Manufacturers "gaming" the CRI (Color Rendering Index) system?

tortuga

tortuga

Code Historian
Location
Oregon
Occupation
Electrical Design
I have been working on a few projects that require high CRI LED's (95+) and was told that lighting manufacturers now game the CRI (Color Rendering Index) system so really 95+ CRI led may not be very good at rendering color and newer TM-30 standard should be used instead of CRI.
I am curious if any of the lighting guru's on here with spectrometers have done any real world testing with both the newer ANSI/IES TM-30 method and CRI and compared it to a manufacturers published CRI as it seems manufacturers do not widely use TM-30.
 
Last edited:
tortuga said:
I have been working on a few projects that require high CRI LED's (95+) and was told that lighting manufacturers now game the CRI (Color Rendering Index) system so really 95+ CRI led may not be very good at rendering color and newer TM-30 standard should be used instead of CRI.
I am curious if any of the lighting guru's on here with spectrometers have done any real world testing with both the newer ANSI/IES TM-30 method and CRI and compared it to a manufacturers published CRI as it seems manufacturers do not widely use TM-30.
Click to expand...

Define "good".

If we're talking about printing, body shop and photography, the name of the game is repeatability and consistency.
If you drive your car into one shop, and hop out and look; then go to another shop, hop out and look, the colors should look the same.

It takes some time to switch gears and some Dodge Vipers were configured such that it hits 60 mph on first gear before hitting the limiter.. so it looks good on 0-60 time, however meaningless that maybe to real world driving.

Tube amps have more distortion than solid state which some consider more pleasing, but nonetheless, more distortion.
 
Flicker Index said:
Define "good".
Click to expand...
Good is light from a PAR 20 50W on a incandescent dimmer.
The dimmer changes the color temperature as well as the level when the lights dim. At low they glow steady and warmer probably 2200k -1800 kelvin.
Meat, produce and fish look their correct color.
Also glare, no visible light source from where a person sits or works.
Smooth lighting, not jittery shadows from multiple little light points.
No flicker.

Here is bad:
In commercial its food especially fresh produce coolers, but also a fish & meat department colors look bad, also the LED light source is 'spotty' not uniform. One store says they saw a decline in sales from their meat counter after a new LED cooler light made all the meat look green. (I found a fix for this)
A hair and nail salon went back to halogen MR16's or at least a few halogens to get colors to look right.

In residential its kitchen lighting especially LED under cabinet lights and task lighting again especially food often does not look right especially red tones. Led's make shadows or slight flicker some say LED's gives them headaches.
 
Flicker Index said:
Define "good".



Tube amps have more distortion than solid state which some consider more pleasing, but nonetheless, more distortion.
Click to expand...
Uuhh... thats just not true... but what does overpowered cars and amplifiers have to do with light?
Dont get me wrong- im all for free thinkers, im just curious...
Or are you being facetious? Maybe im severly dehydrated today..
 
rambojoe said:
Uuhh... thats just not true... but what does overpowered cars and amplifiers have to do with light?
Dont get me wrong- im all for free thinkers, im just curious...
Or are you being facetious? Maybe im severly dehydrated today..
Click to expand...

That they're designed to look good on certain metrics popularity sought after or required in specs.

Having the gear ratio such that shifting occurs at 55mph or 65 mph may or may not be "better" than having it at 60, but if you start the clock at dead stop and stop the clock when it reaches exactly 60, either variants will look worse on that test.

CRI, or Ra8 is just like that. To have the output tweaked such that the average of the eight designated standardized pastel colors is as high as possible. You can have 90+ CRI and still have absolutely terrible R9 (deep red, like tomato), and sometimes even near zero.

800 series fluorescent lamps with couldn't meet the efficacy requirements, but there was a legal exception to lamps with a Ra8 CRI rating of >90. So in those last few years, lamps with lower lumens per watt and 90 CRI were produced with higher lumen depreciation and lower overall LPW, because that passed the loophole.
 
tortuga said:
Good is light from a PAR 20 50W on a incandescent dimmer.
The dimmer changes the color temperature as well as the level when the lights dim. At low they glow steady and warmer probably 2200k -1800 kelvin.
Meat, produce and fish look their correct color.
Also glare, no visible light source from where a person sits or works.
Smooth lighting, not jittery shadows from multiple little light points.
No flicker.

Here is bad:
In commercial its food especially fresh produce coolers, but also a fish & meat department colors look bad, also the LED light source is 'spotty' not uniform. One store says they saw a decline in sales from their meat counter after a new LED cooler light made all the meat look green. (I found a fix for this)
A hair and nail salon went back to halogen MR16's or at least a few halogens to get colors to look right.

In residential its kitchen lighting especially LED under cabinet lights and task lighting again especially food often does not look right especially red tones. Led's make shadows or slight flicker some say LED's gives them headaches.
Click to expand...

Just about everything in your comment is outside the language of CRI and they're "irrelevant" in terms of specifications you mentioned.
 
