J.W. Speaker Denouncing Low-CCT White Light?

[jtr1962]

What are you talking about? Both UMTRI studies showed more blue light producing more glare.

Guess what you can also do? You can also cut down CCT while further cutting down blue. Any sort of improvements of cutting down the blue spike could be applied to lower CCT products to bring their blue light content even lower, e.g. the voilet pump Sunlike.

Here's the problem in a nutshell with your position on this-this obsession with reducing blue light over and above everything else. Neither me nor JustAnOldFashionedLEDGuy are advocating specifically for more blue light in the spectrum. Rather, we're advocating for more light at the wavelengths with greater response in the scotopic and mesoptic. Yes, because of the way CCTs are calculated that generally means higher CCTs, but not necessarily high CCTs. There's a difference. 4000K is higher but not high. Nobody is advocating for anything over 5000K regardless, and frankly it seems there's little benefit in most cases going above the low 4000s for roadway lighting or headlights.

Ideally, it would be nice if through some combination of phosphors and primary pump wavelengths we could get rid of the unfortunate blue spike present in most LED spectra but we have to live in the real world. There are other concerns like manufacturability and repeatability, as well as efficiency. However, the fact remains you need some of that higher frequency light (i.e. shorter wavelengths) to stimulate the rods which become increasingly dominant at lower light levels. You go with a higher CRI spectrum, you cut down on the blue spike, which is desirable (no argument there). But then you're suggesting cutting CCT to reduce blue further. Fine, but this also cuts down on the needed non-blue wavelengths in the mesoptic and scotopic regions, which means at that point you're compromising safety solely to meet an arbitrary goal of having low CCT lighting. And even that goal has questionable merit given the wide range of CCT preferences among different populations. On top of all that, lower CCT LEDs are less efficient. That's an important parameter for nearly all DOTs as most are cost constrained.

Violet-pumped LEDs sound nice in theory but I'm dubious if they'll ever supplant the current blue-pumped ones. The WPE of violet emitters still lags that of blue. They're also more expensive at present. And the Stokes losses converting to white light are always going to be higher. That's just physics. Efficiency is an important parameter in lighting. For most lighting uses reducing the blue content somewhat at the expense of a huge efficiency hit is a nonstarter. The exception might be where you need to mimic a black-body spectrum very closely. We don't need to do that for streetlights or headlights.

You complain me and JustAnOldFashionedLEDGuy aren't linking to research supporting our conclusions. For starters quite a bit of this research is paywalled. I'm not going to link to something that people won't be able to freely access. For another, I've saved loads of stuff on my hard drive (including that UMTRI study you obsess over which I saved back in January 2010) but lots of it is no longer online. There's what is called a working knowledge that I draw on. JustAnOldFashionedLEDGuy has a lot more of that than I do. I've designed LED products but I work mostly on the electronics end, making the drivers. I deal very little with the optical characteristics of those products. As such, I lack the depth of knowledge of JustAnOldFashionedLEDGuy in that area, although I'm sure my knowledge of electronics trumps his. But as a hobby I've had a desire to learn about this topic, especially in areas which affect me personally, light streetlighting or interior lighting.

 

[jtr1962]

Continuously lighting highways is a poor/questionable practice, btw.

And Caltrans' experience with ridding themselves of continous expressway lighting.

Here's something I might actually agree with you on. I don't like HPS and low CCT LED roadway lighting for lots of reasons besides the technical issues already discussed in this thread. They cast a depressing yellow pallor on the street scape, render cool colors poorly, and their light color clashes with moonlight/starlight, effectively ruining the ambience of night. There were even theories floating around that part of the reason for the crime increase in the 1970s/80s was due to the wider use of HPS lighting. I recall a town in the UK where crime dropped when they replaced their HPS lighting with HID, so there may be something to that theory.

Anyway, given a choice of lighting with lousy light sources, versus no light at all, I'll pick the latter in a heartbeat. At least with no light, I can use whatever type of light I prefer while cycling or walking.

 

[cetary35]

Here's the problem in a nutshell with your position on this-this obsession with reducing blue light over and above everything else. Neither me nor JustAnOldFashionedLEDGuy are advocating specifically for more blue light in the spectrum. Rather, we're advocating for more light at the wavelengths with greater response in the scotopic and mesoptic.

You don't have the research to back this claim up, remember those studies on school children you linked you support refuting. Not that they're relevant to night driving anyways. You don't see in hundreds of lux outside at night like the studies you posted you posted. You also aren't trying to make out lettering at short range either. You're trying to identify moving targets.

You complain me and JustAnOldFashionedLEDGuy aren't linking to research supporting our conclusions. For starters quite a bit of this research is paywalled. I'm not going to link to something that people won't be able to freely access.

Excuses, excuses.

For another, I've saved loads of stuff on my hard drive

What's stopping you from posting that here? I'm sure you'll fabricate some excuse like the above. Otherwise stop wasting our time with your opinions.

