We want to own a small LED kit for our company so we stop renting the basics every time. Every spec sheet says CRI 95+. One says TLCI 97, a pricier one lists SSI. I'm a producer, not a DP. Which number should drive the decision, and what is a sensible test before we buy?
CRI was designed around human vision and a small set of colour samples, so a high CRI does not promise good colour on a camera sensor. TLCI is the EBU's metric. It models a television camera and display and scores how much correction a colourist would need. Above 85, cameras usually need little or no adjustment. SSI comes from the Academy's Science and Technology Council. It compares the light's spectrum with a reference such as daylight or tungsten, written SSI[D55] or SSI[3200K]. It is now a SMPTE standard, ST 2122. It makes no assumption about any particular camera, which is why I trust it most for cinema work. Test: shoot a colour chart and two people with different skin tones under the panel and under a tungsten or HMI reference, same camera, same settings. Source: Academy SSI, https://www.oscars.org/science-technology/projects/spectral-similarity-index-ssi
Add a listen to that test. Put each panel at full output in a quiet room, stand where the actor would be, and record a minute. Fans ramp up at high output and some drivers whine at certain dimmer levels. I have had to ask for a light to be turned down mid-scene because of driver noise, and the DP was not happy with me.
We tested three panels with a chart and two of our actors, next to a tungsten reference, same camera and settings. The cheapest one had a TLCI in the 90s and still made one actor look slightly green next to the reference, which the colourist confirmed on the scopes. Bought the middle one, which also had the quietest fan in Vera's test. Cost a morning, saved us from owning four regrets.