The Magical Science of 'Cymatics': Filming Sound & 16K Volts on the RED Dragon & Phantom Flex
I've shot many music videos during the course of my career, but never one quite like this.
Almost exactly a year ago I was approached by Shahir Daud and Nigel Stanford, Director and Musician respectively, to shoot a simply fascinating music video concept for a track on Nigel's album, Solar Echos, entitled "Cymatics" -- the concept of visual sound, that is to say scientific experiments that enable the viewer to directly see the effect sound waves have on matter. Nigel and Shahir showed me some experiments they performed and I was mesmerized. When I was a child, I initially wanted to be a scientist, and my love of science and engineering follow me to this day. I could not pass up an opportunity to shoot a video like this.
As you can see the effects of sound and matter are simply fascinating. I will be approaching this article from the standpoint of a cinematographer, so if you want more information about the making of the experiments themselves please visit Nigel's blog for a full breakdown of the scientific setups.
Cameras and Lenses
At the time we were prepping this, Shahir had a few must have elements concerning the look of the video. He wanted a moody, slick and bigger than life look, extreme slow motion and a final 4K master. The answer was clear: We needed anamorphic lenses, the (at the time) brand spanking new RED Dragon and the Phantom Flex (the Phantom Flex 4K was unavailable in NYC then).
Luckily, my friends the Diamond Brothers had just acquired an early Dragon, one of, if not the first in New York City. Why a Dragon, and not a standard Epic? Simple answer -- a bigger sensor area in anamorphic 4:3 mode that utilizes more of the anamorphic lens circle, and a higher dynamic range that will help us capture the electrical and fire based experiments in a dark moody room, more faithfully and easily. This was my first Dragon shoot and the one that completely sold me on the camera.
After the first time, I couldn't live without it, and here's why -- I learned on testing day that there was very little penalty shooting ISO 250 vs. ISO 2000 in terms of dynamic range. In previous iterations on the RED system, shooting at ISO below 800 meant you lost a significant amount of dynamic range up top in the highlights. This new feature enabled me to create an entire environment in one lighting setup that required very little tweaking for the myriad of different exposure levels needed for this shoot. This gave us two very unique and important advantages. The anamorphic lens set we shot on, Lomo Square fronts, were not particularly sharp wide open at a T2.8, so I had to keep them around a T5.6 to bring them under control optically speaking. This is quite the lighting feat when you need to jump from 25FPS to 96FPS, and maintain the same stop for optical consistency.
I created essentially "two cameras" in one, the first a 250 ISO T5.6 25P +.3ND camera and an ISO2000 T5.6 96FPS camera, and they both had to exist under the same lighting setup with no lighting changes. In general, I try not to use ISO as a means of controlling exposure, I prefer to stick to one ISO and use ND, but the dramatic change in exposure between the standard speed and high-speed material would have required very strong ND and would have caused color shifting. I was able to keep my exposure consistent between the different setups. I had to light the entire set to be friendly to the Phantom Flex and its lighting requirements. The difference here is that I had flexibility in the glass. The 1080p Flex, required spherical lenses to get the most resolution from its 1080p imager, and thus I could obtain optical quality at far lower T-stops to compensate for the skinny shutter of high frame rates by opening the iris. On the Phantom, I shot on Ultra Primes and Van Diemen rehoused Leica Macro lenses. The Van Diemens even made it on the Dragon from time to time for extreme macro work.
Some of the more challenging experiments to capture were the ones that required precise synchronicity between the camera and the experiment. I'm referring specifically to the water tube experiment. The challenge here is to use a skinny shutter to "freeze" the action in the air, while getting the exact frame rate and frequency of the sound coming out of the speaker to match up. Any loss of sync results in the water spiral appearing to travel down the path of the water or appear to climb up back in to the hose. We had to create a custom frame rate of 25.01FPS (if memory serves me correctly) to get the sync perfect and have the water appear to be frozen in time in a spiral.
Personally, I like to light the entire set once, create the world the camera can play in, and then just move around with impunity. Usually only minor tweaks are necessary like adding a bounce fill or a small edge light. The methodology of this lighting style is to light to your slowest lens, camera or frame rate. In this case, it was the Phantom Flex, with the potential of reaching the 2000 FPS mark for some shots. The Phantom sequences can be very light, hungry setups.
I lit the main set from overhead with two 20'x20' Silks, with three 10K fresnel Tungsten ArriMax lights, skirted to form an overhead softbox, all of which was suspended from a box truss we rigged to the rafters of the building. I also had two MaxiBrute 9,000W lights in the corners of the room, on dimmers and in frame to create flares and give a warm edge light to the entire set. There were also two 5K Tungsten pars on rolling stands to enhance the edge lit look if needed in the wings. The lighting plots below will give you a sense of what we did to create the feel of the Cymatics video.
I had been asked several times since this video premiered whether we used a technocrane to achieve the camera moves. In this case, it was a good old-fashioned Fisher 11 dolly and Fisher Jib -- operated carefully. William Amenta, operated this video for me, and really brought his "A" game. There was some post stabilization to remove some of the jitters of having a long jib on a dolly. Some of the side benefits of having 6K of resolution: you can perform your post stabilization and cropping, then still be able to downscale to 4K for a pristine master.
The most extravagant dolly move, the grand finale shot of the piece: The Tesla Jump. Shooting at around 500 FPS we were able to capture our Faraday suit character jumping at an achingly slow speed on the Phantom, but in order to feel the movement with such slow motion, the dolly had to be whipping by the Tesla setup at full tilt. (Quite the difficult feat with only 20' of track.) My G/E department really pulled out all the stops (no pun intended) to get that hulking Fisher 11 up to speed and stop it in time to prevent the camera flying off the track.
One fun little note, when a 200,000 volt, 8' Tesla Coil is running, all your kino flos start glowing, despite no connection to power. The electromagnetic waves stimulate the gas in the tubes, and make them glow. The closer they are to the Tesla Coil, the brighter they glow. Yeah, science!
Bringing to life all of these amazing experiments is quite fun. Each had their own challenges and caveats. This kind of shoot is the most fun, simply because of the challenge, collaboration between departments, and overall cool factor of the content itself. I love my job.
If you missed Part I of this two-part series, in which Shahir Daud shares his experience directing Cymatics, you can check it out here.
[EDIT: I had missed a small detail in my initial post: We shot 25P not 24P for this project for international release. The post is updated accordingly.]