Ideally, it would be nice if through some combination of phosphors and primary pump wavelengths we could get rid of the unfortunate blue spike present in most LED spectra but we have to live in the real world.

Yes, that's why we drop the CCT.

 

[cetary35]

There were even theories floating around that part of the reason for the crime increase in the 1970s/80s was due to the wider use of HPS lighting.

More of your unfounded theories! But hey I'm sure there's like millions of articles on this and getting more then one should be easy? Btw, the study you linked on HPS vs MV is essentially meaningless. Apparent brightness is not target detection or seeing ability or measurable safety.

 

[jtr1962]

Btw, the study you linked on HPS vs MV is essentially meaningless. Apparent brightness is not target detection or seeing ability or measurable safety.

Apparent brightness is just another term for greater response in the scotopic or mesoptic regions. And that is relevant to target detection on roads.

 

[cetary35]

Apparent brightness is just another term for greater response in the scotopic or mesoptic regions. And that is relevant to target detection on roads.

Provide links then that quantify lower reaction times in drivers and greater detection distances then. You won't.

 

[jtr1962]

What's stopping you from posting that here? I'm sure you'll fabricate some excuse like the above. Otherwise stop wasting our time with your opinions.

You can't upload anything except images, so that's what's stopping me. Besides, you already seem to have your mind made up. It's all about low CCT for you, so you'll twist anything I post to mean something different.

 

[cetary35]

You can't upload anything except images, so that's what's stopping me. Besides, you already seem to have your mind made up. It's all about low CCT for you, so you'll twist anything I post to mean something different.

You'd just make that excuse. I knew it! You can work around it like how I worked around getting your contradictory statements on high CCT fluorescent lights in jpg format in the comments.

 

[cetary35]

They cast a depressing yellow pallor on the street scape, render cool colors poorly, and their light color clashes with moonlight/starlight, effectively ruining the ambience of night. There were even theories floating around that part of the reason for the crime increase in the 1970s/80s was due to the wider use of HPS lighting. I recall a town in the UK where crime dropped when they replaced their HPS lighting with HID, so there may be something to that theory.

Old man yells at clouds!

 

[jtr1962]

You'd just make that excuse. I knew it! You can work around it like how I worked around getting your contradictory statements on high CCT fluorescent lights in jpg format in the comments.

Those statements aren't contradictory. Driving is a task, some of those studies on task lighting apply. You're assuming there's never any need to discern fine details or make out lettering while driving. Hello, road signs???? What about being about to detect targets which are cooler or neutral colors? You'll miss those if the light is too warm. Point of fact, the ~4000K, give or take, which many have been saying is optimal, is considered to be neutral white. That means it provides the best compromise at being able to detect targets which are warmer colors or cooler colors.

As for your so-called workaround, so now I'm supposed to convert multipage .pdfs or word documents into pictures just so you'll tear them apart if they don't support your preordained conclusion? I have better things to do with my time.

 

[jtr1962]

Old man yells at clouds!

I'm making a point that for every person who might love a given light color there will be another who hates it. That's why we avoid using extremes unless there's some compelling technical reason. There's no case to be made for 2700K/3000K outdoor lighting beyond aesthetic (and only then for some portion of the population). There's no case to be made for anything much beyond 5000K for similar reasons. So in the absence of that it makes sense to go towards the middle. Most of the data supports doing so from a public safety perspective as well.

 

[cetary35]

Those statements aren't contradictory. Driving is a task, some of those studies on task lighting apply...4000K, give or take, which many have been saying is optimal

Your opinion at this point and not fact. None of the UMTRI studies, UC Davis, or Clanton findings support this.

so now I'm supposed to convert multipage .pdfs

You have more excuses then a kindergartner on his first day of going to school.

 

[jtr1962]

Your opinion at this point and not fact. None of the UMTRI studies, UC Davis, or Clanton findings support this.

And they don't support your view, either. JustAnOldFashionedLEDGuy did a far better job picking apart your argument than I did.

Learn about scotopic and mesoptic responses then get back to us.

 

[cetary35]

JustAnOldFashionedLEDGuy did a far better job picking apart your argument than I did.

No, he didn't. He broke down and tried to fall back onto his imaginary credentials.

And they don't support your view, either.

Prove it.

Learn about scotopic and mesoptic responses then get back to us.

That's cute. How about you learn how to upload pdf's so we can see all those "studies" of yours.

 

[jtr1962]

Here's one that is still online:

https://vtechworks.lib.vt.edu/bitstream/handle/10919/23325/Assess_Imact_Clr_Cntrst.pdf?sequence=1&isAllowed=y

For color recognition, the fluorescent light type provided a significantly greater color recognition distance than did either LED for small targets. For pedestrian clothing color recognition, the 6000 K LED outperformed both the fluorescent and the 3500 K LED light type by approximately 130 ft. Due to the difference between the detection of pedestrians and the recognition of their clothing color, the results suggest that the 6000 K LED may be the superior light for rendering color.

