don't actually disagree here. My definition of double (in this post) is a bit looser than yours, but your point is well taken. Clearly, technically speaking a double is the same part played again, and that has a different sound and feel to two unique parts, but when it comes to symmetry and asymmetry, it's really a function of one guitar v. two guitars notwithstanding other parts from other instruments in the arrangement, and notwithstanding extra guitar parts used for texture.

Enjoy,

Mixerman

 

[JB's Studio]

A quick and dirty way is to: use a modeler and use a patch with dual amp (different amp models) in a stereo patch and add a compressor in front of one of the amps, and a stuck chorus effect to one side to give separation. It can sound huge.

Another quick and dirty way is to reamp and throw a compressor on the output from your DAW to the amp, then change the EQ in front of the amp to give a feel of a different pickup. The compressor will change the attack wave form and the EQ changes the shape.

Uh oh.... my post follows mixerman. I'm in trouble. lol.

 

[Mixerman]

A quick and dirty way is to: use a modeler and use a patch with dual amp (different amp models) in a stereo patch and add a compressor in front of one of the amps, and a stuck chorus effect to one side to give separation. It can sound huge.

Until you put it in mono...

Phase issues make that not the greatest idea. You need a true time differential when you have a split signal of an identical performance, regardless of how you might disguise it with EQ (amp or otherwise) and even chorus. That time differential needs to be excessive enough that one side would actually be noticeably late. It takes a minimum of 22ms to properly throw the signal, and you're still going to have all sorts of phase cancellation issues in mono until you get somewhere above 50ms.

Even when the listener is in the stereo field, you're still going to cause the guitars to wrap around their head to some degree. I personally think this is distracting to the listener, especially on rock music with harmonic instruments as important as guitars. You're far better off tracking a second guitar if your goal is to fill up the sides.

Uh oh.... my post follows mixerman. I'm in trouble. lol.

You sure are!

Enjoy,

Mixerman

 

[miroslav]

I've never tried it (I don't really use modelers, preferring to mic amps)...
...but why would two different amp patches from a modeler have phase issues....?...is it not like layering 3-4 synth patches all playing the exact same sequenced part....?
They would be perfectly time aligned...wouldn't they...?...so why would there be phase issues?

 

[Mixerman]

I've never tried it (I don't really use modelers, preferring to mic amps)...
...but why would two different amp patches from a modeler have phase issues....?...is it not like layering 3-4 synth patches all playing the exact same sequenced part....?
They would be perfectly time aligned...wouldn't they...?...so why would there be phase issues?

Because no matter what you do EQ-wise and compression-wise and amp modeler-wise to the other side, it's STILL the SAME signal. Put the same signal to both sides without a time differential, and they will appear center. Put a 20ms time differential on one side of the signal, and you'll throw the signals to the sides (seemingly) as if they were two independent signals, but you won't throw them sufficiently to eradicate phase issues. The differences that EQ and compression might cause to one side will only alter the frequencies and amplitude of the cancellations that occur. It doesn't eradicate the fact that it's the SAME signal.

Mixerman

 

[miroslav]

OK...they might feel centered, but without phase issues since they are perfectly time-aligned...that's all I was saying.

Though the suggestion was to add a chorus to one signal, I believe, which would/should keep the signals from being identical, so I would think they could then also be panned L/R...without the need for longer delay since the chorus by it's nature adds that delay and makes the signals dissimilar...so the phase issues would not be as present, thanks to thje chorus effect....yes?

Anyway...I would have to try it with modelers to see how much dialing in is needed per amp patch. Like if you have one set for a pure, clean amp and the other for a distorted crunch...that could make them disimilar enough to allow L/R imaging...I would think.

 

[Fletcher]

OK...they might feel centered, but without phase issues since they are perfectly time-aligned...that's all I was saying.

Right -and what you have is the same part - which is not a double, its the same part from the same player doing the same thing. Humans - with very few exceptions - will never do the same thing twice which is what gives you the "size" in a double track situation - the rest is just subtle EQ / distortion character changes which won't give you size.

Yeah you can slide it off - run it through a "doubler" [short delay with a bit of chorus], etc. will give you a "quick and dirty" double but will create other mix issues - as in lack of phase coherence and seeing as this is an audio engineering forum - that kind of mundane shit is actually relevant to the craft. You can create major phase issues with as little as a "one sample" delay between signals [its that sensitive] so you really should know more about the concept of "phase relationships" before doing something like that.

One of the first things I learned when learning music theory is that you can't justify breaking the rules until you know what the rules are - and yes, you can get good results when breaking the rules, but unless you know how to "do it right" you're really just mucking about and hoping to get lucky... once you know what you're doing [and more importantly "the WHY" of what you're doing] then you can attain better results in less time while knowing how far you can go when "breaking a rule".

