In late September I was asked if I would like to broadcast from New Orleans for the Toast of the Nation New Year's Eve event. The idea that New Orleans was going to be in any shape to support a coast-to-coast NPR broadcast was hard to comprehend, but if NPR wanted to do it, I was game.
I flew to New Orleans for the technical walkthrough on Oct. 25. The show was going to be live from the legendary Tipitina's Club. Tipitina's was fully resplendent in a wrap of plywood over the doors. A makeshift spray paint sign announced that NPR was going to broadcast there on New Year's Eve. Despite the outside appearance, the club's inside was in good condition.
There was a huge amount of logistical work be done to make this happen. First and foremost was figuring out how we were going to transmit the show to NPR master control. Jane Holmes, manager, Remote and Mobile Services at NPR, contacted Bell South and ordered six ISDN lines and one POTS line with symmetrical DSL. As a backup, we planned to use a satellite uplink, but that would increase the costs significantly.
We were greeted by this make-shift sign. Fortunately, the club's interior was completely intact.
This trailer housed the broadcast village. The vehicle was slightly wider than originally ordered and stuck into the street.
The announcer position in one end of the trailer. It's not the fanciest place but it worked perfectly.
Duke Markos makes a final check of the audio.
Mike Pappas points Fritz (the dummy head on the pole) to capture the marching line.
The transmission rack for the broadcast.
The walkthrough complete, I flew back to Denver with a list of needed items. The first item was a modular office trailer. In the 26 years that NPR's New Year's Eve Toast of the Nation had been broadcast, the hosting had always been done from NPR in Washington. This year, NPR wanted us to host the whole network from New Orleans. The trailer would house an announce booth and a place for producers, transmission equipment, communications equipment and a 5.1 surround mix location. Unfortunately, office trailers were in short supply. I was laughed at more than once when I called rental places to rent a trailer. I was told that nothing was available and that no one would commit to a delivery any more than three weeks prior to our event date. I had nightmares that three weeks out we would not be able to rent an office trailer, so my plan B was to rent two motor homes in Denver, load all of our equipment in them and drive them to New Orleans.
Because demand was so high, rental fees for everything were almost 300 percent higher than normal, which stretched an already tight budget to the limit. Additionally, Tipitina's power was marginal and wasn't capable of supporting the broadcast village we were going to build. Finding a generator in an area that was aleady short of generators resulted in more calls to generator rental houses and more laughter. Again, the standard answer was to call three weeks prior to the event to see if anything is available. Plan B was to rent one in Denver and tow it to New Orleans.
In the meantime, Holmes acquired the six ISDN lines and a POTS line with DSL and installed them to the phone pole outside of the club. I had e-mailed her a picture of the phone pole in question to make sure we didn't have any issues as to which pole I wanted the lines installed on. Everyone laughed at the pole picture, but the lines were installed where I wanted them.
While waiting for genset and modular trailer dates, I needed to obtain a mixing console, codecs, codec preconditioning equipment, surround encoders and a bunch of communications equipment. I had seen the Digico DS01 consoles in action on the James Taylor tour and at the Surround III seminar at the October AES convention. We were going to broadcast three acts from Tipitina's with a short period of time between them, and having a console with total recall and full 5.1 facilities was a must.
I'm not a fan of most digital consoles for several reasons, but the Digico DS00 was the first console that I felt comfortable enough with to risk a show of this level.
Digico was willing to loan one to us, but suggested the DS00 instead because it is better suited for broadcast than the DS01, which is designed for sound reinforcement. Digico shipped the console to KUVO the week before Christmas so we could configure the I/O for the 40 AES-3 inputs and outputs and 16 analog inputs and outputs and test it. We had the I/O installed, the console configured and all the inputs and outputs (digital and analog) tested in a couple of hours.
Harris loaned us a pair of APT Tokyo codecs to feed Washington. I planned on sending three feeds to NPR: announcers on a mono L2 128kb/s feed, music in stereo using the Apt-x enhanced codec at 384kb/s and a backup that I could route to the announcers or the music feed using Apt-x enhanced at 256kb/s. KUVO has an APT Milano, which I planned to use for the music feed and I was going to use one APT Tokyo for the announcers and another for the backup.
Lots of sources
Because we used so many AES-3 sources, I needed lots of AES-3 distribution amplifiers (DA). ATI loaned us a pair of DDA112-XLR AES-3 DAs for transmission. I used my Z-Systems 16.16 router to handle all of the transmission routing.
