In over 40 years as an audio professional, I’ve never seen anything quite like the sea change in technical creativity and resourcefulness that’s emerged from the social distancing measures imposed on most of us in early 2020.
This article chronicles one such leap made while many businesses, schools, and houses of worship have been closed or seriously hampered by the inability to assemble. Fortunately, some churches can offer live streaming of their worship services. Shadow Mountain Community Church (SMCC) in El Cajon, CA is a good example.
SMCC easily classifies as a megachurch, with an average weekly attendance of roughly 7,500. Most weeks, the church offers a blended worship experience that includes a 120-voice choir, 30-piece orchestra, eight-person rhythm section, and six to eight front line singers.
For SMCC, the challenge became how to retain and present as much of its regular worship experience as possible while complying with the necessary social distancing guidelines. On Easter Sunday this past April, the choir, worship team, and technical crew – led by Tobin Davis (music minister), Michael Sanchez (music director), Byron Mathe (production director and front of house engineer), and Troy Dausend (television production director) – did something remarkable: They presented Andrew Peterson’s “Is He Worthy” featuring a 90-voice virtual choir.
Watching that service, I was so moved by what I saw that morning I couldn’t help but wonder how they pulled it off. I knew I had to find out and share this with others. My first call was to Byron, who filled me in on all the technical details. Shortly after, I called my good friend Jeff Schmitz, director of engineering at Sound Image (Escondido, CA) who’s also a member of the SMCC choir. I wanted to learn more about what this was like from the choir’s perspective.
Initial Challenges
The virtual choir had 94 initial participants, and here’s the remarkable part: All choir members recorded their own cell phone video at home, which was then uploaded to the technical team via one of the cloud storage sites they suggested.
Simply sending in the videos was a significant challenge, because many of the members were unfamiliar with how to upload a large file to a cloud-based account.
Basic upload instructions were provided, but from the perspective of someone who already has a cloud storage or FTP account rather than from the perspective of someone who’s never done this before. Regardless of other talents, not everyone has fluent computer and IT skills. Numerous choir members needed assistance to successfully deliver their clips.
An even bigger challenge was self-consciousness. Many in the choir just couldn’t believe how poorly their voices were coming across as a raw, fully exposed vocal track. This theme harbored much frustration.
Jeff describes it this way: “For most of my life I’ve been involved with live and recorded music as both a singer and musician. This assignment was way harder to do than I expected. All of my many practice efforts sounded bad. The main problems were pitch, slurred words, and that I was singing alone. I couldn’t judge how to blend without hearing others around me. It was really tough to do. I’m used to singing with instruments and other voices, which help hide my flaws. This was a stark-naked track of my less-than-perfect voice.”
Of course, this brings us to a common theme with choirs in general: With less than about 10 or 15 voices, everyone needs to be a respectable singer (at least) for the group to sound good. With a mass choir, most minor flaws are homogenized into a recipe of beautiful music, providing everyone’s singing the same words at the same time.
Each choir member received an email with a link to an attachment and the following recording instructions. The attachment contained a reference video of Tobin conducting, Michael playing piano and singing a scratch track, and a click track.
1. Play the reference video on a desktop, laptop or tablet computer and listen using earbuds or headphones.
2. Use a cell phone, placed in portrait orientation, and record a reasonably tight headshot while singing your part. Use a white background if possible.
3. Rehearse and record as many clips as needed until you get something you think is worthy of submittal.
4. No special microphones needed. Just use whatever is built into your phone.
5. Once your track is ready to send, upload via Dropbox, Vimeo, YouTube, File Dropper, or any other method necessary to handle large files.
Pre-Production
The tech team accepted whatever file format was sent, and the whole collection was compiled as a production collection in a Dropbox account. The pre-production staff, led by video editor Justin Kintzel, then began the tedious process of compiling all the tracks. While there were 94 submittals, only the best 90 could be used. Once all the choir clips were reviewed they were loaded into Apple Final Cut Pro and Logic Pro for syncing and mixing.