Peace.

 

[miroslav]

Yes...I agree on playing it twice is a better way to get doubled guitars rather than always splitting/doubling.
I was mainly responding to the "phase" thing...that two identical signals obviously can't be out of phase....but yeah, even with a small time-alignment difference, you can get some phasing, though I don't think that's alway such a bad thing is is often made out to be.

There are always phase differences when mixing multiple signals playing similar parts...and the chance of one signal's frequency canceling or adding to another signal's at some points along their range is very common. Sometimes it's bad, but that's also where/why a lot of good harmonic interaction takes place.

AFA what happens in mono...yeah, that's still something a lot of people check for, me included...but I don't get too hung up about it being all that perfect sounding in mono since there's always going to be some differences than what you have in your stereo mix. I just listen for any extreme differences...but small, sample level stuff is always going to be there.

I was just working on a vocal track and wanted to use some of the alternate takes just in a few spots to double-up the lead vocal...and no matter how I would "slide" the second track's parts, there as always some phasing between them...but, in some spots it was quite pleasant and in others the frequency interactions were not so good.
So phase differences can/do work...it's all in how they interact with each other...from what I've seen.

 

[Mixerman]

OK...they might feel centered, but without phase issues since they are perfectly time-aligned...that's all I was saying.

Though the suggestion was to add a chorus to one signal, I believe, which would/should keep the signals from being identical

so I would think they could then also be panned L/R...without the need for longer delay since the chorus by it's nature adds that delay and makes the signals dissimilar...so the phase issues would not be as present, thanks to thje chorus effect....yes?

No.

Anyway...I would have to try it with modelers to see how much dialing in is needed per amp patch. Like if you have one set for a pure, clean amp and the other for a distorted crunch...that could make them disimilar enough to allow L/R imaging...I would think.

I love this post of yours. "I would think." "I believe." "I would have to try it."

Did you perhaps consider that I've already done all of these sorts of things before? I've broken this "rule" on many occasions, especially early in my career. I stopped, because it weakens the mix.

There is nothing you can do to prevent phase coherency issues between an identical signal that's split and spread, other than to delay it about 50ms. Chorus won't do it.

Hey, I've got an idea...try it!

Enjoy,

Mixerman

 

[miroslav]

Yeah...that's what one SHOULD do...try!
Rather than just read about it and mimic.

Even though I often do split/pan and delay with a generous amount of "ms" when I double some tracks...
...the point I weas making about "I would think." "I believe." "I would have to try it." is that sometimes phasing between signals WORKS in a nice way to create a certain kind of texture.
If you alays just delay everything by a lot of "ms"...then you may remove the most obvious phasing between the signals...but it may also remove that certain texture, and you just get that obvious *delay* going L/R.
There's sometimes a spot where thet delay is not yet obvious, and yet there is also some phasing going on, but it sounds good.

So yeah..."I would think." "I believe." "I would have to try it." and hear how it sounds in my mix rather than just setting things at "50 ms" or whatever.

 

[Mixerman]

Yes...I agree on playing it twice is a better way to get doubled guitars rather than always splitting/doubling.
I was mainly responding to the "phase" thing...that two identical signals obviously can't be out of phase....but yeah, even with a small time-alignment difference, you can get some phasing, though I don't think that's alway such a bad thing is is often made out to be.

There are always phase differences when mixing multiple signals playing similar parts...

The interaction from multiple signals of a similar part will interact in a phase COHERENT manner. Multiple signals of the same part will interact in a NON phase coherent manner unless there's a rather large time differential.

You're confusing natural chorusing which occurs because of very slight tuning and timing discrepancies of two different signals with phase coherencey. Natural chorus can make it sound like there is momentary phasiness, but it doesn't cause phase coherency issues.

This is like debating whether a small apple and a large grapefruit will fall at the same rate of speed. It's settled science.

Mixerman

 

[miroslav]

You're confusing natural chorusing which occurs because of very slight tuning and timing discrepancies of two different signals with phase coherencey. Natural chorus can make it sound like there is momentary phasiness, but it doesn't cause phase coherency issues.

No...I' m not confusing that.
I was talking about/responding to the comment about two signals that have been altered/processed to a certain point of difference (even if the "playing" is identical)...and then split/panned.
And...I've agreed that phasing DOES occur between them...so no need to lecture on that point any further.
I'm saying it's not always BAD...and that you may not always want a huge enough delay to clearly separate them apart.
I'm saying you should listen to it in the context of your mix and then decide.

 

[JB's Studio]

There is nothing you can do to prevent phase coherency issues between an identical signal that's split and spread, other than to delay it about 50ms. Chorus won't do it.

