What is an acceptable variation between the led light color temp reading & actual WB

T

The Fog

Member
What is an acceptable variation between the led light color temp reading & actual WB

Today I was shooting an interview with a Sony FS7. Normally, I shoot in log, but today I shot in HG and had to do a manual WB. The WB reading seemed to be off compared to what was indicated on the light. The light is an Intellytech Lite-Cloth. Their CRI was rated at 96 on the website.

Anyway, I got curious and started down the rabbit hole of taking my light, a white card and camera, and did a manual WB from 3200k up to 9000 (the color temperature range of the light).
So, 32k read 34k. 4k read 4100 and so on. Until I hit 55K, which read 49k. After that, the difference increased from 1k up to over 2k. That's a big variation if you ask me. At 7k, I got a WB of 62k and at 8k, I got a reading of 6800!

I also have a Lupo Superpanel and it was almost spot on with a CT range from 32k to 56k, the variance was from 100 to 200--pretty good to me.

If you had a light with a wide decrepancy between the actual WB from the camera and the light's reading, do you trust it or return it? I'm close to the returning it option.

Thanks!
 
The color temp numbers displayed in the camera are just for reference and not to be considered an absolute "true" color temp.
 
Why say that?

Anyway it is easy to test with a grey card. In post or with an RGB waveform (or histogram)

I would expect the numbers on a light to be further off than a camera.

An rgb waveform like smallhd and others would enable you change the colour temperatures in the lights until the "matched"
 
There's a trend today to rely on specs, and get cross when they appear wrong, but a tungsten 3200K lamp was NEVER 3200K. Oe could be 3100, another 3250, and of course at some times of day the supply voltage would drop just a few volts and the colour temp go down maybe close to 3000K. In Broadcast TV when I first started years ago, we'd run the lights in the studio at 70% on the fader, because we could then go up to full for nearly an extra stop, or down to 50% for one the other way and of course the CT varied wildly when this was done. With every source a tungsten, the white balance was simply done to give white. Nobody really cared about what it really was on a calibrated meter - the cameras just made it white. I suspect it would have been closer to 3000K in many cases. Then we added discharge lighting and started throwing gel about to warm up or cool down sources. The presets in most cameras would be 3000/3200/5600 or similar steps, and you looked at the light and picked the one that made most sense. If you have multiples of camera models, setting them all to the same one is enough. Doing a fine white balance with a card is often just pointless with the control we have in post.

Treat figures as indications, not accurate specs.
 
is usually around +/-100-200k difference by my experience. Even the HMI from the Broncolor have about +/-100K difference.

Also at different power level the performance varies
 
Don’t want to name ‘n shame, but suffice to say many of the better bi-color, & RGB+, LEDs CT settings are incorrect.
 
You must log in or register to reply here.
Back
Top