(A Note About the Vocal Audio Samples Below: With the exception of the “FINAL MIX”, I attempted to make all the vocal examples sound similar (in panning, reverb, and level), and they ONLY include 4 of the 8 voice parts I eventually recorded). The “FINAL MIX” is an excerpt from the finished track as it will be released.)
Ok. So after months of work I finally had a piano track! Now came the easy part, right?
Let’s review the steps:
1.) Buy the Music.
2.) Learn how to play piano.
3.) Learn how to play “Dirait-On”
4.) Record the piano part.
5.) Learn how to pronounce the French words.
6.) Learn all the singing parts.
7.) Record eight versions of me singing all the parts.
8.) Do post production work to correct any syncing issues, tuning problems, noise removal, etc.
9.) Mix the song.
10.) Master the song.
In this installment we’re going to focus on:
Steps 5 Thru 10 : Recording the Vocals
STEP 5 (Learn the French Lyrics): Nah.. ths is boring. I’m sure I mangled the words anyway. Let’s move on, shall we?
STEP 6 (learn the parts): Since “studio time” in my own house is free, I did this as I went. 🙂 I sung with the music in front of me, and for the first few days of recording, I found myself trying to sing along with 5-minutes-ago-Scott and realizing the moron didn’t even know the words! I battled with myself in this way many times over the course of recording–to the point that it became comical. Not only did 12-minutes-ago-Scott screw up the rhythm in the 2nd verse, 2-weeks-ago-Scott recorded a piano section that was flat out wrong! That f***ing BASTARD! Why 1-week-ago Scott didn’t figure that out I had NO idea. ::groan:: Would have saved us ALL a boatload of time and aggravation. Anyhoo…
STEP 7 (Record eight versions of me singing all the parts): This was the real challenge of this whole thing. And as embarrassed as I am to provide you with some early examples of my attempt to record the vocals for this piece, I will do it anyway. I started with what I had–a dynamic, cardioid Sennheiser e835 microphone running into an M-Audio Profire 610 audio interface used as a mic pre-amp only, and then into my Yamaha Motif XF8 workstation keyboard, which functioned as my ACTUAL audio interface. This initial set up (specifically the mic) made things EXTREMELY difficult, and I’ll explain why after I grit my teeth and provide this… less than stellar example of my 1st attempt.
Now, aside from being badly out of tune and having TERRIBLE vowel singing, there were other major problems with this that aren’t as easy to identify in this particular section. The REASON it sounds so… forced and out of tune is that I had decided to try recording the vocals from a decent distance away from the microphone. This was an attempt to avoid getting too many mouth noises (saliva pops and clicks from opening and closing my mouth, etc.). It was also an attempt to minimize proximity effect from the mic (the extra bass you get when the sound source is placed close to a cardioid microphone). For some reason I thought it wouldn’t be very “choral” to have too much bass (I realized that fear was a bit misguided later). Finally, I thought a little more room sound (in addition to the inevitable reverb I would add later) would MAYbe help with gelling the eight vocal parts together. This decision turned out to be problematic in many ways, the most insidious of which is that because my mic wasn’t very sensitive, I found myself having to sing too loud and in a… less than relaxed way. This created tuning and expressiveness problems because my forcing the volume a bit was making it harder for me to stay on the pitch.
I exacerbated the problem by overcompensating and cranking the pre-amp gain up as much as I could. This resulted in a LOT of noise. I was in fact, just setting myself up to fail by adding solutions that ended up creating more problems. All of this would have made little difference if I was doing a single vocal for a pop song with a drum track and lots of instruments, but my plan was to include EIGHT different vocal tracks in the final mix, and at some points the piece would go to pianissimo accapella with no accompaniment at all. That hiss multiplied by eight combined with me singing super-soflty would be a total disaster. And it was after I recorded the whole song like this, including the final, brutally hard, accapella section at the end, that I started to seriously worry about the successful outcome of this project.
Eventually, after a bit of research, I decided my mic was simply not the right tool for this particular job. I went out and invested in a new Sure SM27 condenser mic. It was much more purpose-built for what I was attempting, and as such, it was MUCH more sensitive, which allowed me to turn down the gain on my pre-amps and eliminate practically all extraneous noise. I hoped that, along with singing a bit closer to the mic, would be enough to fix most of my initial problems. Here was the result:
Better right? Not great, but better. The signal to noise ratio went WAY up, and that alone was worth the investment, but it also just captured the vocal differently: a warmer, fuller sound that had more breadth and texture, and somehow just made the vocals mix together better. I was much more optimisitc about the outcome of the project.
I still had much more work to do, however. It was time to really get down to the nitty gritty about how I wanted to perform this piece. What were the dynamics? What notes did I want to emphasize. How freely, or rigidly did I want to sing it? More rigid (striaght line notes with no vibrato) meant more potential for perfect tuning, and a “controlled” sound. Whereas a more free interpretation would mean less than perfect tuning, but more room for dynamic and emotional expression. My first instinct was to go the more controlled route–only because choral music that is not sung in tune tends to drive me bonkers.
Given my decision, I worked for another week or two on the vocals, and here was the result:
Alright! Now we’re cookin’ with some gas!
STEP 8 (Post-Production Clean Up): I worked long and hard on this version. I did a lot clean up work on these tracks. That included a LOT of micro-edits to get rid of those annoying mouth noises. I also did some audio warping in order to get the parts synched up a little better. Finally, I did a little pitch correction because I don’t care who you are, hardly anyone sings a vocal part perfectly. And since this is a studio recording rather than a live performance, perfection (balanced with expression) is the name of the game.
I was so pleased with this version that I started to think, “Hey, this might be worth a $.99 download to SOMEone… maybe?” And with that in mind, I thought I better take it to a professional to go the rest of the way. It was Friday, and I was set to go to the studio on Monday. It was then that I listened to the whole thing one last time and thought, “You know… I think I messed up.”
STEP 8A (THROWING IT ALL AWAY AND STARTING OVER): I know, this wasn’t one of the original steps, but sometimes… you just can’t plan these things. Remember that decision I made about well-tuned “perfection” versus expressiveness and dynamics? I called that into question at the 11th hour, along with having a problem with the out of control “ee” in my “Dee”rait On vowel choice. I decided to spend the weekend recording an entirely new set of vocal tracks with a little looser feel–a lot more dynamics, and a little less “EE” and just see how I felt about it.
It took all weekend to not only record the vocals but do the post-production work I needed to get it studio ready, but the end result was a much more emotional performance. Although I ended up taking both sets of vocal tracks to the studio, in the end I decided to go with the latest and more expressive ones I’d recorded over that final two days.
STEPS 9 AND 10 (Mixing and Mastering): This was done with Jim Pavett at Allusion Studios in Tucson, Arizona. That was fascinating process unto itself, but I won’t bore you with the details. It took about 3 hours, and Jim gave me the best compliment I could have been given–which was that my vocal recordings were so well done that he had no suggestions about how to make them better other than to stay well hydrated and keep apple juice handy while recording. He said that would help cut down on the saliva clicks and pops I was plagued with throughout the process.
THE END RESULT: So the irony is that, in the end, it took only 4 days to go from piano track to fully mixed and mastered, eight part harmony. But it took 5 months to educate and prepare myself to create those final tracks. Here is the same excerpt you’ve heard before, but now with the piano, in the final mix of the song as it will be released on December 14th, 2011:
Look for it on iTunes starting December 14th. I hope you’ll check it out. Thank you VERY much for reading about the process I went through to create it. Next time I’ll explore some of my adventures trying to learn how to create electronic music, which is an altogether different challenge.
Comments are closed.