Cinema Audio Society (CAS) Award-winning production sound mixer Steven Morrow has recently earned his second Oscar nomination for Best Sound Mixing, and both nominations happen to be for musicals. First, there was director Damien Chazelle’s 2017 musical comedy La La Land. This year, it is director Bradley Cooper’s remake of the rom-drama A Star is Born. Two musicals that required two very different approaches to production sound.
On A Star is Born, Morrow was tasked with capturing live vocals for all of Cooper’s and Lady Gaga’s performances, some of which were filmed in front of real music festival crowds. He also mic’ed up those crowds — from three different angles. And mic’ed up the band on stage. Those concert scenes maxed out at a whopping 61 tracks! But in the end, it was all worth it. “This is the best-sounding film that I have done so far…The final soundtrack for the film just blows my mind because it sounds so good. It feels like you’re there with them,” says Morrow.
Here, Morrow talks about how they arrived at their plan of attack for capturing the concert performances, and how they made the concerts look and sound real — just as Cooper and Lady Gaga intended. He shares his struggles and his triumphs and gives us a look at what’s on his sound cart.
S&P: All the singing/vocals for A Star is Born were captured live on set. How did you prepare for that?
Steven Morrow (SM): We were hired five months before production started. Bradley Cooper had let me know that Lady Gaga wanted to perform everything live, because all of her shows are live and all of her performances are always live. She wanted to have that authentic sound you’d get from performing the vocals live.
We’re all fans of musicals, but there’s that moment when it goes into a song and the ambient world disappears. So the goal on this movie was to avoid that. My first conversation with Bradley was that he and Lady Gaga were going to sing live, and how do we do that.
S&P: You had a live vocals test run at Warner Bros. prior to the first day of filming. How did that go?
SM: Early on, there was the question of whether to record the full band performing live or record just the vocals live to playback of the band. We had to determine which was going to sound better. On the technical side, it was way easier to do a playback of the band and do the vocals live because that way you can guarantee the band will always be in time throughout the day, for ten hours straight. If we used playback, the band was always going to playback at the same speed, whereas if you have a band playing the same song live, over and over for ten hours, by the end of the day (no matter how great the band is) the timing of the song is going to change. That would make it hard to edit from take to take. So the purpose of the test run was to show that either way the vocals would sound the same. We can process the band’s studio performances and make those sound live. The most important thing was proving we didn’t need to record the band live, just the vocals. We always wanted to record the vocals live.
So we set up with the band on one of the Warner Bros. stages and we recorded the song twice — once with the band playing live and the singers performing live, and then again with playback of the band and the singers performing live. Jason Ruder [supervising music editor]mixed the two versions that night and presented them to Bradley and Lady Gaga in the morning. There was no difference in the two recordings, in terms of which one sounded more authentic and live. So that was the proof of concept to demonstrate that playback of the band on set wasn’t going to ruin the live performance feel. It wouldn’t sound like a pre-recorded song playing back in the film. So this test run gave us a process to follow on set to technically achieve the sound that everyone wanted.
S&P: What were some of your concerns going into that first shoot at Coachella?
SM: The main goal from the studio and from Lady Gaga and Bradley was to not have the music leaked. We can guarantee that the crew wasn’t going to record it and put it on the Internet. And we could make sure that the extras that were there getting paid didn’t have their phones. But at Coachella, there are vendors and all sorts of people there. We shot in between the concert weekends, but there are still vendors and people setting up and the roadies offstage who aren’t part of the filming. We can’t guarantee that they aren’t going to record us making the movie and leak that on the Internet. We didn’t want any of these songs to get leaked a year and a half before the movie came out. A lot of the songs were pivotal points in the movie and you don’t want that information out there to ruin the experience when the film is released.
