Yes, you can capture a choir in a home studio!
By D. Bruce Moore
I was recently recording a contemporary gospel tune, and thought wistfully, “Wouldn’t a choir sound great with this!” But, alas, where would I find a choir and the space to record them? While my home studio is adequate for small endeavors, I would have to find a choir and rent space to record them. Since my budget amounted to just about zero I wouldn’t be doing a choir in the near future. However, I decided to try a recording technique that creates a choral sound with only a few singers and a small space—and here is how it went…
Prep, prep, prep
The time frame was really tight—three hours to record all tracks. It had to happen during that session, or not happen at all. There are many things that can “torpedo” a recording session, so I did everything I could think of to prevent the session from failing. These things included:
• providing the singers with an MP3 of the song along with sheet music via my website;
• testing the technique (using myself and my wife as singers) in the planned recording space to see if it would work as envisioned;
• setting up and testing all the equipment prior to the session;
• pre-allocating tracks in my recording software, deciding on panning and basic effects ahead of time;
• estimating how many minutes we could spend per track to get it all done within the three hours, to keep us on schedule during the session.
Based on previous experiences I told myself (and the singers) that getting the first take would be the hardest as we would have to get used to the headphones, get the mic levels set properly and just get the singers used to each other. I figured that if we got the first good take in one hour we would be doing fine. The following takes would go much faster.
To get the choral sound out of three people meant having them do something rather foreign to them. Each singer would sing on every part: the male singer in falsetto for the soprano and alto parts, the female singers in their chest register for the tenor part. (I decided I would drop the bass part, partly because I only had 3 sets of headphones and partly because the women could not sing low enough for it.) This creates a similar blending of voices that happens with a choir instead of multi-tracking the same voice on the same part. They would sing each part three times, making 9 voices per part for a total of 27 voices. The choral track could be sung in two parts easily, so I planned to record the first half, give the singers a short break, and then do the second half.
I used the biggest room in my house (which isn’t very big) as the recording space. The redeeming qualities of the room include a 9-foot ceiling, hardwood floors and openings to other rooms at odd angles (reducing the number of parallel surfaces and standing waves). I pushed the dining room table up against the wall and covered it (and part of the wall) with pillows to reduce any direct reflections. In spite of its size, the room created a short but rich reverb which sounds great on vocals (and drums, incidentally).
The three singers would surround the mic, two in front and one behind. I used an AKG C 3000 in hypercardioid mode to pick up all of the singers, including the one singer from behind. (Remember that hypercardioid mics often have a strong pickup lobe at 180º. I used this to my advantage.)
The control room was in the back part of the house, with cables running through the heating ductwork. A talkback mic allowed me to communicate with the singers via their headphones. The headphones were rigged to play the cue track through only one ear. This is significant for two reasons:
• With only half of the headphone providing the cue there was less sound to bleed through to the recording mic
• The singers needed to uncover their other ear to hear themselves. I did not feed their singing back into the headphones because of latency issues (hearing their own voice with a slight delay would be confusing). This worked well, as the room was nice to sing in and because the singers could control how much they heard of the others by how much they covered their other ear with the silent headphone speaker.
2:00—Two of the three singers have arrived
Unfortunately we got off to a late start when one of the singers was half an hour behind. I didn’t waste the time, though—I went through the plan with the other singers and they were solid with it once we were ready to start.
2:30—First technical glitch
The headphones were cutting in and out. A quick piece of gaff tape stabilized a loose connection (note to self…always have gaff tape handy).
It was immediately apparent that, barring technical difficulties, this was going to work! The singers had an amazing blend. Steve Newransky’s falsetto was up to the task and easily balanced against the two women. Sandra Park brought her Mariah Carey-ish style to the group while Tannis Penner’s choral/opera technique provided a rock-solid foundation for the ensemble. Mic levels were tinkered with, a couple of unsure notes were corrected and we decided exactly where to put the consonants.
2:50—Finished 3 of 9 tracks!
In spite of starting later than planned we accomplished more than I planned! Within the initial hour we had recorded the entire Soprano part and were ready for a well-deserved break.
