Last Saturday night I recorded the Boulder Philharmonic and Boulder Chorale performing Brahms German Requiem. Unlike most of my Boulder Philharmonic recordings, which are a two-microphone M/S set-up, for the German Requiem I had to use more microphones, a lot more.
The problem with recording a chorus with an orchestra is getting the balance between the chorus and the orchestra right. Most choruses simply aren't as loud as a full orchestra, so they tend to get buried during the forte's. The solution in a studio recording is simple - record the chorus at a higher level. But during a live concert that solution isn't so simple. Obviously, you want to record the chorus at a higher level, but when 108 musicians are shoehorned onto the stage, isolating the chorus isn't so easy. Combine this close proximity with the fact that usually the brass section is located just in front of the chorus and you have a tricky mic placement problem.
Normally I record the Boulder Philharmonic solo - I get some assistance hanging and removing my microphones, but the rest of the process is a one-person job. But for the Brahms Requiem I called in some assistance in the form of Fergus Stone (no relation except in spirit) who regularly records the Boulder Chorale and owns an independent recording studio in Boulder. We discussed the recording and decided that he would bring a pair of Neumann microphones and a pair of two-channel Grace M-201s to augment my Schoeps Collette/Grace V-3 combo. We also brought about 400 feet of mic cables and several mic stands.
When we arrived at the venue, University of Colorado's Macky Auditorium, we discovered that the house sound guy had already placed and wired four microphones for the chorus sound reinforcement, arrayed across the stage in two pairs. Fergus and I looked for a spot to place his mics and after about fifteen seconds decided that there were no spare spots, so we agreed to take a patch-in feed from the already-positioned mics. Doing so saved us quite a few minutes of cable-laying time, but we managed to gobble that extra time while hanging the main Schoeps microphones when the dual shock mount that held them came partially undone and I was forced to climb down from the four-story high catwalk to reassemble the shock mount. By the time we got the main microphone positioned the Friday evening dress rehearsal had already begun.
Since both Fergus and I brought two-channel recorders, we had to mix-down our six channels to two "on the fly," during the live recording, meaning we had to get the live mix right since that would also be the final mix. Obviously this increases the level of difficulty some. Our signal chain was relatively simple - Microphones going to Grace microphone preamps, six preamp's line-level outputs going to a Mackie 1202 VLZ mixer, and the two-channel mix-down going from the Mackie to our recorders. The tricky part is determining how much "chorus mics" to add to the main mix.
We spent about forty minutes listening to the rehearsal while adjusting our microphone levels. The final mix was primarily the main microphones, with only about 15 to 20 % of the sound coming from the chorus microphones. In an ideal world I would have preferred to boost the chorus levels up to 25% of the mix, but that pesky problem of the horn section bleeding into the chorus mics precluded using any higher levels.
The Brahms German Requiem also includes two soloists, a baritone and a soprano. Usually I put up a single soloist microphone to highlight the soloists slightly, but for this performance neither soloist wanted a microphone, and after listening to the rehearsal, it was obvious that neither of them needed any sonic augmentation since the main stereo pair was picking them up quite nicely.
The recording on Saturday evening went smoothly, with no equipment meltdowns or malfunctions. The mix I heard through the Focal XS Books that I use as on-location monitors sounded about right, with good image localization and balance, as was the sound through my Grado RS-1 headphones, so I considered the recording a success, despite the set-up difficulties.
On Sunday morning I spent about five hours editing out the extra time between movements and then labeling and reassembling the separate movements into CD length sections. At slightly over 70 minutes, the Brahms German Requiem required it's own 80-minute CD, while the first half of the program went onto a second CD, and the pre-concert lecture went onto a third CD. But what takes the most editing time is converting my DSD 5.6 MHz (128x) DSD files into other formats. I make a complete set of 2.8 MHz DSD, then 192, 96, and 44.1 PCM files from the original 5.6 DSD files. Finally, after all the conversions are finished I create CDs from the 44.1 files. The higher rez files are primarily for archiving and my own listening.
So, how did everything turn out? Not too bad, actually. In a perfect world I would have preferred a slightly higher chorus level, but given the real-world limitations of their physical location, I think we coaxed as much level as we could without screwing up the orchestra's imaging or getting too much horns in the mix.And that's how I spent my Saturday night (and Sunday morning.)