We also needed wireless mics and IFB equipment, along with additional microphones outside of my collection, so my next call was to Sennheiser and Neumann USA. Sennheiser provided a pile of 3000-series wireless equipment along with SK-5200 hand-held transmitters with Neumann KK105 capsules and in-ear monitoring systems, which we used for IFB.
I needed an additional Neumann KU-100 stereo dummy head (aka Fritz) mic for field sound gathering as my KU-100 was going to be hung in the rafters of Tipitinas. Neumann supplied an additional KU-100 and a pair of KM-184 mics, and Neumann Berlin supplied pairs of TLM-193 and TLM-170 mics. Everything arrived at KUVO on Dec. 19.
Communication was significant for this event. We needed eight IFB channels, three channels of RTS PL with four channels of Telex SSA-424 digital two-wire-to-four-wire hybrids, two stage announce (SA) and a telephone hybrid for communication between the Washington director and New Orleans director. I needed three IFB master panels and a backup. My assistant at KUVO, Will Barnette, went through all of the IFB equipment testing and repairing as needed, along with checking all 12 of the RTS-325 TW PL belt packs and headsets.
The plan for the announcer palace was to make the system totally independent from the rest of the facility. There would be a separate mixer, analog-to-digital (A/D) converter, ISDN codec, Neural Ultralink and AES-3 DA, and this would be stand-alone so a failure in the rest of the facility would not take the announcers off the air. In case of a failure with the announcer codec or ISDN line, I could route the output of the announcer A/D to any of the surviving codecs, and I also had the analog outputs of the Midas Venice 320 console cross-patched into two of the analog inputs of the Digico giving me a couple of ways to make sure that at least we would have announcers if the world came to a halt.
At the other end of the trailer was going to be the 5.1 mix room. We packed all of our equipment including 40 tracks of Direct Stream Digital HD recording on my Genex 9048 with EMM Labs converters and 40 channels of Grace Design 801R and 901R remote-controlled microphone preamps. We figured we would need upwards of six 70GB Ultra 320 Seagate Cheetah III AV-rated hard drives to record the event and Sony AIT III tapes to back it up. We arranged to rent additional EMM Labs converters because I only had 24 track's worth.
The idea was to put the Grace preamp by the stage and run the outputs into the EMM Labs A/D converters. We would feed the DSD outputs of the EMM Labs Converters into the Genex 9048 and take the AES-3 outputs into the Digico. A Rosendahl Digital Audio Clock Server would provide master clock for all converters and the Digico.
Neural Audio provided Ultralink codec preconditioners for all feeds to NPR DC. The most critical of these feeds was the announcer ISDN line. The spoken word is tough on codecs because there is no place to hide the codec artifacts. Additionally, I wanted to use the Neural 5225 system to encode the 5.1 signal.
Finding a home
Three weeks from Dec. 31 we found a modular office trailer and a 40kW silenced generator. I got my first night full night's sleep on Dec. 20 after the contracts were signed and all vendors committed to our delivery schedule.
The truck from KUVO Denver was loaded and left the morning of Dec. 26 and arrived in New Orleans on Dec. 27. I flew to Houston on Dec. 27 and drove to New Orleans, as direct flights were impossible to attain.
Once in Houston, I called the trailer vendor to make sure everything was set for our trailer to arrive at 10 a.m. at Tipitinas. Things started to go wrong at this point because the trailer vendor realized that our trailer was sitting in Dallas and hadn't been scheduled to be trucked to New Orleans, and there wasn't a driver available to drive it. Of course, the vendor didn't have the trailer size we needed in New Orleans. I began having visions of us mixing the show on the sidewalk in front of Tipitinas and putting the announcers in the back of our rental cube van. Many calls later the vendor found a larger trailer for us to use. The problem now was that the trailer would hang out into the street by a couple of feet. But, at least it's better than mixing on the sidewalk.
The original plan was to park the trailer at 10 a.m. on Dec. 28 and the generator at 11 a.m. The generator showed up at 10 a.m. and the trailer arrived at 11 a.m. This resulted in the generator being sited at the wrong location, which we didn't discover until the trailer arrived. We enlisted as many bodies as we could find to push the 4,000-pound generator to the correct location. It wasn't pretty but we got it in the right spot and nobody got hurt doing it.
The trailer arrived and was parked and leveled. Only then could I walk into it and determine, much to my horror, that the vendor had shipped us a trailer with a different floor plan. The floor plan we needed had an office on each end with an open center section. The one they shipped only had an office at one end leaving us with no place to put the announcers. After several terse calls back and forth to the vendor, they came out the next morning and constructed a wall with a door for the announce position.
