Another journey through the valley of production hell. I started late on this one, but couldn’t put it away, and ended up not only violating my 2:00 AM stop time rule (I think I finished around 3:00), but calling a mulligan the next morning.
The opening pad sound got looped while I wrote the lyrics, but I didn’t have any melody in mind while writing. That’s not typical, and it came back to bite me during recording. The lyrics are maybe fourth or fifth revision, as I ended up shifting lines around while recording because they didn’t make sense, or had more emotional impact in a different order. I even ended up completely changing the rhyme structure for the chorus because it just wasn’t working.
- What If Everything You Know is Wrong Ray Toler 4:51
The songs I’ve written where everything just pops into my head and I race to get it captured before it disappears are typically the strongest. If it’s not working, and I continue to fiddle and tweak and edit, the results are normally pretty disappointing. If it weren’t for the Song-A-Day challenge and deadline, I probably would have walked away from this one about halfway through and shoved it in the “In Progress” folder. In all likelihood it would stay there until the next hard drive crash, never to become a real song.
But Song-A-Day is a demanding mistress, so I kept with it. With the lyrics written, I decided to get the structure nailed down. I knew how many measures each part would take, but I didn’t know what the chorus was going to sound like, so I left that area blank and just recorded the drums.
There are four discrete drum loops that show up in various combinations throughout the piece, and aside from the vocal work, recording them took a while. Because the sound is very reliant on the effects of the plugin I was using, I ended up recording long takes instead of just recording the two or four bar loop and flying it around.
With the drums set, I decided to figure out the chorus next. I still didn’t have any melody but was just kind of singing all over the place. In the end, the chords of the chorus drove the melody instead of the other way around.
Next up was the bass. I knew I wanted a slow, legato sound, but I’ve been doing too much of that low sub-bass thing this month, so found something with higher harmonics. I was able to control some of the swells with note timing, and worked to get a decent, but not distracting amount of variety into it.
Time for vocals. And this is where things really started getting slow and frustrating. I couldn’t find the melody anywhere. It was either too low, too monotonous, or not monotonous enough, with me flailing around like an R&B diva trying to find the root pitch. At least my hand wasn’t waving around like an idiot.
Finally one of the candidates beat the others into submission, but my backwards composition process came back to bite me again – the final chord in the chorus didn’t resolve back to the root, which is emotionally unsatisfying for the listener. Sometimes that’s what you want, but it sounded awful in this case. I was able to save it by repeating the final line in a quiet single voice. It all worked out, but it took too much work to get there.
By this time, I had vocals for the entire song recorded, though they wouldn’t survive the night. I had enough going to finish the arrangement, though. The chorus needed some oomph to raise it above everything, and the string patch I found had some nice high end harmonic weird stuff going on. The collisions sound like I’m playing individual notes, but they’re just sustained chords with the sounds doing all of the work. The textural noise in verse 3 was added to provide some additional interest, and I decided it was enough.
But man, were those vocals bugging me. I tried tuning them, but there was no saving it. So I rerecorded the entire thing. I should have been standing for it, but I just stayed in the chair, mostly because I was singing softly and didn’t hear how bad it was until I switched back from headphones to the monitors.
With Mary sleeping, I was trying to keep volume levels, low, but even at low volume, the second set of vocals weren’t right. I went through one more time, then decided to double the vocal on the chorus.
Here’s a dirty secret of the recording industry – you can make entirely new tracks from existing ones with the right tools. I made a copy of the chorus vocal, then ran it through Melodyne, doing a lot more tuning than I typically do, including timing changes, to get it to sound like I’d sung a second take. I thought this would be faster, but it ended up taking me a lot longer because I didn’t mess with it enough, and kept getting phase issues.
Whatever. It was almost 3:00 AM. I threw a microshift on the doubled chorus, ran the whole thing through Ozone, and posted it.
The next morning, listening to it at a higher volume, the mix was pretty bad. Vocals were too hot, compression was overly aggressive both on the vocals and on the full mix. In industry terms, I had tried to polish a turd. It was sort of shiny, but it was still a turd.
I re-recorded all of the vocals. Yes, all of them. This time, I stood up, sang more accurately, and even without any tuning it was significantly better than what I had done in the wee hours of the morning. My singing has been getting better over the course of the month (who would have thought – practice actually helps), so I didn’t have to do major surgery, just taming vibrato or fixing intonation here and there or fixing a scooped sub-tone type of thing.
Perhaps unadvisedly, I also ripped out all of the processing plug-ins and started doing more manual EQ across the entire mix. I also remastered the entire thing. I’m still learning to master, and I’m not sure I got the entire mix loud enough, but it’s in the ballpark. I did A/B my manual mix with what iZotope had done automatically, and I was generally pleased with my work.
In the end, I like the track well enough. It may grow on me over time, or I may chalk it up to another good training project. I can hear some really nice background vocal possibilities in it, so it may be worth revisiting in the future. It was definitely an exercise in writing the wrong way, but demonstrated that it is possible to pull a rabbit out of a hat every now and then.
What would happen if you just let go
Let the granite flow
What would happen if you didn’t know
What you think you know
What would happen if you found out
That they’re not devout
Do you think that you could learn to doubt
Or do you only shout
What would you do if none of it were true
What would you say if it all went away
What if you’ve been lied to all along
What if everything you know is wrong
What would happen if you could see
All the secrecy
Do you think that you could set yourself free
Or would you let it be
Could you bring yourself to just let go
And sink into the undertow
What would happen if you didn’t know
What you think you know
Copyright © 2019 Ray E. Toler, Jr. All rights reserved.
- Warm Pad: Kurzweil K2600XS
- Strings, Bass, Textures: Spectrasonics Omnisphere
- Drums: Spitfire Phobos
Next up: Troll