Proper light quality measure - CRI sucks

F

funkytwig

Active member
Just came across an interesting video interview with Alan Roberts, ex BBC tester, who has developed a new way of checking light quality. The video is at http://www.newsshooter.com/2015/02/...pert-alan-roberts-has-the-surprising-answers/ (was a couple of years ago) and all his results are published at http://www.gtc.org.uk/members-area/tlci-results.aspx.

Basically, CRI was developed in the 60s are not much use for measuring light quality for even lights used for shooting news. The results cover a lot of lights (Lightpanels Astra scores very well, a lot of the Bowens ones are awful.

Would be great if he had tested Aputure Amron HR672 but his software is freely available if anyone has a suitable probe. Maybe we should see if we can get some results from other makes/models populer here. It tends to be generally the higher end stuff he is testing.
 
Last edited:
TLCI is a much better metric than CRI for judging the spectral qualities of LEDs, & lamps that score well on a Sekonic Spectometer are usually a safe choice. But even TLCI won’t always account for a small color casts, & lamps with similar scores may look quite different when compared side-by-side.
 
I thought these two articles (and their links) are pretty informative, if not definitive. One by a DP with a technical bent, one by an engineer with a visual bent. Worth reading, imo:

TLCI vs. CRI, CQS, etc.: How do they stack up?
Does TLCI do a better job than CRI, CQS and other color models at measuring LED light quality? Maybe, and maybe not.
By Art Adams January 30, 2017
https://www.provideocoalition.com/tlci-vs-cri-vs-cqs-stack/


ICA’s 2016 LED color quality survey
Indie Cinema Academy has done the legwork so we don’t have to
By Adam Wilt July 01, 2016
https://www.provideocoalition.com/indie-cinema-academys-2016-led-color-quality-survey/


Short version: TLCI is imperfect but useful. Different meter and different cameras see light differently.


What references do other people find helpful?
 
Metering color is hard with the current tools we have available and I think it's a good base to start with and at least you know if it's worth a serious look, but real-world testing also needs to be done. I've found a high CRI is usually a good indicator. LED has advanced so much in the past year that they all seem to pass the test but the big diffrence now is build quality and features of the LED lights.

I find a lot of lights don't live up to the Kelvin ratings that are stated and this causes a lot of issues when using them with other lights. I use a Sekonic C-300 for reviews I do for Newsshooter.com and also at work when I mix brands on a shoot. Having the ability to tune the kelvin temp of a bi-color fixture to dial in a set of lights is an invaluable feature to have. Without that meter I would have to guess and your eye sees those colors diffrent than the cameras sensor does.

As for TLCI I dont have a meter or know of one that is available that does those readings and like some of you have stated it isnt the answer to solving the problem either.
 
Erik Naso said:
I find a lot of lights don't live up to the Kelvin ratings that are stated and this causes a lot of issues when using them with other lights. I use a Sekonic C-300 for reviews I do for Newsshooter.com and also at work when I mix brands on a shoot. Having the ability to tune the kelvin temp of a bi-color fixture to dial in a set of lights is an invaluable feature to have. Without that meter I would have to guess and your eye sees those colors diffrent than the cameras sensor does.
Click to expand...

One of the reasons I've always recommended not mixing LED brands in the same shot/shoot(not necessarily not matching temp wise, but green push, etc.). I've recently added two Hive 100-C's and they do not match my LitePanel Astra's at 3200K. Not even close.
 
Run&Gun said:
One of the reasons I've always recommended not mixing LED brands in the same shot/shoot(not necessarily not matching temp wise, but green push, etc.). I've recently added two Hive 100-C's and they do not match my LitePanel Astra's at 3200K. Not even close.
Click to expand...

Yeah it's really tough. At work I decided to get four ASTRA's for our main lighting to just have enough of the same fixtures on set but I always bring in some extras because fernels are also needed but it all seems to work out :-D
 
Tm30

Tm30

Interesting no one is talking about TM30. It's not a single number (2 and a few charts) but seems very useful.
 
You must log in or register to reply here.
Back
Top