Yo-Yo Ma performing with the Colorado Symphony Orchestra and the Neumann U67 (foreground).
Photos: Darius Panahpour
When the Colorado Symphony Orchestra (CSO) began to feel an economic pinch this year, promoters came up with several novel ways of addressing potential budget shortfalls. In addition to across-the-board pay cuts, the CSO teamed with Colorado Public Radio for a three-day pledge drive that culminated with a concert by acclaimed cellist Yo-Yo Ma and the CSO.
CPR broadcast the concert live to 22 radio stations in Colorado in DTS Neural 5.1 surround. All told, the drive raised more than $600,000. For the broadcast, CSO chose independent consultant Mike Pappas to oversee the recording; Pappas has previously worked with the orchestra and often works with KUVO, Denver's 24-hour jazz public radio station.
Pappas has been recording since “I sold my car and bought a Scully 280 2-track, ¼-inch tape recorder that I used to carry in a flight case and lug around to gigs,” he recalls. “It weighed like 90 pounds. I had 14-inch reels on it so I could actually record for an hour.”
Mike Pappas mans the DiGiCo board while David Day and writer Candace Horgan look on.
“[CSO] called us a couple years back to do 10 days of Beethoven, so we loaded in on May 28 and were out on June 13," he says. “There were usually two performances a day, or a rehearsal and a performance. None of those rehearsals and performances were the same, so you'd rehearse one thing and they'd be performing something different that night. I think we had, at one time, 12 different sets of tape marks on the hall floor as to where microphone moves were for specific events. We had to take the stage apart from the morning rehearsal and then put it back up for the evening performance.”
The symphony's Boettcher Concert Hall is idiosyncratic, with several problems an engineer has to consider when setting up a recording. “The hall is in a really weird spot,” Pappas says with a laugh. “Half the audience is behind you, and none of the seating areas are symmetrical. It's hard to find a centerline in that room. It has quirks, like players on the right hand of the stage can't hear players on the left hand of the stage. All that plays into the gear we use and how we get it to work.”
Classical Miking Scheme
On a tour of KUVO, Pappas' assistant Will Barnette unlocks the gear cabinet and reveals a Neumann-lover's paradise, with M150s, U87s, a U67 and more. For the Yo-Yo Ma concert, Pappas turned to a prototype Neumann microphone, a KM133D digital model. The mic uses an M50 titanium capsule, which Pappas says is ideal for many classical applications.
“It's mounted in a sphere, so what happens is at low frequencies, the microphone is an omni, and as we get above 1k, the mic develops a pattern and it's cardioid, and then above 16k, it's a hypercardioid,” he says. “What you can do with it is use the fact that it's an omni at low frequencies to integrate some of the room tone. The other cool thing about omnis is they are extremely flat in their frequency response. Those capsules on those digital mics, or the M150s for that matter, are 3dB down points at 2.5 Hz. We have lots of low frequencies, and that's good because you don't want to have roll-off that's going to look back anywhere into the audio passband. You want to keep that as far out of the audio band as possible, because the minute you have a roll-off you have a phase shift. Omnis are cool for that, but the problem with omnis is they are omnis, so you want something with some directivity at high frequencies because it allows you to aim the mics into the orchestra and use that directivity to help bring out certain sections of the orchestra.”
Pappas used a total of five in front of the stage, using tape to mark the placement of each to within a quarter of an inch as they had to take each mic down after each morning's rehearsal. The mics were placed 10 feet, three inches above the stage, aimed slightly down at a five-degree angle. Barnette and Joey Kloss helped Pappas with mic placement and removal after rehearsal.
David Day, freelance writer Gary Krakow and consultant Mike Pappas discuss mic placements.
The digital microphones have an NMC DMI-8 8-channel control box that feeds into a Macintosh computer running RCS software. The control box provides phantom power, in this case 10 volts at half an amp, because the mics have a gain stage, A/D chip and DSP chip inside the bodies. From the computer, via the RCS software, Pappas could set many parameters, including compression and limiting, though he used neither. The control box also lets you set the sample rate, which in this case was 24-bit/44.1 kHz.
The reason for the lower sample rate is because the DiGiCo DS-00 broadcast console is limited to 44.1. Pappas has a new SD-7 on order, which will be configured to record up to 24/192 and have built-in Neural Surround. For the CSO broadcast, the DS-00 was used to route the signal.
