Note: While this post about Sticking Your Neck Out appears after That Little Yellow Car, this track was really the first one I completed. For whatever reason, I forgot to post the blow-by-blow account of this track first.
Last summer I had the great fortune of attending Steve Vai's "Vai Academy". Not only did I get to meet two of my all-time favorite guitarists (Eric Johnson and Steve Vai), but I also met so many other cool, passionate guitarists.
One of the highlights of the camp is getting the chance to jam with Steve and his amazing band which was the chance of a lifetime—but also terrifying. Each night, a block of time was set aside for each "camper" to come up and jam with Steve. Steve always left it up to the camper what they were going to play. Sometimes it was a cover-tune, sometimes it was some chord progression, sometimes it was just a blues-jam.
My slot was at the end of the week which meant I had several days to observe the jams. One thing I desperately wanted to avoid was wasting time on-stage going back and forth with Steve with "what do you want play?" "I don't know, what do you want to play?"
On the afternoon of my "jam night", I went back to my room and started screwing around with various chord progressions. Eventually I settled on this simple Dm7-G7 two-chord jam. When it came time for me to walk on-stage I gave the rhythm section a quick description of the feel, busted out these opening chords and we were off! It was a terrifying and exhilarating experience that I'll never forget.
When I got home from camp, that little two-chord vamp wouldn't get out of my head. Eventually I started cooking up one part after another until I pretty much had an arrangement in my head. What started as a simple jam progression quickly bloomed into a full song. It was time to get it out of my head and into something I could share. That humble little two chord progression evolved into Sticking Your Neck Out:
I find the art and science of recording and mixing endlessly fascinating. If you're like me, then you'll enjoy the blow-by-blow breakdown of this tune that follows…
My DAW of choice is Logic Pro (please, no DAW-holy-war emails). So the first task was opening up a new Logic Pro project, figuring out the basic tempo and laying down the rhythm guitar parts that I'd figure out so far. This process was much simpler than usual by virtue of using the Kemper Profiling Amp. I just plugged the Kemper straight into my audio interface and into Logic and was ready to start recording—no fussing around with mics and cabinets.
My first pass through the arrangement was just against a simple click. I wasn't ready to think in detail about the drums yet and I didn't want to get lost on some rabbit-hole of fiddling with the drum parts. The primary goal here was simply to capture the idea. Once I had the basic parts down, I added arrangement markers to the project for basic navigation.
Logic has both arrangement sections and markers which appear at face value to do the same thing. Of the two, markers are actually more useful for navigating around a tune. But the arrangement markers turned out to be key in getting the drum tracks setup (more on that later).
The next step was to lay down a bass track. Since I had already figured out the song's arrangement, when it came to track the bass, I could focus more on the execution and less on orchestration. This meant it took much less time to record the bass tracks. For these I just plugged my G&L straight into a LA-610 compressor and then into my audio interface. I wasn't totally happy with the tone but I wasn't going to let that interrupt the creative process. Onward!
My primary musical identity is as a guitarist, but in some alternate-universe I've always imagined myself as a drummer. I have a drum kit and even took lessons for a few years. So it's safe to say that I have some pretty strong opinions about how I want my drum-tracks to sound.
In Logic Pro X, Apple introduced a new "Drummer" track which I've played with a bit, but always felt like I had to give up too much control. In this song I laid down a Drummer track just to get some basic scaffolding in-place, assuming that I would end up replacing it with a hand-crafted drum part.
The surprising thing is, the Drummer track turned out pretty good and with just a little exploration and tweaking I found Drummer wasn't just adequately managing the task, it was actually coming up with some delightful drum parts. At this point I made the decision to forgo programming the drums myself (a very laborious process) and I was just embraced Drummer fully.
Where Drummer really shines is when you treat it as if you were in the room with a real drummer. Unless your a complete musical dictator and/or you're paying the drummer, you don't really get to tell a drummer exactly what to play note-for-note. Instead you have a conversation. You talk about some stylistic guidelines. You talk about how the energy needs to ebb and flow in different parts of the song. You talk about how the drummer should change the instrumentation in different parts. Using a Drummer track in Logic is the same thing. You have the same conversation, but instead of words, you use the Drummer interface.
When the drum part needs to change, you simply make a new Drummer clip (or slice one in two with the scissors tool) and tweak each clip. One important thing to understand about Drummer clips is that your drummer will often put fills at the end of each clip. So if you have a very specific place you want a fill, break the clips up at the point where the fill should end.
Crafting the Lead
While I had a really strong sense of the basic song arrangement and its harmonic movements, I didn't have the foggiest idea of what the lead was going to be. I always imagined this tune to be entirely instrumental, but I wanted it to have a strong melodic component and not just be a vehicle for flashy shredding.
