I went into the writing session with a list of potential song titles, and this one stood out. “Happy New Me” was inspired by the decision to do music full-time. Co-writers George Paolini, Bill O’Hanlon and I had all made the switch from lucrative careers in different fields to focus on what truly makes us feel alive: music. It seemed fitting to write about it when we got together in Nashville.
Here is a snippet of the iPhone demo from the writing session:
And here is the final product:
After the session, Bill said it was up to us if we wanted to do anything with it, and he was not interested in participating in the recording. A perfectly reasonable approach if you just want to focus on songwriting. We agreed and parted ways.
George and I felt the concept was strong, but weren’t sure it was melodically our brand, or contemporary enough as is. It had too much of a 70s folk-rock feel for a contemporary Christmas song. We almost passed up on recording it until George had the idea of going for a Motown vibe. We had been listening to a lot of Motown lately, and were inspired by the sheer number of amazing female acts that made the sound such a classic. With a few minor changes to phrasing and groove, we felt it could fit on a holiday party playlist.
George did most of the pre-production. He picked a reference track, and a drum loop to get started. He added a bass track, an organ, and a Motown-style upright piano. Once we confirmed the key would work for me, he added a sparse guitar ringing on the verses, and a rhythm guitar to the chorus. It was then ready for a scratch vocal.
I started with a scratch lead. Once that was done, I went straight into the vocal production, adding harmonies on the chorus and a few key words in the verses.
I introduced some call-and-response background vocals in the chorus to fill the space in between the lead phrases.
One of the most fun parts was to try a few different ad libs for the last chorus.
Finally, I added some long “oohs” and “aahs” for contrast and support in the pre-chorus and a few other spots. George contributed a few harmonies too.
There are more and more layers as the song progresses, ending with a big stack on the last chorus. Songwriting coach Robin Frederick always encourages us to make sure each section sounds different. One should be able to tell the difference between the two verses, or the different choruses. The vocals are only one of the layers in which I ensure this happens.
In addition to being my favorite part to do, I also start with the vocals because I know I’m going to have a lot of harmonies stacked and want to make sure I create the rest of the instrumentation to support them instead of fighting them.
Instrumentation & Editing
First, I replaced the drum loop by recreating the groove in MIDI and replacing some of the stock sounds to get the tone I wanted. I recorded shakers, tambourine, sleigh bells, hand claps and layered them with MIDI. I also experimented with a few small percussion instruments I have lying around the studio. Once I had the basic outline of the arrangement, I meticulously edited the percussion group such that, if you only played the drums and the vocals, each section would be easily distinguishable and the percussions got bigger and fuller as the song progressed.
I recorded an actual toy xylophone for the glockenspiel. I used a MIDI vibraphone and tubular bells to harmonize or double the glockenspiel in different registers.
After playing with the bass for a while, I decided to go for a short muted rhythm instead of sustained notes for the verses to give the rest more space.
Next to the drums, the horn arrangement took the most time to get right I think. George created the arrangement, but I struggled to get the articulations right during the editing process. I finally got a little closer when I hooked up my keyboard faders to automate the swells. Samples tend to sound good in certain registers and not so good in others, so my solution was to cut out some of the parts that weren’t sounding quite right no matter how much I muck with the articulations and automations. Sometimes less is more!
The verses, though short, were feeling a little sparse, so I added a toy piano, finger snaps, strings, and a few other little things.
I decided to also add strings to the chorus to fill it in a bit more. The string arrangement is not quite traditional Motown in style, but I felt it added just enough warmth to round out the sound.
Generally, the editing process is where I tend to go down a rabbit hole. I’ve been told by mentors to make decisions and commit to them, so I tried to do so on this song. I also kept that in mind when tracking vocals, and only redoing a part when I felt I needed a better take. After comping the guitar and vocal takes, it was time to tune the vocals.
I’ve gotten pretty good at singing harmonies with the exact same phrasing so they would line up right. Pitch-wise, it was close enough that I could more or less do a quick pass with Flex Pitch at ~80-90% and only had a few things to manually adjust. Still, with 22 vocal tracks total, it did take a few hours to complete.
