When the music video for Ariana Grande’s single “Thank U, Next” debuted late last year, it broke records. Within 24 hours, the clip had been viewed 55.4 million times, YouTube said, a new high-water mark for the service. The song lyrics earned buzz by mentioning four of Grande’s past boyfriends, including erstwhile fiancé Pete Davidson; quick-eyed viewers may notice references to those relationships, but the concept for the video revolves around tributes to female-driven films from the 1990s — Mean Girls, Legally Blonde, Bring It On and 13 Going on 30.

Cinematographer Christopher Probst, ASC, insists there’s no higher meaning or thematic thread tying the video’s different settings together. As directed by Hannah Lux Davis, “Thank U, Next” is mainly a nostalgia trip taking viewers back in time with references to some well-loved movies. And he’s reluctant to take any credit for the video’s bar-raising performance. “‘Thank U, Next’ was on Spotify cranking up the numbers prior to the video, anyway,” he says. “Her fan base was waiting for this, regardless of what we did. They would have been thrilled if we had shot on Hi-8.” Lucky for the fans, Probst, who is well known for his commercial and automobile work as well as music videos for the likes of Eminem and Taylor Swift, brought his usual high standards to the production, using an abundance of soft light, Steadicam, and the Red Weapon Monstro to follow the rhythms of the music and make sure Grande looked her best from every angle.

Because of the number of locations required, wardrobe and make-up changes scripted and celebrity cameos involved, the shoot was scheduled over three days — a relative luxury for a modern music video. On the first day, the production shot on stage, with set builds including a beauty parlor, a bathroom and a bedroom. The second day was shot at a private residence in the Valley standing in for locations from 13 Going on 30, including interior shots as well as a backyard pool party. Finally, the production’s third day was spent on the campus of Alexander Hamilton High School, which passed for an Ivy League university campus in shots harking back to Legally Blonde. The school’s auditorium (stocked with 70 extras who were shifted around in the seats to make the theater appear packed from different camera angles) was used for the video’s Mean Girls tribute, and its athletic field hosted a climactic cheerleading sequence echoing Bring It On.

Watch the video, below, then read on for details on the shoot.

Camera Choice

Probst shot with a Red Weapon camera with the 40.96 mm x 21.60 mm Monstro full-frame sensor. He likes shooting with a fairly large sensor because it gives him plenty of room to manipulate depth of field rather than restricting his options. “As you go into large-format sensors, you get the option to have wide shots that can still have shallow depth of field,” he explains. “Conversely, if you were doing videos back in the day on 16mm [film], you’d have a hard time getting anything to have shallow depth of field, because of format size. You wind up using wider and wider lenses on a smaller format for the same equivalent field of view, and as the lenses get to shorter focal lengths the depth of field increases. So for a certain angle of view you’re locking your depth of field in partially based on the choice of format. These larger sensors open up possibilities in terms of which lenses you use and what depth of field you can harness.”

On the other hand, Probst noted, shooting with an even larger sensor like the 54.12mm x 25.59mm A3X chip in the Alexa 65 could reverse the problem so that a DP would fight to maintain deeper focus. It also means working only with optics designed for 65mm cameras. Probst said he finds the full-frame form factor to be a “sweet spot” on the spectrum between 16mm and 65mm.

One of the tactics favored by director Davis is the generous use of high-speed shooting on set, which gives her the option later on to use slow-motion moments in the edit. That meant Probst had to make sure he lit for a faster exposure and, in some cases, adjust his camera settings — while most of the video was shot with the Monstro sensor in 7K mode, shooting at higher frame rates meant dropping to 5K or even 4K modes. “We would often run the camera at 48 or 60 frames for everything, even with the performance of lyrics,” Probst says. “And she can always speed up the shot so that a vocal performance looks like it was shot normal. It winds up using extra hard drive space, and editorial assistants will have to speed up the shot 2x or 2.5x and re-sync the vocal with the playback, but it gives her that option.”

Most of the camerawork was on a Steadicam rig operated by Ari Robbins, well-known for his operating on La La Land and other films and music videos. “He was vital,” Probst says. “There’s a fine art to Steadicam in music videos. It depends on how much of a musical sense you have. There’s one tier of competency — holding the camera so it doesn’t wobble and the horizon line isn’t sea-sawing. But then there’s another level, which is where you choose to execute your move in terms of the song. You’re usually looking for a three-second editorial clip. A good operator who has a musical sense, or perhaps even a background as a musician, can feel the song and feel the transitions and create shots that work for the editorial timing.”

Lenses and Looks

Because digital sensor technology has become more consistent across vendors as it has matured, Probst sees lenses playing an increasingly crucial role as influences on the image. “Especially in the digital realm, I wind up using lenses akin to what cinematographers used to do with film stocks and processing — to create looks,” he explains. “Lenses can alter the way the image hits the sensor. It’s not doing something artificial, like post-filtering or trying to induce some sort of grain or film look. The actual light interacts organically, if you will, with the glass elements.”

Specifically, Probst appreciates the way vintage glass can affect contrast. Veiling glare, for instance, adds what Probst calls a gauzy or creamy quality. And that creaminess can be just the thing when shooting a full-on pop star like Ariana Grande in close-up. But shooting with vintage glass can mean losing practical, modern conveniences like the refined mechanics and ergonomic design of contemporary lenses. With all that in mind, Probst shot “Thank U, Next” with the brand-new Atlas Orion anamorphic set, which was designed to incorporate some of the visual characteristics of vintage anamorphic glass while integrating modern lens technology.

