Happy Day is a song that I wrote for a songwriting competition in 2014. While I didn’t win anything, it was remarkable fun to write, record and mix. To make it better I was able to bring on a few new friends to contribute.
This song was written specifically for the songwriting competition. I was somewhat writing with the judge specifically in mind as my audience. The criteria for the competition are very simple – the judge’s favorite songs win. As I sat down to begin working on it, I was attracted to the chord turn-around in the A-section of the old gospel hymn, Oh, Happy Day. If you can listen to the version recorded in the Whoopi Goldberg movie Sister Act 2, it’s pretty cool. (48 seconds into this clip)
The turn-around got me into the theme of a happy day, and knowing that my competition judge is something of a sucker for love songs, I thought about how my wedding day was an incredibly happy day, as are many of the days that I have since spent with my wife. The verses started coming together, and eventually filled out into a standard 12-bar blues, with the gospel turnaround B-section. I’m not sure where I got the idea for the twist at the end where it’s revealed that the singer is actually trying to get his baby to accept his marriage proposal, but I think it’s kind of cool.
I knew this was going to be a sort of piano blues, since I’m a piano player. I could hear the horns early on and knew that I wanted to take a shot at arranging and recording them. The horns are a standard pop horn section consisting of trumpet, tenor sax and trombone. Once I had the bones of the arrangement down with piano, drums and bass scratch, I worked at the horn lines on a keyboard, working with various arrangements and harmonies until I had pulled out of my head what I was hearing in there. It was a surprisingly arduous process, but I’m really happy with the results.
At the time my digital piano was on the fritz, so I was unsure how I would record a good piano line. Eventually I borrowed my daughter’s laptop before she left for college, loaded up the free version of Presonus Studio One, drove to my church and plugged in to their digital piano. I recorded six or so takes, came home and edited them together into what you hear in the recording.
The biggest blessing of putting this recording together was bringing on two of my new friends to play electric guitar and bass. I had the bassist, Dan Short, over to my house to lay down his track. That was fun and I’m just getting to know Dan. He has groove oozing out of his pores and it’s a pleasure to play with him.
The guitar part was laid down by Frenchie in his home studio and the cool thing is that I didn’t produce his part at all. I just gave him the mp3 of the song without the guitar and told him to go for it. What he sent back fit the feel of the song perfectly. He struggled a little with some of the technical complexities like the intro and the horn break near the end of the song, but I was able to cut and paste from his recording to work everything together like he played it all end to end. At its core it’s all him.
I did something a little unusual for me with this mix – I didn’t bounce all the instrument tracks from MIDI down to audio before mixing. You can see on the tracking window, right, that the horns, drums and piano are all still MIDI tracks. I’m also experimenting with different color schemes for the tracks, since I never liked the Sonar default Crayola colors anyway.
At the top is the vocal line, followed by the horn section, the piano and drums, the muted scratch bass line I played in, and with bass and guitar audio tracks at the bottom. Some things to note about this view – since I left everything in MIDI mode you can see some of the controllers I used to program the horn lines. There’s also a little bit of volume automation on the vocal and guitar. The guitar I wanted to feature at different places in the song, and thus used the automation to do that. The vocal I just wanted to boost a little at the end to increase the intensity.
I’m playing with a brand new plug-in for this song. Recently Eventide released their new UltraChannel plug-in for free for a limited time and I grabbed it. Looking at the screen shot you can see that it is a full channel strip with a gate, two compressors, 5-band EQ, pitch shift and delays. You can reorder the effects any way you want and turn them on or off selectively. So far I’ve been really impressed with the quality and I like having everything laid out in front of me where its easy to see and manipulate.
Some notes: I doubled up on the EQ. On the Sonar channel strip you can see that I low- and high-pass filtered the signal to limit the bandwidth, going for that 50’s sound. I again attenuated the low and high in the UltraChannel post the compressor, with a slight boost around 3k. I found this brings out one of the few nice ranges in my voice. Note that I bypassed the delay on the UltraChannel because I wanted to use delay as an insert.
