Filming An Advertisement
In a departure from my bread and butter as a documentary filmmaker, I recently stepped back into the commercial world to be the Director of Photography for a couple of walk-and-talk advertisements as part of a political movement. The budget? Small. The talent? Policy experts instead of actors. The staff? Limited experience. Here’s how we did it:
These advertisements were originally pitched to our video team as a 3-5 minute set of informational videos in order to educate the general public around inflation and its many-faceted reasons for increasing the price of everyday goods and services. The videos were to be shot in a studio, with a grey backdrop and a talking head and were to be distributed on the internet as part of a larger campaign. As I read through the draft script, it struck me as the distribution plan was perhaps a bit stale: tested and safe to be sure, but with our limited audience, we would have to primarily rely on ad buys to get any sort of traction on the finished product - which was the plan. In talking with a couple colleagues, I pitched the idea of changing the plan a bit: instead of pre and post roll YouTube advertisements (where a lot of our ad-related content lives), we could try something new: those annoying TV screens on gas pumps. (Well, petrol pumps… but I’m living in the US now.).
The idea was simple: instead of the entire ad buy going to old sources (for which the effectiveness is questionable at times), we’d try a new avenue of audiences - one that resonated with the exact kind of person who’d be watching it: the everyday American feeling the pinch of higher prices. So, we got to work on the script - instead of introduction the policy people with “Hi, my name is….”, we’d take a more traditional TV advertisement approach in the form of a tight 30 second advertisement for each script. We’ll focus on the more dynamic one: a person pushing a grocery cart, talking about inflation directly affecting groceries. The talent would have a cart, unload it into his car and walk off screen. At the behest of the producers - and with a reminder that we’re not working with professional actors - we’d put a teleprompter in front of the camera.
Given as I’d pitched this idea, my task was to take away from the pitch and put together something where I could show all of the interested parties exactly what I wanted to do and the costs involved. I asked for a $50,000 budget - thinking we could hire a team of grips, M18s, the lot. $5000 was approved - and we used every penny.
Here was the original plan for the grocery shoot - a wide establishing shot, followed by a tight shot of our talent walking into the controlled lighting area. We’d then start our push in and cut to the B Cam to hide any roughness. The talent would then slam their trunk shut and walk off screen.
The original idea was to push 2 x M18s through heavy silk to create our beautiful key light, with a Quasar R2 in the trunk acting as a key for the B Cam. The camera itself would be on a dolly & tracks to get that smooth movement that we’re all used to in cinema, and we’d do multiple takes with multiple lenses to create enough coverage for our editor to work with and master in UHD.
Let’s take a moment to focus on the technical side of things - after all, that’s where the bulk of my interest and expertise lays - so we’ll break it all down.
Camera wise, we were shooting with Ol’ Faithful - the company’s trusty FS7 MK2. I know this camera inside and out, and especially know its quirks. We would be shooting in S-Log2 to deliver in REC709A, and since it was perhaps going to be unseasonably sunny during our shoot, I wanted to have the option of using the inbuilt variable ND as well as the option to put ND on the camera’s matte box. It is, however, a large camera - something that became a concern later - but suited for our task. S-Log, as a rule of thumb, is pretty forgiving in the grade - as long as I controlled the highlights on our talent and we didn’t have a stark white car, I should be good.
That said, S-Log is still log, and can be unforgiving if not exposed correctly. This is especially true for highlights - you can recover a lot of shadows with some careful grading, but if your highlights are overexposed, there's nothing you can do to bring them back - because they've clipped, there's no information there. S-Log is a well known beast at this point, and prior experience tells me that if I keep my skin tones under 60% IRE, the chances are that the subject - by far, the most important piece - will be correctly exposed under a grade. That said, I should really do a deep dive into IRE values as they relate to exposure. It's all about gate voltages, pretty cool stuff.
For my lens choice, there was only really one option that I thought we'd be able to get in our wild 4-day turnaround from greenlight to production: my personal set of CineAlta Primes. Now, these are PL mount (I rent them with an adapter here actually) and beautifully resolve with the Sony color processing in my opinion - but although they're 5lb lenses, they offer some benefits over the stock 24-70 that is my go-to all-rounder lens for 80% of my work. Firstly, they're T2 - and when I was planning on running 1/128 ND to keep our background as bokeh'd out as I could, that's needed. Secondly, they have moulded-in Mod 0.8 gears, which would work great for our remote focus pulling needs - otherwise, we'd have to adapt a 24-70 with a plastic gear which mostly works some of the time. Thirdly - familiarity with these lenses. This is the second time I've mentioned knowing gear inside and out, and there's a reason for that: if, as a DP, you know your gear to the point of muscle memory, you can also push that gear to its limits if needed. I knew that these lenses needed support to use with the FS7, so I accounted for that with a set of 6" rails that could also house the follow focus. I also knew going in that chromatic aberration wouldn't be an issue with these lenses and that they are all color matched: meaning I wouldn't have to micro-adjust the camera each time I changed lens. These little things all add up - knowing the gear is also knowing how each bit of gear interacts with one another and how those small working details can delay a production.
Talking of little things adding up, we need to talk about the big thing - and that's the grip requirement here. I asked for a piece of classic hollywood: the Fisher 10 dolly. To reliably execute on the same path each and every time in a smooth, there's very little that can beat sheer weight on a set of rails. And that's exactly what we did.
As 4:30AM blared through my phone, the day began. It began with a load-in in the office: all of the stuff is stored in our offices in a manner than needs to be reorganized. That said, we'd all planned to meet there, load equipment into a couple of trucks and get on the road to the location: an hour away in the back of a church parking lot. The editor of the piece - a former member of the church - had arranged the location. Perhaps predictably, DC weather wasn't on our side, hitting 95F at 9AM.
There was also a new factor to consider that hadn't shown its hand when I visited the site the day before for a scout: roadworks had started in the previous 12 hours, and there were a couple of very happy jackhammmer operators going about their day. So, we pivoted to be further away, sent our PA off to get some snacks, waters and tents, then set about building our set.
Part 2 to come soon!