The biggest concern was how to achieve what we needed to technically achieve while still keeping the music secret. Bradley sang live at a real concert a couple times — once at the weekend after Coachella, which is Stagecoach. He was out there right before Willie Nelson’s band came on so he was singing in front of roughly 50,000 people. In between sets, he jumped on stage and that scene is the beginning of the movie. We had 10 minutes to shoot it. And again, we didn’t want the music to get out. So how do we shoot it and get what we need to get while still protecting the material?
We had a couple of ways that we approached it. When Bradley jumps up on stage at Stagecoach and at Glastonbury (in front of 120,000 people), we were filming between sets so that it looks like the crowd is there to see Jackson Maine. It looks like he has this giant fan base. In those scenarios, because we didn’t have a lot of time to set up and not much leeway in what we could do, I would send a feed to the monitor mixer (stage left) and he would play that playback track into the wedge speaker in front of Bradley on stage. The music was played back kind of low — just loud enough for Bradley to hear it. Then, we take the feed from the stage mic that Bradley is singing into and run that into the recorder, but we wouldn’t amplify the music or his vocal to the crowd. When you are on a bigger stage like that, sometimes the first row of the crowd is 40 feet away. From that distance, you really can’t make out what he’s saying, especially for the more intimate songs.
The real crowd that was there all day, listening to music and enjoying the bands, suddenly got very confused because there was Bradley Cooper on stage and he’s singing and playing the guitar but nobody could hear him. So the first run through of the song everybody was excited because they were seeing it. Then on the second run through, they were a bit confused because the crowd couldn’t hear the music. It was fun to look through the newspaper the next day and see that technical problems were reported, that the mics were shut off. That was the goal — to make it look real but not have the crowd hear the music. So it was nice that the music wasn’t leaked despite there being over 50,000 people there.
At Coachella, we had time to set up and we had film extras there (not real crowds). So for that, we had the in-ear monitors so the performers could all hear the music. We deadened the drums as much as we could and we muted all the amps. So the band would play along to the playback track, and Bradley and Lady Gaga would sing (but we didn’t amplify their performance to the audience). Again, the audience was far enough away that they didn’t hear the song. If they heard anything, it was out of context and there was no music to it. Also, the muted drum kit isn’t truly quiet, it’s just deadened. So that overwhelmed the vocals for those standing in the crowd.
S&P: What was on your sound cart for these live performances?
SM: Basically, the cart didn’t change whether we were shooting the live performances or dialogue. The idea was to always set up the band performances as playback but there was still a possibility of doing a whole song live. There were times when the band would play a song that they hadn’t recorded in the studio yet, and so we would film it and that would be fully live. These were songs that Jackson Maine was playing (not with Lady Gaga) and those were performed in more controlled environments (not at Coachella or Glastonbury). We were always prepared to record the band whether or not they were going to play live. We wanted to make sure that we covered every instrument and every amp, every possibility, so we could flip a switch and pull up the faders and do one live whenever.
As you start counting instruments, you start counting microphones and you make a dream list of what you want to record if you can record everything. We got up to 61 tracks of audio. So it was this monster amount of information coming into the recorder. That also included a 5.1 surround sound mic that we set up in the audience to get clean crowd sounds. We also set up mics on the left and right of the stage aimed at the audience to get all of their whoops and cheering. We put mics in the back of the crowd aimed at the stage, to get a far perspective of the actors on stage. The drum kit was mic’ed and the guitars had direct lines in and also the amps were mic’ed. Suddenly, you’re at 61 tracks.
I had two Midas M32R Mixers, each one with 32 inputs, and through a Dante network we fed those into two Sound Devices 970s, which can do up to 64 tracks but I had an extra one running for backup. I didn’t want one to fail and therefore ruin the movie.
I had my main mixer on the cart where it usually is, and I had the additional mixer on the top shelf of my cart so that we’d always be ready to move. If you’re in the shot and have to get out of the way, that’s a ton of cabling and equipment that you have to move all of the sudden. We always try to set up in a spot where we won’t have to move but we were fully prepared to disconnect and run-and-gun that way.