“We have every imaginable facility on the console, but we don't use it,” Pappas says with a laugh. “For big symphony projects, we don't use any EQ, dynamics processing or the eight stereo reverbs. It's a big level control and router because we route stuff to get it to the broadcast guys.”
In addition to the KM133Ds, Pappas used a stereo dummy head, the KU100 nick-named “Fritz,” in the back of the hall hanging from the ceiling 10 meters above the floor to create the surround rears. “Fritz is an analog box and has standard [analog] outputs,” Pappas says. “We try to not run analog mic cables very long. The cabinet right above Fritz is a Grace 802 unit with a built-in A/D converter that outputs AES. We feed that AES signal down 400 feet of Cat-5 cable.”
In determining how to mike Ma's cello, Pappas had considered employing a KM140 figure-8 in cardioid in a mid-side configuration so that he could control the stereo width. However, his assistants persuaded him to use a Neumann U67 running into a Grace 802 unit in the stage rack above the DiGiCo controller.
When mixing the show, Pappas had a hard-right/left pan of the outside microphones, and the middle three were mixed Lc/C/Rc with the U67 mixed just left of center. Fritz was used for Sr and Sl.
Getting the Right Cables
Not many engineers are concerned with things like audiophile-grade cable, but Pappas has 3,000 feet of Cardas microphone cable for his setup. “We did studies of cable capacitance and the microphones' ability to drive live cables,” Pappas says. “Where this came into play was with applause on live recordings. Typically, you'd have 150 to 200 people and it sounded great; you get to 1,500 or 2,000, and it sounds like crushed pink noise.
“We came to the conclusion that it's not a frequency-response problem or a filtering issue; the microphones run out of current to drive a load so you get this insidious slew-rate limiting distortion,” he explains. “What's happening is the amplifiers in the microphones can't deliver enough current to charge the capacitance at a high level with high-amplitude signals. Once we figured that out, it was an epiphany. We have an Excel spreadsheet that allows you to enter the parameters of a microphone and dial in your cable specs and how much capacitance it has per foot, and it will calculate how much cable length at what frequency you start rolling off and what the slew-rate distortion is, and once we got that done, our applause in the rear channels sounded much better. Low-capacitance cables on analog microphones are the key.”
In addition to high-end cable, Pappas has high-end gear for monitoring the mix: five Avalon Acoustics speakers driven by Jeff Rowland Class-D 1,000-watt amplifiers. The hall's control room was filled with TubeTraps to control echoes.
Controlling the Signal
Pappas paired with David Day of DaySequerra to supply the surround encoding gear. Out of the DS-00 console, the signal was routed into a DaySequerra NLC5.1ST for encoding in DTS Neural Surround. The NLC5.1ST helps control intermodulation distortion without artifacts so that it increases Time Spent Listening (TSL), a key metric for measuring the success of radio programming.
“The DTS Neural loudness algorithm isn't a multiband processor at all; it's a loudness processor,” explains Day. “It speaks to how the ear hears. Its processing is set to be sensitive to the critical bands in your ear. We are the only people on the planet who have anything close to this. Over 250 DTV stations and HD Radio™ stations in most major U.S. markets are using DTS Neural Loudness Control.”
After going through the NLC5.1ST, the feed went to a DaySequerra DTS DownMix surround encoder and then via ISDN using Telos Zephyr codecs to Colorado Public Radio.
All the digital units were clocked via an external Rosendahl word clock using co-ax. The Rosendahl clocked the Grace M802R preamps, the DaySequerra units, the DS-00 and the Neumann DMI-8.
“What we've found is that jitter is the defining factor in digital audio quality,” says Pappas. “Especially once you start getting into the A-to-D conversion process, if you have a lot of jitter going on at that point, there's nothing that fixes that later — you're stuck. Early on we figured that out and decided we wouldn't buy anything that had anything greater than 500 picoseconds of jitter in the audio passband. Most of the stuff we have now runs in the less-than-50-picosecond range, including the Rosendahl.”
In addition to broadcasting, Pappas and his team recorded the concert in PCM, via Logic 9 on a Macintosh, and in DSD. The DS-00 sent one stream to an EMM Labs up-sampler, then to a Genex 9048 for DSD recording.
Candace Horgan is a Denver-based writer.