[ This article was first published in the April, 2010, issue of
Larry’s Final Cut Pro Newsletter. Click here to subscribe. ]
Friday, April 9, we loaded up our rented Buzz Mobile Studio and trundled our way across the desert to Las Vegas and the 2010 NAB Show. In this report, I want to describe what we did, and how it turned out.
NOTE: For a list of all our shows and coverage, visit: www.nabshowbuzz.com.
This was our second year as the Official Podcast for the NAB Show. Our goals for the show were to create at least six hour-long programs and at least 35 five-minute News Briefs. By the time all the dust had settled, we had created 10 hours and eleven minutes of long-form programming and 37 News Briefs containing roughly 240 minutes of news. Whew — more than 14 hours of programming produced in a week!
It would not have been possible without our crew of 22 staff and volunteers.
We setup into a 20×20 foot booth, with the broadcast table at the back, editors on the right and reporters on the left. We had very limited Internet access provided by the venue, so we supplemented it by a series of cell-phone-based USB wireless devices for key computers.
Last month, I wrote an article detailing our technical gear (you can read it here). The only changes we made from that article were in experimenting with Vericorder technology using iPhones for news-gathering, and testing fiber optic digital cabling for our final signal path. More on these in a bit.
One of the lessons we learned last year was the importance of a good monitoring system for key members of the staff. Last year, we had speakers in the booth, but no staff headsets. This year, we brought the speakers, plus headset set amps and headsets so that we could have up to eight people all listening to the shows at the same time.
The content for our News Briefs was gathered by a team of reporters racing around the trade show floor and covering the convention. They captured sound bites using Marantz 661PMD digital recorders and Vericorder-equipped iPhones. We used the Marantz recorders last year, along with a Shure SM58 mic, and they continued to work great. Fast, reliable, and the WAV files they created were easy to edit directly in Soundtrack Pro.
All our audio editing was done in Soundtrack Pro, feeding a central server (a Mac Mini RAID) with storage on a Drobo Elite (see the review later in this issue). Once we figured out a naming and file storage convention that worked – which took us a day – technically, the system worked flawlessly. (Well, except for a bad Ethernet cable that kept taking our server off-line. Once we found that problem, which took a couple of hours, everything else was seamless.)
While the reporters were gathering news, I was busy in the booth doing extended interviews with key industry leaders. In the booth, I used three Beyer-Dynamic headset mics, feeding into three Aphex 230 Master Voice Channel Processors for pre-amp and audio processing. (Sigh… they helped us sound so nice!)
From the Aphex, the signal went into an Alesis MultiMix 8 FireWire eight-track mixer and were output via FireWire as three discreet audio channels recorded on my MacBook Pro laptop using Soundtrack Pro.
Prior to the show, I created a template for STP that preset audio inputs, applied standard filters and settings, and had the track layout I wanted to use. Then, at the start of each interview, I would open that document, save it under a different name, and record the session.
Each session was recorded and saved in a folder named after the company and the person speaking. I also collected all audio files from that recording (using Save As…) so that I could move all the files associated with that recording simply by dragging the folder from one place to another.
(For a complete tutorial on how we used Soundtrack Pro for our editing, check out this tutorial bundle. We created these four movies specifically to train our editors.)
Each show was created by editing each interview as a stand-alone AIF file, then I recorded the open and close from our booth on the show floor, and stitching it all together in Soundtrack Pro. The final mix was output as an AIF file, compressed into both MP3 and AAC files for posting to the web, and delivered to Greg Clarke, who handled all our web posting for final uploading to our web site and iTunes.
Our workflow for the news was different. Each reporter kept track of all their interviews on a written field log form, along with the business card of the person they talked with. Since a digital recorder creates a single file for every recording, the reporters were told to only record one file for each interview. (To keep things even simpler, we recorded everything using one mic into a mono file.)
When they were done gathering interviews, they returned to the booth, where our operations manager, Laura Peters, copied the files from the recorder to the server, renaming files based upon our naming convention and creating backups — just in case. Then logged all the interviews into a database so we could track them later. (Which was a good thing as we accidentally erased one file during production.)
From there, the editors took the source file and edited it into a sound bite. This edited AIF file was then reviewed by the show’s producer, who, along with two writers, wrote the scripts; including lead-ins to each interview.
The script was delivered to me as a print-out. We debated going entirely paperless for the show, but developing a network-based database for scripting and clip tracking wasn’t worth the effort given that we needed to go from setup to full-operation to tear-down in four and a half days.