The basic meter of the tune and the cadence of the chords felt like the song broke down logically into a series of four-bar "chunks". So I setup a new track in Logic for lead guitar, connected the Kemper and set the cycle marker for the first four bars. At first I'd just play along and try out different ideas until I came up with some rough idea of what I wanted. Then I'd hit the record button and do anywhere from four to ten takes on that idea with different variations. Once I felt like I'd captured one or two good performances, I moved the cycle marker to the next four-bar section and repeated the process.
Along the way I hit upon the main melodic elements which are repeated in the tune. The rest of the tune I left as a blank canvas for different lead lines and bit by bit the song came together. I got most of the lead lines done in a single evening session of a few hours. My fingers were absolutely trashed by the end, but I had made a lot of progress and knew I'd be able to use what I'd recorded.
One important technical note is that intentionally disabled all of the lush, juicy reverb and delay that I usually play with on the Kemper. Those effects sound great on the Kemper on their own, but I wasn't sure how they would fit into the final mix, so I recorded all of the guitar parts "dry".
But hearing those bone-dry parts while recording them can throw me off my game a bit. So a good trick I used was to insert a nice plate reverb (EMT 140) in my Apollo audio unit on the monitoring side. This meant that I could dial in a little of that comforting reverb that was only present in the speakers while I was recording, but the original dry signal is what went to Logic.
One other thing I did was also record a plain direct signal. I have a Radial JDX-48 direct box which let me split the signal from my guitar into both the Kemper and into my Apollo directly. The direct signal isn't useful by itself, but it's a nice insurance policy to re-amp if the need arises. This turned out to be useful later.
After getting the arrangement of the song figured out, I moved into the phase of capturing good tones, trusting some of the refinement to be taken care of later in the mixing process. I was pretty happy with the guitar sounds, but the bass just wasn't happening. Since the Kemper worked so well for the guitar parts I figured I'd try the same approach with the bass. I grabbed pack of bass patches for the Kemper, plugged in and in fifteen minutes I landed on a much meatier bass tone that fit the song much better.
This may have been one of the smartest decisions I made with the track. If I had gone with the original bass tracks I imagine I would have spent much more time trying with EQ and compression on the bass tracks trying to get them to sit right with the drums. But by getting the tone right earlier in the process, it made mixing so much easier. The trick is to develop enough intuition for how basic tones will evolve with post-processing (EQ, compression, other effects) without worrying about them when the focus is on capturing performances.
As I hit each major milestone in the evolution of the song, I would bounce a mix to iTunes and listen to it later with undivided attention. The more I listened, the more I felt like the rhythm guitar part was too monochromatic. I loved the barking, aggressive sound I got with the opening chords but it was too fizzy sounding in other parts of the tune.
So back to Logic I went, opening up some new tracks in the hopes of finding some complimentary guitar tone. I went back to the Kemper and started fishing around for some kind of tone that was very different from what I had originally tracked. The original rhythm parts were done with a Peavey 5153 profile. What would the opposite of that be? How about a Vox AC 15? Maybe we'll add a little pitch-shifting and oh my I think we're onto something.
What started out as a big Eddie Van Halen-fest now had a little mid-80's Alex Lifeson injected into the mix and suddenly the tune exploded with so much more life. There was now wiry, ethereal counterpoint to the aggressive bite of the 5153.
Editing and Tweaking
The recording phase was finished once I felt like I had tracks with good performances and good tone. Recording all of the parts "dry" (without effects) was extremely helpful in being able to hear the details of each part and listen critically.
Once these tracks were down it was time to move onto the polishing phase. Before jumping into mixing, I wanted to trim off the rough edges of what I'd captured so far. This next phase is where DAWs really shine—though it's also full of potential time-sucking pitfalls if one gets a little too obsessive.
One of my favorite features in Logic is the ability to synthesize a "comp", which is a composite track stitched-together from several takes. Each of those four-bar loops had several performances to choose from. For each of those little four-bar chunks I had several takes to choose from. I'd cycle each four-bar chunk, tweaking the comp until I had a performance I liked.
In some cases I could just select a single take as the best one for a given four-bar chunk. Other times I had to assemble these from parts of takes. Fortunately Logic's "quick-swipe comping" made this extremely easy.
Bouncing and Cutting
Once the comping was finished and I had chosen the best performances, I bounced all of the multi-take performances down to single-region audio tracks (hiding the original multi-take tracks). Then I applied Strip Silence to each of these to cut the regions apart, removing the quieter sections which tend to build up noise and various unwanted artifacts.
Then for each region I zoomed in and resized the region to the very beginning and end of audio, with a nice little cross-fade for each to avoid any "popping" artifacts.
After all that, there were still three or four little spots that needed the musical equivalent of Photoshop tweaks. These were a few bent notes that weren't quite in tune, or a few bass-notes that fell off the beat. Enter Logic's Flex Time and Flex Pitch tools.