I wasn’t physically mixing the whole time on that top mixer. I was just recording the inputs at good levels and then Jason [Ruder] and the post sound team could sift through it and mix it down how they wanted.
We always recorded all of the instruments, even if they were muted. A muted drum kit can at least tell you when the drummer hit the foot pedal, for example, and all that information provides a sync point for editing. You can use the production recordings to sync-up the studio recordings. When it all lines up then you know you’re solid. Or, if the guitar player was doing a cool riff or if the bass was doing something cool — which the players often did — that makes it unique to the movie and they have that material to mix in. It was the same musicians performing the music, both on set and in the studio. It was the same guys so they knew the songs well.
S&P: What about the dialogue? Were Bradley and Lady Gaga always wearing lav mics?
SM: In the film, you don’t see it so much, but Jackson Maine really interacts with his band a lot. I think most of that was cut out for timing. But we had the band mic’ed up every day, including Bradley and Lady Gaga and whoever else was with them in the scene. Everyone was wearing a lav mic. In the concert scenes, the cameras were everywhere. There were three or four cameras that were covering every angle and we wanted to make sure that we captured any conversations that they had. For example, in the “Shallow” song sequence, Jackson Maine goes over to Ally and he’s trying to talk to her over the roar of the crowd. That’s the actual crowds cheering. We didn’t have the crowds do the normal thing where they just mime. Since we wanted it to feel as real as possible, we just let the crowds go crazy. That’s why Jackson had to lean in and loudly whisper into Ally’s ear. Those are the moments that make it real. Sometimes actors don’t consider what noise will be playing in the theater while they are delivering their lines, especially if it is silent on set. So we let the crowd dictate how loud they wanted to be and how excited they wanted to be that way the actors could react to it. We always mic’ed the actors, to get those moments and any unscripted dialogue they might have before or after a song.
S&P: What were the lav mics that you used?
SM: We used the Sanken COS-11s on Lectrosonics SSM transmitters.
S&P: Going back to capturing the crowds, what mics did you use and where did you place them?
SM: We set up two Sennheiser MKH 416s on the left and right side of the stage aimed at the crowd so you’d have a great split on a stereo field. Then, for the surround mic that we placed in the center of the crowd at eye level, we used a DPA 5100 surround microphone. And the mics behind the crowd pointed at the stage we put two wireless Sanken COS-11s. Those weren’t capturing much other than the distance of the room.
We also used those mics to capture impulse responses (IRs) from the venues where we recorded. You play this awful sound that goes from 0 Hz all the way up to 20 kHz in a linear band through the speakers as loud as you can while the crowd stays quiet. It takes about a minute and a half to go through the whole frequency band and it throws out these computer noise pops. Once you bring that information back, you tell it the mics you used to record it from the stage’s point of view. The drum kit mics even worked well for this process. It tells you how big the room is, how much echo there is, and how live it is. And it does that by listening to the sound that was sent out and to its resulting slap back. At Coachella, we did it on the main stage and the signal went out 10 or 20 miles, bounced off the mountains, and came back to us over a second later. Those IRs were used in post production on the studio recordings, to give the intimate studio recordings a custom reverb sound that matched the space where the song was performed in the film.
S&P: What was the most challenging live performance to capture?
SM: One of our hardest days on set was actually the second day filming. It was challenging because this was our second day at Coachella and we were moved from the main stage to the Mojave stage. Part of the montage of Jackson Maine and Ally performing together was this huge Coachella crowd and then, for the song “Always Remember Us This Way,” which Lady Gaga performs on the piano while she sings, that was shot on the Mojave stage on the same day. Part of our challenge for sound was that this was a three-hour set-up and then two-hour tear down. So we moved everything a half-mile from one stage to the other, set it up for three hours, filmed all morning, then when the crew went on break for lunch the sound department spent two hours tearing everything down and moving it to the other stage and spent another three hours setting it all back up. That’s five hours of time when production can’t do too much filming because there is no sound. They can only do so much without sound. For that day, the general call time was 7 AM but the sound department was there at 4 AM setting up. Then, we wrapped at midnight and didn’t get out of there until about 2 AM. So it was a 22-hour day and we worked through the whole thing. And that was the second day of filming. So for us, it was physically difficult but not difficult in terms of capturing the performances.
