Wednesday, March 27, 2013

Layering Guitars

Okay, so I'm going to talk a little bit about layering guitars. It's something I'm very familiar with; I record a lot of rock music with lots of guitar layers. Sometimes the easiest way to get a big wall of guitars is to just play a track on the left, then double it as tightly as possible on the right. Sure, that's the old standby and that almost always works. But what about three layers? Or Four? Or Sixteen (which I have totally done before - not recommended). What about when you don't want the guitar on the left to sound like the guitar on the right? 

Sometimes all you need is these two parts. It's often "bigger sounding" to have two parts and nothing else. My favorite "goto" is to have two well written parts, one left, one right, and any leads in the center. Various parts in the song may contain other little guitar riffs or whole note chords on the chorus or last note of the song, but nothing drastic--the frame is built around the two main guitars.

So, here are some tips when it comes to layering guitars. The best guitar tones Ive ever gotten as far as "layers" go have come from:

a) carefully selected guitars
b) carefully selected amps/pedals
c) carefully written parts
d) vision

I'll tackle these one at a time.

a) Carefully selected guitars. My favorite guitar layering sounds come when the tones match. So for example, if you're going for "big rhythm" a lot of times that means two guitars, left and right, probably humbucker but sometimes single coil. IF you're going for more chimey guitars (think 90s rock, Country, Pop) you might be better off using a single coil, or something like single coil left, P90 on the right.

The tones you choose don't have to match exactly, it depends on the role they're playing. Big chorus guitars that create a wall? Probably big humbucker sounds as aforementioned. If your goal is to make it sound "doubled," then DOUBLE IT. If the goal is to make it sound more like a BAND, then you may be better off playing two separate guitars.

I generally try to match the LOW end on the guitars. Like, You generally don't want a really thin guitar on the left and a really thick one on the right. It's generally something like, slightly warmer on the right, slightly edgier on the left. Sure, you can clean up the low end in the mix with filters, but it's generally best to think of this during tracking. So like, two Les Pauls make sense together. A Les Paul and a Tele will be a little harder to match in terms of low end, but there are a few tricks:

-Have the Les Paul Play Higher Up / Tele Play Lower
-Use a "bigger sounding" amp for the tele and a "smaller sounding" amp for the LP.
-Try neck pickup on the Tele and Bridge on the LP.
-Try a ribbon for the tele and a 57 for the LP.

You really just want them to sound like they belong together in the context of the band. Their midranges will differ, which is good, that's what gets you separation. The low end will define its size. So just be cautious when you're matching up guitars.


b) Carefully selected amps/pedals. Though I find that most of the time guitars sound best going straight to the amp, but sometimes pedals can provide a special sound that's not achievable by any amp - like a Tubescreamer does "a thing" that amps don't really do. Pedals often have a sort of compression in them that can help sometimes depending on the part. I generally find that MUCH LESS GAIN is needed when adding guitar layers. I like to think of it this way:

-The more notes being played in each chord, the less grit you can get away with. Solos can sound great with tons of distortion. Big chimey G/C chords? No way. Power chords? Decent amount, but not as much as lead (generally).

-The more layers of guitars you plan to have, the less distortion each one needs. Distortion adds up and starts to sound like "one fizz" coming from the speakers. Heavy rock guitars aren't really that distorted - it's that they need to be played evenly and tightly. It's usually "just enough gain to sound chunky."

-Distortion is essentially clipping. Part of what we "like" about distortion is how it controls the level, not just how it sounds. However, you don't really get "PUNCH" always - it starts to sound well, clipped. Thus, I find that recording mildly driven guitars and compressing them to heck sometimes sounds a LOT better than recording super dirty guitars and not touching them with a compressor. I find that the compressor helps it have the evenness of a distorted guitar sound (which makes the player happy) but it doesn't sound like buzz buzz buzz (which makes you happy).

c) Carefully written parts. I think this is rule #1 (even though it's letter C...) IF the part conflicts with the vocal's range, it will always conflict with the vocal. Think about that. If you're playing the exact same notes that the vocalist is singing, no matter how you EQ it, those notes will be in there. Then you'll be upset and say "aw man I hate making my tone sound like crap in the mix - it sounded sooo good before!" Well, the tone can be perfect, but if the part is wrong, you can't EQ it out.

Beware of where the VOCAL sits. Generally speaking, power chords are easy to fit in a mix (provided the tone is right) because that rarely clashes with a vocal. It's when you start getting into the E-B and G strings that you really start to conflict. In that case, try to avoid playing the exact same notes the singer is singing (unless you're doubling the vocal melody on purpose, which can sound really cool). Another tip that can help here is doing what...oh...LAYERS. Record the "big guitars" down on the low three strings with your meaty drive, but then record cleaner, even lightly driven guitar parts on the higher strings. On both parts, avoid crossing over to each others territory, ie., try to stick to the low 3 or 4 strings on the power chords, and try to stick to the high 3 or 4 strings for the cleaner parts (mostly).

d) Vision. Before you lay down a single guitar, you NEED to have vision of the picture you're about to create. I always like to make a "roadmap" so to speak of the guitar arrangement, because it's so easy to just add and add and add. If you know where you want to end up, you can carefully select tones. Remember before when I said basically that the more guitar parts you have, the less big each one needs to be? Well, if you start by thinking "this song just needs two guitar parts" and then you even up adding 5, they may NEVER fit together because you didn't consider all five in the process.

So. Listen to the song, the rough, the demo, whatever, and try to map out what's needed. Here's an example.

Intro - 5 guitar parts, two power chord Les Pauls, fairly gritty. Then, two chimey Teles with LIGHT grit. LPs Hard left and right, Teles 50/50, lead guitar, creamy distortion, panned center.

Verse 1 - Les pauls drop out leaving the Teles. Probably need a little more grit, but not necesarilly low end or body. This less the low end of the verse drop out a little so it feels smaller but not wimpy. Both teles playing carefully written parts

Prechorus - lead part comes in in the background, not super full, don't step on vocal.

Chorus - Les Pauls come back in, which creates the illusion of the song exploding (as the other guitars are just 50/50, these are 100L/100R.)

Etc.

With this type of vision, you can plan out the tones for each section, each part, and try to envision how it will all come together before you even put up a single mic. TRUST ME, it will save you a lot of pain and suffering later down the road. In the digital world, you can fix a LOT of things. One of the only things you can't hardly fix is over distorted guitars. When in doubt, less gain can work just fine. When in a LOT of doubt, always record a DI.

Another side tip - a lot of n00bs will pull the whole "mic up an amp with two mics and pan one left and one right." As someone said before, that will do nothing by take up a bunch of useless space. It will NOT sound big. Come on, just play it again, is it that hard? It sounds SO much better that way!!! Nevertheless, there are a few times when I will play with out of phase sounds on purpose. I once did a track that needed "something" in the verses with the drums, bass, vocal, and main guitar. But we didn't want an organ or pad, because they wanted a pretty close representation of their sound live, which is just 3 guitars, bass, drums. In that verse, the middle guitar was the only one playing, while the other two sat out. What we ended up doing was using a half-cocked Wah pedal filtering out all the top end of a guitar, and mic'ing up a cab with two 414s in a weird stereo placement. I then flipped one side out of phase and panned them hard left and right. This created the typical "behind my head" sound that out of phase stuff creates, but with the part SUPPOSED to be almost more felt than heard, it worked out beautifully. In the track it sounded so cool because you couldn't really pinpoint what it was...or where it was coming from. It was VERY low in the mix.


I hope these thoughts have been helpful to someone out there!