Lining up the tracks was a little harder than expected because there were no film clapper boards or other “start” markers at the beginning of each track. Just dead air before the first words were sung. As Byron notes, “If we ever do this again, we’ll have them give us a hand clap or two up front, at the beginning of a specific measure, so we can more easily align the clips.”
The choir doesn’t enter until the fifth measure of the song. Also, the first word sung is a mezzo-forte “we” and there’s literally no transient or plosive attack on the letter W.
Once the starting point of each track was locked in, the next step was to normalize the levels on all 90 tracks. No active mixing was done, nor dynamics processing or EQ added. For this project, normalizing just meant the tracks were all adjusted to be of similar volume relative to what was needed for the vocal blend within each voice-range classification, i.e., soprano, alto, tenor, baritone, and bass. I asked Byron about what went better than expected, and he replied, “When blended together, the poor cell phone audio actually sounded better than I would have thought.”
After all the audio tracks were ready, the final mix-down was done as a stereo submix. A touch of reverb was added to the 2-mix bus to further blend the ensemble and to approximate the ambience of the live tracks.
Five more audio tracks were also synced up. The third channel was the click track. The fourth and fifth channels were a submix of a previous live performance of the same song, which Byron says, “needed a lot of cleaning.”
Bringing in the existing track was only possible because the rhythm section almost always works with a click track. By matching the click from the previous recording and using it for the live event, all the critical tempo and timing issues were easily accomplished. Michael and the rest of the band are very comfortable performing with a click track in their in-ear monitors, so this was no different than their usual setup.
The last stereo pair to be synced was a re-mix of the orchestra recording from the previous performance. Byron explains that he used very little of this sub-mix because this year’s Easter service didn’t include the orchestra, so hearing much of it would be out of context with the livestream visuals.
The final video mix consisted of five discrete output buses mapped across five large screens. This is the template SMCC regularly uses. Two large direct view LED screens are positioned on the far left and right sides of the stage. These are typically used for image magnification (IMAG).
Three smaller projection screens are located above and behind the choir. Their system is configured so each projector has a dedicated HD-SDI input, and discretely serves one of the three screens. They don’t use edge blending because the center screen moves in and out when needed.
As you can see from the photos (and a video clip of the virtual choir’s performance is available at Vimeo here), three screens were used for lyrics and two for the choir’s video box display. By cropping and pixel mapping each choir member’s file, they were able to show their faces in two 45-box images, which neatly filled the mid-left and mid-right screens.
Live Performance
The three SMCC Easter service broadcasts were only available via live internet stream. Further, the onstage worship team was limited, for most songs, to just a four-person rhythm section and a six-person front-line vocal team.
For the “Is He Worthy” performance, the choir video and lyrics were handled by a FOH computer running Pro Presenter 7, with a Black Magic Design DeckLink 8k Pro card. The DeckLink card supports four simultaneous 12G SDI output feeds. Those four feeds were routed to the master video switcher, which also handles all their manned and robotic camera inputs.
Two switcher outputs were routed to the projectors showing the choir’s headshots. The lyric feeds were bused to the IMAG and center projection screens.
The three center screens are 16:9 format, meaning the total horizontal pixel count for all three can easily be 5,760 (1,920 x 3). And, because the DeckLink card supports such extreme ratios, 1,920 pixels can be allocated to each of the three output channels and show a composite, full-motion graphic with a 5.33:1 aspect ratio.
The seven audio channels were fed from a Dante Virtual Soundcard in the production computer to a Dante card in the Soundcraft Vi7000 FOH console.
Prior to the audio channels coming into the console, all timing and latency issues had been resolved. It was then a matter of blending the live band mix with the pre-recorded sub-mixes while feeding the click track to the IEM monitors on stage.
Final Thoughts
SMCC Pastor David Jeremiah’s post-Easter newsletter included these statistical responses: 45,200 viewing sites logged into the Easter service via its various streaming platforms. Further, social media engineers tell us that on average each viewing site represents 2.3 viewers. That brings the likely number of viewers to slightly more than 100,000 worldwide. This is a huge number for any church; showing that even under the most difficult conditions, the human spirit will prevail.
Stay strong, kind, considerate, creative and faithful.