3:00—Record Alto part
For the Alto part I wanted to get more of the alto sound that Tannis provided, so I moved her in a little closer to the mic while shifting Sandra a little further away. Steve stayed put with his booming falsetto. The singers decided that it would help them keep the consonants together if someone conducted, so Steve was nominated for the job. We immediately fired off three good takes in rapid succession.
3:15—Record tenor part
It was time for the women to sing in their “boy” voices and for Steve to open up in his Tenor range. The amazing blend continued as the two women managed to balance off against Steve.
Uh oh. The “Joys of Computer Based Recording.” I told the singers to take a break while I rebooted and tried to remember the last time I clicked the Save button. It turned out it was before I started recording the Tenor part. The tracks were still on the hard drive, but their positioning in the song was lost.
I needed to import the files and try and get their start positions correct. The good news was that I was saving each part (Soprano, Alto, and Tenor) in separate folders with the part name as a prefix. I just opened the “Tenor” folder and saw several Tenor WAV files inside. I imported one and fiddled with it until I got it into the correct position. Seeing how long that took I decided to import the others after the session and continue recording with the single Tenor track for a cue.
We finished the Tenor part. It was time to give the singers a break and for me to do a full backup. I copied everything to do with this song onto another computer through a network connection. Whatever happened from here, I would still have the first half of the choral track.
4:00—Record choral part #2
I took five minutes with the singers to go over the second part, singing it for them and working out where to put the breaths and cut-offs. We tried a take and Steve forgot that he needed to switch back to falsetto—we were back to singing the Soprano, Alto and Tenor parts (in that order).
4:15—Another computer lockup
You start learning to hit the Save button after every take when recording with computers. Fortunately, I had just done so—we were back at it in about 3 minutes.
4:20—Give me a “P”
It’s hard to get a good “P” sound without popping the mic. The singers were about a foot and a half away from the mic which was great for a blended sound and avoiding pops, but not so good for capturing the elusive “P” on the last word of a phrase (“hope”). We hammered away at this to little avail—and would have started doing creative and drastic things had a larger issue not raised its ugly head…
4:30—The “Russian” take
After recording the Soprano part I called the singers to the control room to hear the non-existent “P.” We were all surprised to hear that they were significantly rushing their part—the last couple of takes just wouldn’t work! I muted them and sent the singers back to get the timing right.
4:45—The “Russians” are here to stay
For some reason they could not seem to reign it in. I would have loved to stop and work it out beat by beat, but we only had 15 minutes left and about 1/4 of the song to go before one of the singers had to leave! I listened to the “Russian” takes again. The material had several breaths where I thought I could do a digital edit and move the singing back into time. I decided to take my chances and fire through the rest of the tracks before the clock struck 5!
4:57—Down to the wire
We finished the Tenor part with just minutes to spare, so I got the singers to do another two Soprano takes to replace the extremely rushed initial takes.
And on time. It took serious planning, flexibility, and a bit of a gamble, but we were finished by 5. Tannis said “bye” and ran off to a choir rehearsal while we started tearing down the gear.
With the singers rushing the second half of their track it was time to find out if I really could “fix it in the mix,” as the old saying goes.
First of all, I had nine separate tracks, three for each part. I didn’t want to edit each of these separately, so I bounced them all down to one stereo track. I panned each part hard left, hard right, and slightly off center for a nice stereo spread.
I fired up my digital editor and took a look at the vocal parts to try and isolate the rushing. Sure enough, it was easy to see—the vocal entries were about 1/8th of a beat early in each case. To maintain the overall timing of the track, I cut an 1/8th of a beat from the breath after the phrase and pasted it in front of the phrase. This moved the vocal part over enough to put it in time, but didn’t throw off any of the material that followed.
How about tuning? Tuning was great overall—and a good thing this was, because pitch-correcting plug-ins don’t work well on an ensemble. There seemed to be one voice in the bunch going just a tiny bit sharp in the last two chords, though. My solution was to isolate that section and bring the whole ensemble down in pitch by about 20 cents. The sharpness disappeared and the rest of the ensemble was still close enough not to sound flat.
In the end, did it work?
The planning, preparation and risk-taking paid off! A true choral sound emerged out of three singers, providing a gospel vibe to the tune that took it to another level of musicality—all in the space of a large closet!