“I got my set a week before the shoot,” Probst recalls. “I had seen some early prototypes. A lot of people were very skeptical of these lenses, but they got a number of preorders and started delivering the first set of three late this past summer. And they worked really well for me.” Probst supplemented the three focal lengths in the Orion set with a Vantage Hawk anamorphic on either end of the focal range for both wider and longer options, but says he shot 96% on the Atlas set.

The large-format sensor helped Probst control depth-of-field; anamorphic glass gives out-of-focus areas a distorted look

Probst favored anamorphics for the stylized spin they put on the image, especially in the bokeh. Anamorphic lenses generally distort blurry objects, rendering orbs elliptically in the background of an image. He cites a theory about why this works so nicely: “In terms of our human visual systems — the evolutionary, fight-or-flight impulses that keep us from being eaten by lions — we sense out-of-focus objects and try to identify what they are quickly,” he explains. “When those images distort, they’re rendered in an unfamiliar way that relaxes our paranoia and allows us to focus on the area that is in focus, and not pay attention to the out of focus image. A lot of people feel that anamorphic photography is more pleasing because, in addition to the 2.40:1 aspect ratio, it draws attention to the subject and the background falls away in a unique rendering that allows our visual system to turn off.” In music video terms, that means your attention is focused where it belongs — on the artists performing for the camera — as the background trappings recede.

Soft Lights and Steadicam

Probst’s lighting was predominantly LED units, including Arri SkyPanels and LiteGear LiteMats. He favors the LiteMats partly because they’re available in different sizes that correspond to familiar fluorescent fixtures despite being much thinner and lighter. LEDs are also fully dimmable, with bicolor or full RGB color options, and wirelessly controllable. It beats cutting gels, which absorb light and thus cut overall output, he argues. Additionally, he declares an aesthetic preference for soft light sources, noting that they help with beauty work by cutting down on conflicting shadows on faces.

The SkyPanels were mounted overhead and softened through light banks. The LiteMats were often handheld by Probst. Since he had a Steadicam specialist relieving him of his usual camera-operating duties, the DP was free to manipulate the lighting as both Grande and the camera moved during shots. “I had a light that I could hand-hold and have some modeling light on her skin that I could move around on a whim, because I know where I want the light at any given moment,” he explains. “If you’re doing a scene where a character is walking all around a set, it’s difficult to create a lighting scenario where you can swing a Steadicam around 360 degrees and have every position on the set with perfect light for beauty work. Moving the lights around with her makes sense, and the best guy to do that is me.

“There were a lot of comments on social media like, ‘Oh, this looks great! What light did you use?’ Well, it’s just a LiteMat. But I’m walking around with the Steadicam operator and holding it right there, just out of frame.”

For the Mean Girls number (above), in which Grande, Elizabeth Gillies, Alexa Luria and Courtney Chipolone dance on stage in Santa Claus outfits, one of the high school’s students who worked as a theater tech was invited to operate the standard house lighting on a bar above the stage. A spotlight effect was created by a Jo-Leko — a Joker HMI 400W in a source 4 Leko for a light with strong punch and a long throw — and LiteMats were used to cast soft light from above.

Matthews Max Menace Arm

Probst also set up one of his go-to pieces of gear, the Matthews Max Menace Arm. “A key grip named Richard Mall designed it and received a technical Oscar for it a few years ago,” he says. “It’s the best thing ever. You can tuck it into a corner and stick a lamp way out, so you can get a lamp up there that you couldn’t rig into the location. It’s insane what the thing can do.” He used it to mount a SkyPanel with a light bank as a 3/4 frontal modeling light to make the performers pop on stage, and to augment the bluer feel of the spotlight. The rest of the scene was filled in by the practical theater lighting, as well as light shone into the ceiling of the auditorium to create ambient top light.

Atmosphere and Baking in the DP’s Intent

Finally, from front to back, the whole video was shot with generous atmosphere — enough haze to give the picture that certain something. Again, the tactic helped advance Probst’s strategy of capturing his desired look in camera, rather than counting on a colorist to find it in post. “Something happens in terms of how the haze reacts to light,” he says, comparing the use of atmosphere to the effect of certain lens choices. “If you have daylight streaming in a window and you have no haze, that’s one look. If you have haze in the room, you feel the light in the air. That’s a totally different quality, and you can’t create that without having haze in the air. The shadow in the corner might be a little bit milky, but that’s affecting the look in camera. When you look at it in a monitor on set, that’s exactly what it looks like when we finish it.”

Probst says the ability to have such precise monitoring capabilities on set is both a blessing and a curse — it’s easy to get tied up with a client who feels empowered to weigh in on every possible aspect of the high-definition image. Maybe he’s only half-joking when he threatens to create a “Video Tap app” for the video village that takes production back to the mid-1980s, when everyone had to look at a grainy black-and-white image flickering on a monitor and put their trust in the cinematographer.

“That’s not going to happen,” he admits. “We’re not going backward. But I don’t see the need to sit in a DIT tent, trying to create a look, and be disconnected from what’s happening on set or on camera. Creating that look on set is paramount. It shows your intent, and bakes in your intent. Through the optics and the haze and the depth of field and the format and the camera settings, it’s protected. As it goes through editorial, people get attached to that intent and end up sticking to it throughout post-production and in color. You may have more people commenting on the image [during production], but you can be much more specific, and also get more into the headspace of the director.”