As LED Lighting continues to mature, more and more users are searching for LEDs that can replace HMIs.
LED technology for video lighting is in an interesting place right now. I’ve been shooting video and digital cinema with LED lighting now for more than a decade, as many of you have also been. My older Tungsten video lights like my beloved Arri Softbank IV kit are seeing less and less use. I now only pull out my Arri fresnels possibly once or twice a year and the tungsten lighting is almost never used on talent, they’re mainly used for lighting backgrounds in interviews and narrative scenes. More on why this is later.
My first pro-level LED video lights were the Coollights LED panels. These were made in China by a small company ran by an American businessman. I spent some time speaking with him about the lights and giving feedback on how they worked in the field on real shoots. I used the Coollights for quite a few years as they worked well, despite the limitations of relatively low output and they had a very distinct green spike in the output that we tried to mitigate through the use of minus green CC filters, which worked okay.
Once I decided to replace the fading Coollights with some new LED panels, I researched the subject and ended up, after much renting and trying out friends’ LED panels, with a pair of the Aputure LS-1S Lightstorm panels a few years ago. They were much more powerful than my Coollights and had more options like wireless control, built-in external heat sinks and the light, control box and power supply were modular, so the power supply was able to be placed away from the light fixture itself, meaning that when the panel was 10 feet up in the air, the controls for the panel weren’t, and I could place the lighting control unit down at ground level where I or various gaffers I work with could access the light’s controls. The Aputures were a step up, they had probably three to four times the output of my previous panels, they were daylight balanced and the wireless control was very handy when I was shooting alone and didn’t have a gaffer to man the light.
The Lightstorms are far from perfect. Their configuration requires that you have an AC cable that goes from the AC outlet to the power supply, which is a small, rectangular box that sits on the ground. You have a four-pin DC cable that goes from the power supply and plugs into the light’s control box that’s usually placed hanging from the light’s yolk or lower down on the light stand if the light head is placed high in the air. There’s then another Hirose-type cable that must be run from the control box to the Lightstorm fixture itself. Once all of this is set up, if you have to move the Lightstorm on its stand, you have to also move the AC cord, the power supply, usually sitting on the ground, along too. Frankly, it’s a pain and messy. But most challenging of all is the lack of sheer output, the Lightstorm is still an LED panel. LED panels typically have hundreds of 5mm LED bulbs installed and the resulting shadows from these hundreds of individual bulbs cast hundreds of separate micro shadows on a subject and on walls behind the subject. You end up needing to run most (not all) LED panels into some sort of diffusion, softbox or diffusion frame. This further reduces the LED panel’s output.
Enter The LED Single Source Light
Up until a few years ago, we really didn’t have single source LED video lights. The technology hadn’t yet evolved to a single large LED emitter being possible, largely because of heat buildup and dispersion and largely because there hadn’t yet been enough demand for a single source LED from the non-video lighting markets. Eventually, single source LEDs came to the video market. One of the first and most successful more affordable single source video lights also came from Aputure, the 120t. This was a single source LED instrument that was first only available in Tungsten 3200k versions, but it was relatively power, affordable and had a whole lineup of accessories available for it including an attachable Fresnel lens that meant the 120t could be used bare for output to a parabolic or traditional softbox or diffusion panel or the Fresnel lens setup could be mounted, giving users a focusable, high-quality LED Fresnel spot source.
There were many other single source LEDs that began flooding the market as well from various lighting companies, many of them established like Altman, Arri, Mole-Richardson, as well as many others from startups, Kickstarters and many various Chinese companies with brands you weren’t familiar with. Soon the market was relatively flooded with different LED single source open face and Fresnel lights. Some challenges soon became apparent when trying to choose which LED single source lights to rent or buy. There’s no standardized set of measurements on exactly how to rate an LED light’s color accuracy or output. We arrived first at using CRI (Color Rendering Index), but using that spec alone soon proved to be close to useless as wild, unsubstantiated claims were made that upon careful examination, didn’t hold up. Using just CRI soon evolved to adding a second color rendering spec, TLCI (Television Lighting Consistency Index) that, although not an approved international standard, is recommended by the EBU and is finding success among manufacturers.
As far as lighting output though, I’ve found that the standard output ratings of LED instruments vary widely. This isn’t always the fault of the manufacturers either, LED lights behave differently and the output varies in many ways from the output of traditional Tungsten and HMI video lighting sources. LED instruments, in comparison, consume far less wattage and amperage than traditional video lighting for supposedly equivalent output. I’ve heard formulas that manufacturers and users utilize such as, “Take the LED’s wattage consumption and multiply it by three to four times to derive it’s output when compared to a traditional Tungsten light.”