This one looks interesting also:

https://www.fhwa.dot.gov/publications/research/safety/15047/15047.pdf

I didn't yet read most of it since it's 240 pages. It has some parts dealing with mesopic vision applied to the driving environment.

 

[cetary35]

Here's one that is still online:

https://vtechworks.lib.vt.edu/bitstream/handle/10919/23325/Assess_Imact_Clr_Cntrst.pdf?sequence=1&isAllowed=y

For color recognition, the fluorescent light type provided a significantly greater color recognition distance than did either LED for small targets. For pedestrian clothing color recognition, the 6000 K LED outperformed both the fluorescent and the 3500 K LED light type by approximately 130 ft. Due to the difference between the detection of pedestrians and the recognition of their clothing color, the results suggest that the 6000 K LED may be the superior light for rendering color.

For one that study...

did not consider the impact of glare.

Furthermore, the study did not mention the CRI of the tested lamps. It's kind of an important factor when you're claiming something like this and the title of your research is, "Assessment of the Impact of Color Contrast..."

For pedestrian clothing color recognition, the
6000 K LED outperformed both the fluorescent and the 3500 K LED light type by approximately
130 ft.

The study is also limited to whatever the lamps they had on hand that day.

The results of this study are limited to the light sources and the targets tested in this
investigation.

The results of the detection tests were also mixed leading to this.

The results indicate that a blanket factor cannot be applied across all design scenarios.

 

[cetary35]

This research, from Alan Lewis of the IES showed minimal difference in detection distances between 2700K, 3500K, and 6000K LED. Further, it's funny you pulled some of Ron Gibbons research as his studies from 2016 showed HPS having the best detection distances in all but rain. In which case 6000K narrowly edged out. Btw, that's another aspect that I don't believe was mentioned in the 2011 study you linked.

Also, wanted to add this from Dr. Mario Motta's website. It's of the latest IES Guidance. It advises against the use of high CCT outdoor lighting.It's a pretty succinct explanation. This represents the culination of multiple studies. More importantly, it presents a consensus. It represents the collective findings of Ron Gibbons, Nancy Clanton, and so on.

"Shorter wavelength light does allow better color rendition and peripheral acuity, but warm white provides very good visibility on major streets while reducing perceived glare and lower traffic areas do not need maximum brightness as headlights combined with less bright street lights work at slower speed limits."

 

[bykfixer]

And actually "is" isn't....or is it?
This is like watching a 2 proppelerheads ping pong match on an Atari. You fellas sure know how to have a good time.

 

[cetary35]

I'm making a point that for every person who might love a given light color there will be another who hates it. That's why we avoid using extremes unless there's some compelling technical reason. There's no case to be made for 2700K/3000K outdoor lighting beyond aesthetic (and only then for some portion of the population). There's no case to be made for anything much beyond 5000K for similar reasons. So in the absence of that it makes sense to go towards the middle. Most of the data supports doing so from a public safety perspective as well.

You know what your commentary reminds me of, smoking. Your commentary reminds me of the smoker, someone who acknowledges that what they do is bad. However, they try and find some odd/arbitrary middle ground in which everyone loses. I'm sure similar thought processes yielded this ineffective "compromise".

 

[jtr1962]

For one that study...

I knew you would bring up glare. Be aware that glare is more a function of light going where it's not wanted, and source size, then it is of CCT. We've seen lots of complaints about glare with LEDs owing to their much smaller source size. Sure, we could put a diffuser over LED fixtures similar to the ones used for HPS/MV but for some reason, probably efficiency, we didn't.

Furthermore, the study did not mention the CRI of the tested lamps. It's kind of an important factor when you're claiming something like this and the title of your research is, "Assessment of the Impact of Color Contrast..."

The study is also limited to whatever the lamps they had on hand that day.

The results of the detection tests were also mixed leading to this.

No kidding. I didn't say it was a perfect study, but you're limited to what is out there. I'd love to see a study using LEDs of similar CRI/identical intensity, in identical fixtures, and at CCTs of 2700K, 3000K, 3500K, 4000K, 4500K, and 5000K. Moreover, I'd love the study to do both on and off-axis target detection, color recognition, reaction times, etc. In the real world these studies cost a lot of money and time. All too often they're commissioned by someone to justify a predetermined conclusion. So we have what we have.

This research, from Alan Lewis of the IES showed minimal difference in detection distances between 2700K, 3500K, and 6000K LED. Further, it's funny you pulled some of Ron Gibbons research as his studies from 2016 showed HPS having the best detection distances in all but rain. In which case 6000K narrowly edged out. Btw, that's another aspect that I don't believe was mentioned in the 2011 study you linked.

And speaking of research designed to justify a predetermined conclusion, here you have it. More of this blue spike nonsense from the AMA. This is sticking their two cents in an area where they have no expertise. The superior off-axis detection distance of the 6000K was simply dismissed as not important, even though off axis target recognition is very relevant in city driving.