As for the performances, they were always kind of nerve-wracking because you never wanted to have anything mess up. You didn’t want to have anything go wrong while recording Lady Gaga or Bradley performing live. You never want to have to say, “Hey, I need to do another one for sound because that one wasn’t so great.” You just want to have it perfect every time so that their options in post were always dictated by the performances and not by the quality of the production sound.
“Shallow” was probably the most difficult song to record because it starts quiet and then Lady Gaga is belting it out. You have your hand on the gain and you are ready to crank it down as soon as she starts belting it out because you don’t want to blow out the recording and ruin the take. “Shallow” was difficult because of the drastic volume change from the verse to the chorus.
S&P: What was the most challenging dialogue-driven scene?
SM: Their house — where Jackson Maine and Ally live — is a practical location. It’s a real house. The bathroom scene where he comes in drunk and she’s egging him on before their big fight was challenging because it’s a real bathroom and there were a bunch of crew people in there. She’s in the bathtub and so you can’t mic her; you have to boom her. The bathroom is like this bad echo chamber-type room that you have to make sound good. It all comes down to the performance. The goal is to never throw the audience out because it sounds weird. That does happen and there are times where you just can’t control it, but you do your best to make it as good as it can be for that scenario. So, that was a challenging scene and bathroom scenes in general are difficult to record because of the hard surfaces and the parallel walls.
Some of their stage scenes were challenging for dialogue, like when they’re practicing for the Grammys and Jackson finds out for the first time that he’s not the singer; he’s just the guitarist. When you are on a stage like that everything (from the lights to the digital displays) has fans that are blowing and making noise. That’s not an issue when they are playing music but when they are having intimate dialogue on stage, then those things come up. If it’s a shot that’s super close-up, you can easily boom it. But you have Bradley and Sam Elliott, who generally talks down towards the floor, and so it sounded better on a radio mic. It’s a head-and-shoulders shot so we could’ve put the boom anywhere we wanted but I knew the radio mics would be better because it would get rid of most of the fan noise and the dialogue would be clearer. But we shot it with both the radio mics and the boom so that they had the option to choose one or the other in post production.
For the boom, we use the Sennheiser MKH 50. We pretty much use that for everything. If it doesn’t sound good on that mic — and this sounds crazy because I know a lot of guys who only use shotgun mics outside — then the sound probably won’t be much better on a Sennheiser 416. At that point, you’re probably better off using a radio mic. My gauge is that the MKH 50 should work for everything and if it doesn’t then you should use a radio mic. Even if you’re getting it 5% better on a shotgun mic, it will be 40% better on a radio mic.
S&P: What are you most proud of in terms of production sound on A Star is Born?
SM: I think in general, it’s the movie as a whole. When the audience comes out of the theater and says, “Man, that’s incredible!” that feels really good to hear — even if the audience doesn’t know the reason why. It wasn’t easy to get this real concert sound and Bradley was really focused on that in post production. He was really involved with how it sounded and he really wanted it to sound authentic.
I would say this is the best-sounding film that I have done so far, and it’s the film that I’m most proud of. There could have been so many pitfalls while recording the sound; the sound could have been a disaster but we came up with a plan. I will say that the test session we did at Warner Bros. was the Friday before we started shooting. Our first shoot was on Monday so we had a weekend to figure it out, not, like, a month to figure it out. So we had that proof of concept that it could work and in the end it all came together. The final soundtrack for the film just blows my mind because it sounds so good. It feels like you’re there with them. The post magic in the mix really amps up that feeling. The success of the sound in the film helps the audience to fall in love with these characters even more.
Images courtesy of Warner Bros.