The electric guitar has fairly straightforward processing. It starts with a high-pass filter to take out any extraneous rumble from the recorded track. A quick note on how I do this: I engage the filter with a fairly steep slope (just enough that it doesn’t resonate at the cut-off frequency) and begin sweeping up with the track playing until I begin to hear it affect the audio (it starts getting thinner sounding). I then back it off until I can’t hear the effects any more and leave it there. This process removes what I don’t want to keep and keeps what I do.
The Rough Rider compressor is used to tame the dynamics a bit to get the track to sit a bit better in the mix, and also add a bit of saturation. Finally, the SlickEQ is used to do some gentle tonal shaping. I’m really liking this EQ for this purpose, and it has supplanted my old favorite the BootEQ.
The bass chain is very similar to the eletrcic guitar chain. I didn’t filter out the lows since the bass is holding down the low end. The ThrillseekerLA compressor is set to a fast opto setting, which gives me a little bit more sustain while still letting the attack of the notes through. This is important for bass since the attack is what defines the notes. Without attack, the bottom end becomes smooth mud, which is not what this particular track needs. At the end is the SlickEQ again, this time with a gentle shelf attenuating the very lows, and a slight bump just above 900hz to help define the attack. I found the frequencies and gain amounts by first setting large gains, sweeping the frequency until I found what I wanted to change, and then adjusting the gain until it was right.
By now you should have noticed the Slate Digital RC-Tube Channel on every channel. Long time readers will remember that this is used on pretty much every channel I mix, simulating running the signal through a mixing console. Several people I respect who were huge advocates of out-of-the-box summing have said publicly that this plug-in changes all that.
To emulate that old 50’s rock sound I decided to run the guitar through a spring reverb. This ended up being a little more difficult than I anticipated. I tried several impulse responses (IRs) that I had already but the results were less than spectacular. Even at low levels and with an EQ filter the reverb tended to build up too thick and I was left with a useless mush. Internet to the rescue!
First I tried the Devil’s Tube Spring Reverb, a stand-alone spring reverb vst effect. I couldn’t get it to work for me. After a bit more searching I stumbled across the Adventure Kid impulses, which are actually generated from his own spring reverb simulator, rather than a real spring reverb. Real or not they worked the treat, and I’m glad I found them.
The final element in the mix is the space where the instruments played. Everything in this project was recorded direct (MIDI or DI) except the vocals, so there was no space on any of the tracks. A number of years ago I found some nice live room samples when I was in my IR collecting phase. Unfortunately I can’t find them online, now. I think I got them from NoiseVault, which is currently updating their website, so the IR’s are unavailable. I like the open, spare feel of the space and have used it on a number of other productions.
The stereo buss received the most processing, which is fairly typical for me. There’s still not a ton, though. The major piece is the SlickHDR from Bootsie. It is billed as a “Psychoaccoustic Dynamic Processor” which purports to compress the audio in three different ways depending on what type of peak it presented. Theoretically it can behave like three different compressors simultaneously, and I have to admit that the results are really surprising. He published very simple instructions for it’s use, which I found applied a bit too much compression, however, leading to a fairly squashed sound, inappropriate for the style of music. But, by backing off the drive on the three channels from their “optimal” settings I was able to get a very pleasing result. Tip: use the trim to back the level off to the incoming volume for more truthful A/B comparisons – an increase in volume will be mistaken by your hearing for an increase in quality. (I tried this processor on a few tracks on my Moments With Him project, but it provided too much compression for the folk/classical style of the music.)
Also in the chain are the Slate RC-Tube Muixbuss and Boost11 peak limiter. I needed 4.4 dB to bring the song up to a good, healthy level. It was never a really dynamic song because of the style, which kind of keeps coming at you. At the end is the T-Racks Metering Suite, which puts peak, RMS, phase and spectrum analysis all in one simple package. I probably wouldn’t use this for a song that will be receiving separate mastering, but for a one-off like this it does the trick.
This was a fun track to write, record and mix. It was a bit outside of my comfort zone and I think I rose to the occasion. The horn arrangement came off great and I think that it is my favorite part of the song. Also, it’s better to work with friends. Don’t ever forget that part.