Flex Time analyses a track and breaks it up into discreet time sections. You can then move these sections within the region without moving the others allowing you to move that one errant bass note back on the beat.
Similarly, Flex Pitch analyses tracks and places them on a piano-roll interface and allows the smallest of adjustments to pitch. I had a few bends that were just a little north or south of an acceptable pitch. Rather than re-record the entire part, I just pushed and pulled those few notes into a range that sounded better. One thing I noticed was that pushing those notes to be pitch-perfect often sounded more out-of-tune than where they started. I suspect a lot of this has to do with the short-coming of scale tempering as well as how the modern-ear interprets micro-tonal changes, especially in bluesy string-bends which often have a lot of character by not being perfectly in-tune.
Is all of this cheating? Not to me. My goal was to get the best expression of the musical idea down. Sure I could have gone back and re-recorded those parts to get them right, but the net-effect would have been the same but with twice the effort. I have more important fish to fry.
As I moved through this process I came to understand that there was less and less significant change I could effect on the track. The performances and tone I captured during the recording process had the biggest impact on the overall tune. The editing and corrections I made in the second phase had less effect, but were still significant. By the time it came to mixing all I could really do was polish. This wasn't a place for "correcting" anything. If a part wasn't good enough by this point, no plugin was going to save it.
Reverbs and Delays
At first I started throwing delays and reverbs around just to add a little space to an uncomfortably dry mix. But I realized I didn't really know what I was doing and was likely to make it harder to get the overall mix right. So turned all of those plugins off and setup a single Aux channel with a single instance of a EMT 140 plate reverb plugin (from Universal Audio).
Plate reverbs have been used on so many records it's impossible to recount them all. To my mind, the first two Van Halen records are the definitive plate-reverb rock and roll records. I simply love plate reverb. So this first Aux channel was going to be my go-to for establishing the space of the virtual room in which this tune was going to be heard.
The lead guitar needed a little more "epic-ness", so I added a stock Tape Delay plugin along with the bus send to the reverb Aux channel. After some careful listening to a few rough mixes I felt like the reverb and delay weren't playing well together. So I changed my routing on the lead tracks and moved the Tape Delay to another Aux channel and routed the guitar to it. Then on the Tape Delay channel I sent the delayed signal off to the plate reverb Aux. This had the effect of adding reverb to the delayed signal, but not to the original dry signal which helped the lead stand out more without sacrificing the space.
In the end, that single plate reverb Aux handled all of the reverb for the track. If I had added reverb to multiple tracks individually I was much more likely to create an incoherent mess of reverbs that would have signaled different physical spaces to the listener. One reverb was plenty.
Compression has to be one of the least-understood parts of mixing. That's not because people are dumb, but because compression does so much more than simply control the dynamics of a signal. It wasn't until recently that I had the "lightbulb moment" and compression really clicked for me. I would by no means call myself an expert, but I have a much better idea of what problems compression solves (and creates) than I used to.
The textbook definition of a compressor is that it's a "leveling amplifier". That is, a device that controls the dynamic range (volume) of a signal. If you want to set your mixer levels and want to avoid overloading the signal path with a few odd loud transients, a compressor is a good way to chop down those peaks. Compressors are great for this.
A funny side-effect is that when you chop down those peaks, you typically need to turn the entire signal back up (called "makeup gain") and when you do this, suddenly a part that seemed to come and go in the mix now comes through much more loud and clear than before.
A side-effect of that "raising of the floor" is that the tonal character of the signal now changes. On a bass for example, the signal seems to get richer and more full-sounding. Some compressors also add a certain amount of subtle harmonic distortions that further enhances this effect.
On drums, compression can have a pretty big effect on the tonal character of the kit—especially the kick and snare. A drum sound typically has a very quick initial peak (called the "transient") and body that decays pretty quickly afterward. If you want that transient to stick out more, put a compressor on the drum and set the "attack" to around 10-15ms and the release to something longer like 100ms. This is enough time for the transient to get through before the compressor clamps down on the body.
But, if you want to bring up the trailing body, you can do the opposite. Put a compressor on the track and set the attack to zero, with a quick release (10-15ms). In both cases you're using the compressor to change the balance between the transient attack and tonal body.
Modern mixing has been accused of over-reliance on compressors, particularly in the mastering stage where the dynamic range is crushed. I was extremely cautious about where I added compression to avoid this. I put some on the bass to fatten it up and level out the dynamics of the performance. I setup a separate Aux track with extremely heavy compression for just the kick and snare. I sent some of each to this compressed track and then slowly brought up the Aux to mix the super-compressed signal with the original which resulted in a fatter drum tone (this is known as parallel or "New York" compression).
One thing I did not do was put a compressor on the final master output bus. A lot of mixers like to do this, but I felt like I should be able to create a mix that stands on its own without that final "glue" mix. Plus, if ever got this mastered by a professional, they would likely add a touch of mastering compression and I'd end up with the modern over-compressed goo I tried so hard to avoid.