I then narrated the script live to our streaming server, playing all the sound bites manually on my laptop using QuickTime player. The audio traveled from my computer to a separate channel on the Alesis, into an Aphex 320D Compellor for final audio processing before traveling to the streaming server.
The audio was recorded at the streaming server during the broadcast – just in case I made a mistake during the production.
Greg Clarke, our streaming engineer, cleaned up the recording, if necessary, compressed it into both MP3 and AAC for posting, and the news hit the web.
During the course of the four days we originated news from the trade show floor, we generated 225 interviews, edited 180 of them, and broadcast 37 News Briefs containing 140 interviews.
THINGS WE LEARNED
Smart decision #1: We decided to go with an all-hard-wired network. There’s so much RF and wireless activity at the NAB Show, we wanted to avoid any problems, so we built our own network. 16 computers into one NetGear switch, feeding an OS X 10.6 Server running on a Mac Mini, with files stored on a Drobo Elite. Worked great. (At one point I measured the server CPU load. Under full production we used less than 25% of the server.)
Smart decision #2: Noise canceling headsets for our editors. This reduced the ambient noise from the hall and made it easier to hear what they were editing.
Smart decision #3: Lots of headsets to monitor news feeds and recordings. PR people, team members, engineering, and producers all needed to listen at some point. We had almost ten monitoring headsets available at any time.
Smart decision #4: We planned our recording schedule so that we didn’t have to do too much during the first day. This gave us plenty of time (well, OK, barely enough time) to get the system setup, debugged, people oriented, trained, and figure out where our planning failed so we had time to solve problems before the madness hit.
Smart decision #5: We assembled a REALLY good crew who got into the spirit of the event, worked tremendously hard, and turned out solid results very, very quickly.
Almost everything we did worked fine. But there were some glitches.
Glitch 1: Signal path. The Aphex Compellor has both analog and AES/EBU digital outputs. In talking with the folks at Aphex (who are wonderful to work with), we decided to use the digital output. This required that we use a TOSlink converter to take the AES signal and convert it to fiber optic for input into the Mac (see the article later in this newsletter about TOSlink).
We bought a Hosa S/PDIF Optical to AES/EBU Digital Audio Interface. For the fifteen minutes that it worked, it was great. After that, it would arbitrarily drop signal – generally 5 to 50 seconds after we started recording. We could never get the converter to work properly. So we gave it up and went back to all analog. This meant that we were not able to fully utilize the power and control of the Compellor.
Next year, I want to tackle this problem and see if I can figure out a solution. An all-digital signal sounds better and the Compellor REALLY handles level changes nicely.
Glitch 2: We needed a better intercom system. When you are doing live events, it is critical that all team members be able to hear AND talk. We had really good monitoring, but the ability of team members to easily communicate with each other was very limited.
THOUGHTS ON USING VERICORDER
Vericorder provided two reporters for us to use during the event, so that they could showcase their iPhone software, as well as test their new video recording software. The two reporters (Brian Pellot and Erica Zucco) worked really hard tracking down interviews. We only had one problem with the system. On the last day, the external microphone for one unit was assigned to a different project and unavailable to us. So, the reporter used the built-in mic of the iPhone. The resulting quality was very poor, the iPhone wasn’t able to filter out the ambient noise of the event the way an external mic could. The interviews for the last day were unusable.
However, the buit-in audio editing capability of the Vericorder system, in the words of producer Cirina Catania: “we were grateful because the pieces [edited on the Vericorder] actually cut down on our workflow time.”
We also has some initial difficulties figuring out how to capture the audio files recorded on the iPhone to our computer network. Once the files were in the system, they were as easy to edit as any other files. But getting them in required a dedicated computer, some careful thinking about setup, and a fairly technical operator.
I want to say a special thank-you to Gary Symons, CEO of Vericorder, for providing both iPhones and software, plus the services of Brian and Erica for our NAB coverage.
All-in-all, this year was very successful. I want to thank Marantz and Aphex for the loan of their gear. NAB for providing us a booth and related services to use as home base. And our sponsors for all our broadcasts: Adobe Systems, Data Robotics, Focal Press, Pond5, and Carl Zeiss lenses. Without their financial support, this would not have been possible.
I also want to thank our team of staff and volunteers. It was hard work and they did a great job:
It was a GREAT team!