We had a tight schedule to complete the announce position because NPR and WBGO wanted the announcers to feed spots to Washington on Dec. 30. We went ahead and built the announcer palace and hoped that the wall construction would be where we needed it.
Because of the floor plan issues, we had to put the announcers at the end of the trailer nearest to the generator, resulting in a fair amount of acoustical noise. The NPR folks weren't all that happy with the amount of noise the mics picked up.
While I was dealing with the floor-plan issues, Will Barnette and Dave Kunian constructed the 5.1 mixing room, and John Mikity organized a maintenance area in the back of the cube van and coordinated the cable runs from the trailer to the stage. I powered the ISDN equipment and tested all six ISDN lines. One of the SPIDS was programmed wrong at the central office, so I called Bell South to fix it.
On Dec. 30 we continued to build the various systems. The mic preamps were placed in the club, wired and checked all the way through to the Genex and Digico. I had most of the communication equipment up and running, including the two-wire-to-four-wire hybrids that were interfaced to the Digico so Markos could talk directly to the stage by hitting a talk back button on the console. I also provided a 50W UHF base station with external antenna along with three hand-held radios for roaming.
A balance of power
KUVO has used a Furman IT-1220 balanced power system for the last three years for all events. I have found that balancing power results in a 6dB to 8dB improvement in the noise floor of the system. I had forgotten that the Furman units provide over-voltage protection until about 2 p.m. when the technical power system shut itself off. Initially, I couldn't figure out what had gone wrong because within seconds the technical power was back on line. I noticed that the technical power had turned off but the rest of the trailer was running. I also noticed that the lights had gotten quite bright and when they dimmed the technical power returned. I jammed my Fluke meter probes into an ac outlet in the trailer and measured120Vac. I set the meter to record minimum/maximum and waited to see what was happening. About 10 minutes later, the technical power went offline again. The multimeter read a max voltage of 175Vac. We had a generator with a major voltage regulator problem.
The good news was that the Furman saved all of the technical equipment because the over-voltage system disconnects the ac if the input exceeds 140Vac. Without this, we might have cooked a bunch of equipment. The generator vendor dispatched a tech to correct the problem, and the generator behaved itself from that point forward. But I never took my eyes off the Fluke, which ran in min/max mode for the next 72 hours.
While I am getting gray hair by the volt with the generator problems, the rest of the crew completes building all the facilities and performs a full check. Friday morning, Dec. 30, rolls around and the ISDN SPID problem is resolved. I can check all the lines with NPR master control. NPR has run out of facilities and we are limited to running the backup at L2 joint 128kb/s instead of Apt-x enhanced at 258kb/s. Additionally, NPR wants to check the lines using Prima codecs on the announcer and backup. Because the Prima codecs don't have a data handshake, the APT Tokyos have to be manually told that there is a Prima on the other end and then re-booted. We get everything configured and dialed up and everything runs fine without any drop-outs or data problems.
Day of the show
On New Year's Eve and we arrive on-site at 9 a.m. The bands are scheduled to load in at 10 a.m. We are going to record and feed a parade through the streets of New Orleans to Washington. Our crew and the band's crew start building the stage. The plan is to sound check from the last act to the first act. We didn't get a full sound check from the last act, and the second act didn't have a full complement of musicians. Additionally, the first act was missing the main musician for the sound check.
I attach the KU-100 to a 7' pole and feed a Marantz flash recorder and begin recording the brass line in the parade. After the performance, I gave the flash card to Josh Jackson for editing and to FTP to NPR. The DSL, which up to this point had been screamingly fast, is now so slow we can't FTP the file in time. NPR Master dials our ISDN lines and we ship it in real-time until I get a call from NPR asking why the feed is in mono. A check of the APT Tokyo shows that it is connected with G.722. A call to NPR Master and I find out that today they are using Telos Zephyrs instead of the Primas we used on Friday. The Zephyrs have the data handshake but because the Primas don't, we had to manually configure the Tokyos. I reconfigured the Tokyos for auto handshake, reboot, then have Master reconnect. Luckily, we manage to get it there just in time. At 5 p.m. we go live with the announcers to NPR.
The doors opened at 8 p.m. and the first act was onstage at 9 p.m. The sound check issues haunted us. The second act decided to sing into the horn mics for the whole set. And the third act sax player wasn't happy with the monitor mix (because he didn't show up to check it during sound check) and it fed back a couple of times.
This is one of those once-in-a-lifetime events and I know that the satisfaction of putting New Orleans back on the map after the devastating hurricane made all the effort worthwhile.
Pappas is chief engineer of KUVO-FM, Denver.