Enjoy,

Mixerman

* problem -- I'll admit it. I'm a hack on the guitar. I'm a keyboardist. If I've got a complicated section, no matter the amount of practice, it ain't coming out the same twice in a row. I know ... loop record until it does.

That's what I've done. But I'm getting lazy. What I'm trying to get is two guitars playing in unison panned to the sides. One is going to change to a harmony. Yes I could use a harmonizer, but I've learned something else >>> never write guitar parts on a keyboard -- when I did this I used the crappy guitar synth in "alsihad" (yes I've read parts of your stuff which was why the "uh oh.") -- because on the keyboard my classical training comes out and the harmonies are never straightforward.

So anyway back to the discussion we're having about chorus vs. delay. You're saying that using a 50 ms delay set 100% wet with no feedback (just so the note sounds approx 50 ms later on one channel than the other) would give a better result than using the chorus for separation without phasing issues? then to test it run the clip in mono to check for anything nasty. If I hear nothing nasty in mono then we're good, right?

I'll actually try this when I get my DAW computer back on Monday. In the shop. Driver issue in Pro Tools..... plus I'm getting it fully checked over because of a bunch of stupid DAE errors when I use more than 3 VIs. I want to rule out hardware before I spend money on a phone call to Avid tech support which I'm coming to the conclusion is a total waste.

 

[miroslav]

You're saying that using a 50 ms delay set 100% wet with no feedback (just so the note sounds approx 50 ms later on one channel than the other) would give a better result than using the chorus for separation without phasing issues? then to test it run the clip in mono to check for anything nasty. If I hear nothing nasty in mono then we're good, right?

Not in any way trying to answer for Mixerman.... ....but that's pretty much what he was saying, and yes, that WILL effectively reduce/remove phasing between the tracks. No doubt about that at all, and no argument about it...there never was.

One thing though...you will hear a distinct delay rather than the unison you are seeking. When I use that approach, I compute the delay based on the BPM and then use the shortest possible fraction of that computation for the delay to get that L/R separation, that way at least it will work with the beat and not sound so "seperated" from the other side, but as always, let your ears decide how much delay is best...and that's all I was getting at earlier, that you should try things within the context of the mix, rather than apply it as a formula whenever you do the split/pan of tracks.
Like I said...sometimes a bit of phasing AND delay in combination will sound good, maybe even better, as you might not need as much delay, and so you get more of that unison vibe you are after. I think that's what you were already kinda' doing with the chorus...as that creates a combination phase/delay vibe.

And again...mono is a good thing to check...but you might not necessarily want to mix for that so much. Lots of good, stereo mixes sound different in mono, and it's not unusual that they sound different and that there some "skewing" of image and frequencies when going from your stereo mix to a mono check.

 

[Mixerman]

* problem -- I'll admit it. I'm a hack on the guitar. I'm a keyboardist. If I've got a complicated section, no matter the amount of practice, it ain't coming out the same twice in a row. I know ... loop record until it does.

That's what I've done. But I'm getting lazy. What I'm trying to get is two guitars playing in unison panned to the sides. One is going to change to a harmony. Yes I could use a harmonizer, but I've learned something else >>> never write guitar parts on a keyboard -- when I did this I used the crappy guitar synth in "alsihad" (yes I've read parts of your stuff which was why the "uh oh.") -- because on the keyboard my classical training comes out and the harmonies are never straightforward.

So anyway back to the discussion we're having about chorus vs. delay. You're saying that using a 50 ms delay set 100% wet with no feedback (just so the note sounds approx 50 ms later on one channel than the other) would give a better result than using the chorus for separation without phasing issues? then to test it run the clip in mono to check for anything nasty. If I hear nothing nasty in mono then we're good, right?

50ms is a HUGE delay. So while that's probably enough to eradicate any phase coherency issues, it's not a reasonable delay between sides musically speaking (not usually anyway).

A 3ms delay is enough of a difference between a player being right on the beat, and laying back (depending on the tempo). Even if you straddle the beat by moving the first part 25ms ahead of the beat, and the second 25ms behind the beat, this would be quite disconcerting for the listener unless there's a very soft attack, the kind you might have from a synth string pad. For any part that has a definitive rhythm, a 50 ms discrepancy between parts is nothing short of unruly.

Without having heard the track, I don't know where the part is or how often it happens, but if, for example, the guitar part is in the chorus, and you actually played the part for every chorus, you can use the C2 part in conjunction with the C1 part, and vice versa. BAM, you've got a dbl.

Other than that, you're far better off playing the dbl. Even if it takes you hours, it's just practice, so it certainly can't be viewed as a waste of time. Can it?

Enjoy,

Mixerman