DTV Audio Group Event Focuses on NFL’s Artificial Crowd Noise

Online panel reveals that curated sound is fine-tuned

Crowd-Less Sports Audio III, NFL Edition,” the DATV Audio Group’s third look at the phenomenon of “curated” prerecorded fan noise for games without fans in the stands during the pandemic, took place online Sept. 30, and focused on the NFL’s experience.

This event, moderated by DTVAG Executive Director Roger Charlesworth, occurred during the league’s third week of regular-season play — a season that began without a preseason opportunity to adjust to the use of crowd sounds recorded at NFL venues in 2018 and 2019 as part of NFL Films’ updating its in-house audio library. As authentic as the sounds are, they nonetheless presented challenges when it came to choosing, sorting, categorizing and editing, with the intention of providing actual sounds taken from each team’s stadium.

During the first period of the DTVAG event, Vince Caputo, VP/senior sound supervisor, NFL Films, explained the process of preparing the sound files; developing, with consultant Robert Brock, director, education, Conservatory for Recording Arts and Sciences, and Fred Aldous, senior audio engineer/consultant, Fox Sports, the Wwise authoring system-based platform for accessing and processing the files; and recruiting and training 32 crowd-sound mixers and their backups.

Conservatory for Recording Arts and Sciences’ Robert Brock (top row, center) explained the NFL’s approach to crowd sound to a virtual DTVAG panel.

Caputo described the entire undertaking as “ambitious, to say the least. We could have settled on using a generic sound-effects library, but, if you’re playing at CenturyLink Field or Lincoln Financial Field, why not let it sound the way those stadiums do?”

Brock added that the Wwise platform, designed for videogame authoring, was a good choice for the NFL. The league, he explained, needed to be able to not only access different types of audience sounds, from cheers to chants and play reactions, but also to be able to make smooth transitions between them and between the various levels of intensity programmed into the system. The goal was to make it easier for the crowd-sound mixers, who were put to work on national-network games with just a week or two of training.

“The Wwise is set up under the hood for these kinds of transitions,” he said. “We worked to make it less of an engineering tool and more of a crowd tool.”

Caputo noted that not until the league’s third week of play did any of the 32 crowd-sound mixers have more than one game under their belt: “Up until then, we were all first timers.”

The A1 Experience

What also emerged from the DTAVG event were the somewhat disparate experiences of A1s on NFL games. Jonathan Freed, who mixes Monday Night Football for ESPN, said he takes the four-channel mix sent to him from the crowd-sound mixer and essentially leaves those faders on his console alone throughout the game, relying on the mixer to follow game dynamics.

“I preview the sounds for the week and set my baseline levels,” he explained. “My goal is to leave them alone, not micromanage it, to leave it to the crowd mixer, unless there’s a [network-level] request to turn it up or down.”

On the other hand, Wendel Stevens, who mixes the NFL for NBC Sports, thought he would approach crowd sounds the same way but quickly realized that he would have to ride their levels along with other audio elements of the game.

“Every play,” he said, “you ask yourself mentally, What would the crowd do here? I realized in the first quarter of the first game that I would have to manipulate the levels up and down. If a pass got dropped, I would pull the level back 10 dB to make it sound natural. You know after doing this a long time what feels right, but, when it [doesn’t], when it feels unnatural, you turn you head and squint a little and fix it. It wears me out.”

Lee Pfannerstill, senior audio engineer, CBS Sports, also finds himself questioning the authenticity of the “crowd” reactions.

“I do move the fader more than I want to at times,” he noted, “because sometimes the reaction lengths seem too long, or it’s an overreaction, like a big cheer for an extra point. So I’ll pull the fader down quicker and then back up so there’s ambience under the next play.”

A key skill, one that the A1s are schooled in, is looking at the big picture to anticipate events and crowd reactions to them. For instance, Pfannerstill said, fans at the game will notice an open receiver downfield, and their noise will swell accordingly, intuitively prompting him to push his ambience channels up in anticipation. But, if the crowd-sound mixer does the same thing at the same time, the effect is overpowering and inauthentic. “We’re getting better at that as we go along,” he noted.

He added that the network announcers, who are working remotely, often clamor to hear more of the crowd sounds in their earpieces.

“They’ll usually have the crowd pouring into their headsets through the crowd mics and the parabs and natural bleed that comes into the booth. But they don’t that have anymore,” he explained. “They really feed off that. I’ve gone from –10 to zero for them on a game.”

Continuous Training

Noting that NFL Films makes constant adjustments to the velocity and other parameters of the sound files to address issues like that, Caputo pointed out that such miscues are diminishing as a result of accumulated experience and his regular after-game virtual meetings with the crowd-sound mixers.

“Our main teaching point has been the need to jump on reactions early and build reaction shots, so they don’t sound like they came out of nowhere,” he says.

Other items on the DTAVG agenda included better integration with venue event-production teams, whose music, effects, and other elements pumped into the PA system find their way into the broadcast mix.

Another issue has been where to position the crowd-sound mixers. In some cases, location is determined by the availability of and proximity to broadcast comms in the venue. In at least one case, the mixers have been positioned outside in the truck compound, but the approach of winter will prompt a search for other locations.

What most A1s have been happy about is how available is audio from the field of play and, particularly, the scrimmage line without the masking effect of actual crowd noise. Freed attributed that to the infrastructure that sports broadcasting has built up in stadiums over the years, with player mics and parabolic collectors covering almost every inch of the playing field.

“I can hear more detail, more banter at the scrimmage,” he said. “I’m not struggling to hear the quarterback. And the announcers are working off it.”

Stevens cited sound effects that he had never heard before, such as the sound of cleats making a sharp turn on artificial turf. However, he added, it’s critical to achieve a newly balanced mix that blends curated sounds with actual ones. Otherwise, it can sound “naked and minimal … not like an event.” The mix, he emphasized, needs to achieve the league’s professional “stature.”

In fact, according to submixer Justin Blackwood, the curated crowd sounds can actually help make the transitions between nat-sound sources. Once those audio levels began to become more consistent, he noted, he was able to push his effects sounds harder and feel them more seamlessly integrate with such elements as the PA and music. “It really helps tie them all together.”

A larger truth emerged from the session, however: an appreciation for the value of fans in the stands.

“Without fans,” said Stevens, “it doesn’t sound like the same thing. I’ve come to have a real appreciation of the crowd.”

Password must contain the following:

A lowercase letter

A capital (uppercase) letter

A number

Minimum 8 characters