Our choirmaster, having set plans to drive out to Minnesota to visit our beloved St. Gregory's alumni Kerri and Jen at Good Courage Farm, needed some content. In particular, he needed preludes — the music that plays before each service. And so he e-mailed me and asked if I would like to prepare a presentation of the tune "How Can I Keep From Singing?" Perhaps a choral arrangement for three tenors?
I have never done this before, but okay.
I'm not well plugged in to the choral world, but I have done some audio production over the years. I've also done some of these style of recordings with the St. Gregory's choir before. So I had some opinions about the tools we used.
Previously, we've used an app called A Capella. There are a lot of nice things to be said about A Capella app: for one, you can point a chorus member to the app and send them to a link and you're off to the races collaborating. It does the video and the mixing and all that for you for the most part.
Unfortunately, it has other drawbacks. As a singer, if one makes a mistake anywhere in the recording, the whole take has to be redone to fix it.
This might not seem like a big deal. After all, isn't a live performance a single take? But recorded harmony is not created in the same way as live harmony.
Live harmonies, and particularly harmonies sung a capella, leap into creation simultaneously. No voice has priority over any other. When there is an error in pitch in one of the parts that makes the harmonic relation imperfect, the voices adjust to one another to correct it. The harmony is sung together, as a team.
So what about recorded harmony? Recorded harmony is constructed: each note is placed next to the previous note. If they fit together correctly, like a jigsaw puzzle, you get harmony. But making harmony against a recorded part is like playing tennis against a brick wall: it's not sung together, it's sung against a fixed point.
This is harder to do than live choral singing. The team element is gone. If singer 1 records a minor imperfection, singer 2 has to match it to make a lovely sound.
So this problem of needing to re-record the whole take if any error is made is a big one. It is frustrating as a performer, and sucks a good portion of joy out of participating in the larger effort.
A Capella also sucks for mixing. I'll say no more about that.
Anyone evaluating a software tool should also think long and hard about the economics of what they're using, too. And from the word go, I didn't like the economics of A Capella App. Think about it:
I may be wrong about all this. I'm not in the business, after all. Maybe they're making good money. But it looks bad to me.
I ended up going with a pretty standard production workflow: I recorded a guide track with the melody and a click track in Ableton, and asked my collaborators to record on top of it. I would mix in my DAW and assemble the video in FCPX. All our intermediate work products would be shared in a DropBox folder.
I knew that the greatest advantage of this process would be flexibility: as long as I received an audio track and a video track, I could assemble the finished product. This is a drawback for communications, though: "Use whatever you like!" isn't a model of clarity. In fact, one of my initial collaborators bailed: they had only allotted a small time to record, and were unable to overcome the tooling issues. I was able to find a backup, though, and we got to the finish line.
We ended up having two different workflows. The first collaborator used the simpler one: he had a desktop, and was able to play the guide track and sing directly to his laptop. Simple enough! And identical to what would have been done on A Capella App.
The second collaborator only had an iPhone. No computer! I hadn't thought through what this would look like. I suppose I hand waved it away: "Oh, they'll have a second device. Who doesn't have a second device, anyway?" (The answer is: many people.)
Here's what we ended up doing:
Recording the take was straightforward. GarageBand on iPhone can import the backing track from DropBox into track 1. Then, the vocalist puts on some headphones and records their take into track 2.
Then, export to DropBox. I'm still not clear on exactly how the vocalist got this to work. They struggled with the tool, and I was unable to provide any guidance (I have an iPhone, but I never use it, much less GarageBand), and then he got it to work. Okay.
For video, it was necessary to go back and lip sync. Playing the track and recording one's voice at the same time with just an iPhone requires yet another app; you can't do it out of the box. For this, we used Video Star, an app I'd used previously on some Theme Music collaborations. It works well enough; I'm not sure it syncs perfectly, but oh well.
Would I do this again? Would I do it for a larger choir?
Honestly? I'd like to. And I say that knowing full well that there would be some challenges.
First would just be communication. The lack of clarity I sallied forth with into this project would not work. I would want to settle on two recommended workflows: a straight to video workflow, and a multitrack recorder workflow.
The multitrack recorder workflow is harder, but I believe it will pay off. I am certain it will improve the quality of the performances; I suspect it will be more fun for the performers, too. A clear how-to will be required, though. It is more complicated, too, which runs the risk of driving off some collaborators.
It's worth considering ditching the straight to video flow: after all, it's one more thing to talk about. But in this case, I think a fallback is necessary. Some may not be able to get the multitrack workflow going, or may have other unknown technical limitations (like, for example, owning an Android phone).
And what about Android phones or other outliers? Hopefully I won't have to handhold these folks too much. If so, I'll have to come up with a recommended straight to camera app something like Video Star.
And I guess that's all I have to say about that.