I didn't do much EQ tweaking on the parts. I put a high-pass filter on the sparkly Alex Lifeson guitar parts and dropped everything below 200 Hz. That's typically a place where a lot of muddy build-up happens and I wasn't losing anything significant by chopping off that region.
On the plate reverb bus I added a low-pass filter and chopped off everything above about 9 kHz, to avoid too much "sizzle", which can be fatiguing in large doses. One guiding principle here was that I only cut frequencies and never boosted any. It's tempting to try heavy-handed tonal surgery with EQ, but I figure if it comes to that, you've already lost the battle. Again, it's important to remember that your options get increasingly limited the further along you are in the process.
Compression is a good tool to get the parts to sit right with each other, especially in a mono mix. But when you can take advantage of stereo, you get a whole new dimension to exploit. In a tune like this with several guitars at once, it's easy to get a pile-up of similar-sounding tones. Compression is one way to separate them out dynamically and, to a lesser degree tonally. EQ is another way to give each instrument its own space. But once those are exhausted it's time to turn to the stereo spread and separate the instruments out spatially.
I knew that the lead needed to be straight-up in the center. That's the main focus and it needs to be right where the listener expects it to be. The stereo tape delay and reverb I added spread the lead out a bit from dead-center so I had to be aware of making it sound big without taking over the entire left-right spectrum.
Since the tone of the lead and opening rhythm parts are so similar (both used the same patch on the Kemper), I decided to push the rhythm track out to the sides. At first I tried a widely-spaced stereo delay on the rhythm tracks but it muddied up the mix and was distracting. So I copied the rhythm track and panned the first one to about 9 o'clock and the second one to 3 o'clock.
Now you might think that there wouldn't be any sonic difference to a single mono track and two of the exact same singal panned to different sides, but there is. You can here it at the bridge when the band drops out and a single rhythm guitar plays those chunky chord fragments. It's especially clear with headphones and adds, I think, a nice touch of stereo spread to the overall sound without something as blatant as a stereo delay.
Compression is good for setting some rough boundaries on levels as well as "warming-up" a sound, but it doesn't remove the need to occasionally ride the faders. Frankly a mix with faders set once for the entire song is going to be pretty boring. So the next task was to setup some track automation to add some dynamics to the overall mix.
Like everything else in this mix, I approach fader automation with the attitude of "less is more". It's very easy to get lost in the minutiae of whatever you're working on at the moment and lose sight of the bigger goal. So rather than endlessly tweak volume automation, I stepped back and thought about how I wanted modify the mix to highlight certain parts of the arrangement. Instead of fine-grained curves, I usually had a few hard break-points where volume on a track changed 1 or 2 dB in either direction right on the boundaries of a song section.
I had one particular section that needed a quick bump to highlight an important part of the song. The three chords leading into the break was this little gem of a progression I came up with:
I loved that way the re-voicing of chords that were already in the song did a nice job of bringing the listener down from the high-energy of the chorus section into the break. But in the first mix those chords got lost. So I carved out a quick bump of volume right into the bridge.
Volume wasn't the only component I automated. From the beginning I always imagined the opening chords to sound far off, awash in reverb. Then as the band came in, that reverb would be removed instantly for the surprising effect of immediately bringing the band to the forefront.
As the song enters the second chorus I was already aware of not trying to repeat too much of the first section. I had already come up with a different lead line that made heavy use of double-stops for a different rhythmic and harmonic feel. But I wanted the difference to stand out even more. Here I brought in a very heavy flanger effect on the lead guitar and carefully sculpted the bypass automation for th effect to hit right when those chord-fragments are played.
One thing that I found surprising was how the heavy flanger effect sounded as if it were applied to the entire mix, even though it was only on the lead guitar. I was happy with the overall sound and decided not investigate further, but this probably warrants some further psycho-acoustic investigation.
The last bit of "special effects" was the ending of the song. As the final climax builds with the plucked guitar part and the harmonized lead lines I wanted the ending of the song to be abrupt to catch the listener's attention. During normal playback there is enough delay on the guitar that when the track ended the repeats really stood out and minimized the impact of the quick fall-away.
The solution was to quickly bypass the delay at the end of the track for that "hitting the wall" feeling at the end of the song.
For many musicians, the recording process is mysterious and frustrating. It's as much art and science as learning to play the guitar itself and requires the same dedication and willingness to learn from mistakes. While you may have picked up a few good tricks, my hope in presenting this is that you get a sense of the level of detail and attention required to make good recordings.
I'm no professional, but I know enough to marvel at the depth and complexity of the topic. Hopefully you don't find it terrifying and off-putting but, instead, a challenge and an opportunity to have a deeper understanding of music.