Through experience, I’ve found that these formulas and manufacturer suggestions on actual light output are completely unreliable. Using my light meter and measuring output at 1m is fine and will give me a rough idea of the light’s actual output, but the fall off of LED light output is different than Tungsten or HMI. So the actual light output that falls on my subject is dramatically different than the normal lighting output falloff that I’d experience with a Tungsten or HMI light in the same situation.
In The Real World On Set
Case in point, a couple of months ago, I was hired to DP a set of interviews to be shot on set of a national TV commercial. Talent was all celebrities/actors. If you’ve worked with name-brand talent before, you know that when you’re going to shoot interviews or scenes with them, they come to set with managers, agents, publicists, their own hair and makeup team, personal assistants and sometimes a load of other entourage. As the DP, none of this matters to me, but in a way, it totally matters as my producer expects me to have the set completely lit and the background squared away well in advance of talent because we’ll often only have them available for as little as five to 10 minutes total to shoot their interview or scene.
There were numerous challenges such as the house the commercials were shooting at was physically small and there was literally no room to set up our interview set, so the Ads relegated us to set up in the front yard. These interviews were to be shot in bright, 2 p.m. daylight, so we couldn’t merely set up a camera and single-source bounce card. In order to make the talent look their best, I had to remove the overhead sun from the equation and then light the talent. I utilized an 8×8 frame with a heavy silk overhead to remove most of the overhead sun, side negative fill to give me back some contrast and I then needed a key source that I was going to utilize camera left, placing my talent in their chair camera right and my interviewer just to the left of my lens.
We performed this same shoot the week before at a different location and because of a lack of budget, I had utilized a 6×6 Ultra Bounce as my key source, which had worked well from a light perspective but from a talent perspective, some of the talent had been squinting too much from having a large reflective source lighting them. I’ve found that talent varies widely as to how light-sensitive they are, and my producer and editor requested that on the next round of interviews, I mitigate that issue by not using an Ultra Bounce as my key source. I was able to get the client to give me a little more budget so that we could rent a large LED source to use as a key source. We needed to rent an LED as we had no budget and resources for a generator for a larger HMI or Tungsten key light.
The Mole-Richardson Senior
After doing some research and looking at the photometrics, I chose to rent the Mole-Richardson Mole Senior (5K Equivalent) Varicolor Fresnel. This is a top of the line LED Fresnel with 2700 to 6500k color temp adjustment range unit with plus and minus Green adjustment, DMX, Wireless control. Its power consumption is 800 Watts. Notice how the light is advertised (by the rental house, not Mole) as a 5K equivalent? The first question is that comparison to a 5K Tungsten or a 5K HMI light? Assuming they mean a 5K Tungsten light, that would mean a 6X output factor when compared to the Tungsten instrument. I’ve shot with 5K Tungsten lights for exteriors, and if the Mole had the equivalent output, that would be enough for how I needed to use it to compete with the high ambient levels. I was skeptical, though, but only had enough budget and resources to rent the largest, most powerful LED instrument I could find for the budget the client had given me.
We needed to see plants and trees and some set stuff happening behind talent as these interviews needed to look like they were shot on set. So that negated being able to build a “box” of Duvetyn around the talent to knock down all of the ambient light, so I did the best I could with some negative fill on the sides. The Mole Senior was a nicely constructed light with a 10-inch lens. It was also a fairly large and heavy light, requiring a junior pin light stand, but at 800 watts power consumption, it could be plugged into the house AC, which was necessary because even the production didn’t have a generator, they too were using LEDs and HMIs that could all be ran off of house current.
I had originally wanted to key the talent with the Mole using my Chimera Medium Quartz softbox, but I quickly found that with the high levels of ambient light I was fighting, I didn’t have enough lighting “horsepower” to achieve the light level I wanted to use the softbox, so I had my gaffer move the Mole more to the side and just put a sheet of diffusion over the bare LED source with the Fresnel lens moved out of the way.
In this way, I was able to achieve the output I needed without blinding my talent, the setup worked and looked good, but I wasn’t able to use a softbox in front of the Senior because it simply didn’t have the horsepower that the rental house claimed it had, at least in a real-world situation. In my experience, the output was roughly the same as a 1.5K HMI, which for our situation, wasn’t enough. After using the Mole, it was a nice light with great lighting quality, but for working outside, I needed an instrument with more output to achieve my lighting plan.
Mole rates the output of the light at 19,000 lux at 6500k at 10 feet, but when I added ambient sunlight and diffusion in front of the light, I simply needed more horsepower to get the job done. We made it work, but it wasn’t ideal. Lesson learned; when in doubt, it’s always best to have a surplus of horsepower and turn it down than to have not quite enough.