Once that was done, it was time for mixing.
I first started out with gain staging so everything, without plugins and effects, was averaging at -18 dB and peaking at -6 ish. This is because digital plugins work best at that level.
Most of my mixing time was spent carving out little EQ pockets for everything. When I don’t know what will sound good, I swipe around to find the area of the spectrum where I think the sound best represents the instrument in question, then try to cut this area from anything else that also lives in that space.
The arrangement plays a huge role in making this possible. Some of the adjustments I had to make to the arrangement included moving the piano an octave higher, and doing similar octave changes in the horns to help each shine in its best register in the song’s context.
Throughout the EQ carving, I had everything panned center and the output set to mono. Once I was in pretty good shape, I then panned. The only trouble with this method is that some edit issues didn’t show up until proper EQ revealed them, causing unanticipated delays. Still, I was rather pleased to be able to hear everything without any panning. The result sounded great in both mono and stereo.
Once I started panning, I was really happy with the contrast in stereo image between the verses and chorus. Another layer of distinction between the two — yay!
I made a few minor edits to introduce drops, swells, risers, chimes and other ear candy along the way. I wish I could have finished all the edits before the mix, but I always find myself getting new ideas for edits while I’m mixing.
Next was automation. Starting with volume, I used automation to draw attention to particular instruments in different spots. This is a tip I got from Robert Venable when he listened to one of my previous mixes. Robert is a GRAMMY-winning producer who has worked with Megadeth, Kelly Clarkson, Twenty One Pilots, and many, many more. So when he said I was playing it safe, and should go bolder with my attention-grabbing moves, I took note and did as prescribed in every mix I’ve done since.
During the mixing process, I constantly referenced multiple other tracks with similar vibe, from different eras. This helped me gauge the amount of reverb, and the relationship between the snare, vocals, and other parameters.
One other technique I tried in this mix, is to gain-match every change, not just with compression. That and adding SonarWorks Reference 4 to my workflow made a huge difference in the clarity I was able to achieve.
I posted my mix in a couple of feedback groups, one at the beginning, and one at the end. This helped me notice anything I may have missed due to ear fatigue.
Finally, at around the 4th or 5th revision, I started noting down every single change I made. This had two effects:
- It forced me to be more intentional and focused on each change, rather than just twiddling with knobs.
- Given how much time I had spent editing, it was entirely possible my ears were fatigued. Should I regret the changes later, I could easily retrace my steps.
- Side note — ear fatigue is a real issue. On one of the previous mixes I’d done, I realized the next day that I’d made terrible decisions because of ear fatigue.
This is the first time that I actually tried to master a track myself, and not by just throwing Ozone on automatic. It took a long time but hopefully it translates well.
First, I listened through to identify any areas that might need some gentle EQ and compression. After these tweaks, I loaded up the stock EQ and compared the sound distribution for my track versus the reference. I found I had a tad too much sub-bass, so I tweaked that in the mix itself.
Lastly, I added a limiter and cranked it until I was getting 2-3 dB of gain reduction, limiting the peak to -1.0. I used PureMix.net to check whether my master would suffer any compression or limiting from Spotify, Apple Music, YouTube, Tidal and SoundCloud and tweaked as needed. When all the boxes ticked, I tested the master on the car speakers, phone, laptop, Hi-Fi system, earplugs, monitors and headphones.
Were there things I wish were better? You bet. But at some point you just gotta ship something.
I tend to work at relatively low volume, because I work mostly at night, and also loud headphones hurt my ears. So one thing I wish I had paid a bit more attention to was how the mix sounds when it’s cranked super high. Oh well. Next time!
If you found this helpful, share it with a friend! That’s what friends are for!
Also, it would be really awesome if you’d follow my blog by subscribing below. I write about sync licensing, songwriting, production and the business of music.
Did I miss anything? Let me know in the comments below. You can also find me on these other platforms:
So… What do you think?
Would love to know what you all think, and of course would appreciate any streams. Happy holidays!
In the playlist below you can hear it in the context that it was meant to live:
Finally, here is a playlist of all the holiday music I have made. I produced and mixed a good chunk of them, but not all.