Flicker Index said:
Just about everything in your comment is outside the language of CRI and they're "irrelevant" in terms of specifications you mentioned.
Click to expand...
My bad, I thought you asked for a definition of a "good lighting" I am not sure what 'language' you speak of but I am sure my discerning customers don't speak it either they just want "good lighting". Ever since California started requiring a minimum Color Rendering Index (CRI) of 90 all of a sudden everything here now says "90 CRI" on the package and I am curious if manufacturers improved anything or just figured out how make a LED that renders 14 pastel colors correctly but nothing else, or if anyone even checks that the CRI is what the package says. It seems these "lighting facts" labels are a self certification like CE.
 
tortuga said:
My bad, I thought you asked for a definition of a "good lighting" I am not sure what 'language' you speak of but I am sure my discerning customers don't speak it either they just want "good lighting". Ever since California started requiring a minimum Color Rendering Index (CRI) of 90 all of a sudden everything here now says "90 CRI" on the package and I am curious if manufacturers improved anything or just figured out how make a LED that renders 14 pastel colors correctly but nothing else, or if anyone even checks that the CRI is what the package says. It seems these "lighting facts" labels are a self certification like CE.
Click to expand...
My understanding of your situation was that you have a project which has a line item "shall have a CRI of 95 or greater".

If you're just required to come up with something that meets specifications, it's not your responsibility to satisfy the flaws in the specifications that someone else wrote, or the legal requirements.

If something says locate two related facilities. "Two facilities shall be no more than 3/4 miles apart" and doesn't specify otherwise and there's no convention in how such is dealt, the below meets the requirements even if it realistically is 40 miles apart, but that is the specifier's fault.
 

Attachments

  • one mi.jpg
    one mi.jpg
    214.3 KB · Views: 14
I think you can more or less tell if the CRI is low

Put your hand out into the light and you will see red light missing below 90cri

Below 80cri it gets bad and below 70 is what I call Walmart / zombie movie lighting

Is there a scientific way of testing? Especially with the switchable color temp lighting
 
B677 said:
Is there a scientific way of testing? Especially with the switchable color temp lighting
Click to expand...
I have seen expensive light meters film and video people use. It would be fun to borrow one but I would never buy one.
There is probably an 'app' for that but I have not tried any.
 
B677 said:
I think you can more or less tell if the CRI is low

Put your hand out into the light and you will see red light missing below 90cri

Below 80cri it gets bad and below 70 is what I call Walmart / zombie movie lighting

Is there a scientific way of testing? Especially with the switchable color temp lighting
Click to expand...
tortuga said:
I have seen expensive light meters film and video people use. It would be fun to borrow one but I would never buy one.
There is probably an 'app' for that but I have not tried any.
Click to expand...

CRI is antiquated, but still widely used industry standard.

https://www.waveformlighting.com/tech/cri-ra-test-color-samples-tcs
CRI used commonly only takes into account for R1 to R8. Ra is the average of these values.
The red rendition is generally related to R9, which is outside of scope of common CRI.

Your screen can't render them quite right, so remember those are just approximate. If you want to know what I mean, just look at pictures of clothes on website vs the actual one side by side.

More reading if you want:


Another real world thing they fail to consider is the UV effect. Many bright white printing papers and detergents contain optical brighteners that fluoresces with bluish light from UVA input. The bluish fluorescence combined with dingy yellow makes it look vibrant white. This is similar to how solid state fluorescent lamps (commonly known as white L.E.D. technology) uses yellow light from phosphors blended with blue to make white light.

Optical brighteners used in laundry products and paper production requires UVA to activate. They work in sunlight and traditional fluorescent lighting, but do not work with royal blue pumped L.E.D. type lights, because they do not emit anything in the spectrum needed to activate the optical brightener.
 
ron said:
I am no guru, but for rendering color, I would also consider the temperature, as the higher temps such as 5000K or 6500K is better than "warmer" colors (like 3500K)
Click to expand...
You are mixing up CCT (corelated color temperature) and CRI (color rendering index) which is a measure of light quality.
 
Flicker Index said:
CRI is antiquated, but still widely used industry standard.

https://www.waveformlighting.com/tech/cri-ra-test-color-samples-tcs
CRI used commonly only takes into account for R1 to R8. Ra is the average of these values.
The red rendition is generally related to R9, which is outside of scope of common CRI.

Your screen can't render them quite right, so remember those are just approximate. If you want to know what I mean, just look at pictures of clothes on website vs the actual one side by side.

More reading if you want:


Another real world thing they fail to consider is the UV effect. Many bright white printing papers and detergents contain optical brighteners that fluoresces with bluish light from UVA input. The bluish fluorescence combined with dingy yellow makes it look vibrant white. This is similar to how solid state fluorescent lamps (commonly known as white L.E.D. technology) uses yellow light from phosphors blended with blue to make white light.

Optical brighteners used in laundry products and paper production requires UVA to activate. They work in sunlight and traditional fluorescent lighting, but do not work with royal blue pumped L.E.D. type lights, because they do not emit anything in the spectrum needed to activate the optical brightener.
Click to expand...
Thank you very much for the info. "solid state fluorescent" is a new term for me.
 
You must log in or register to reply here.
Top