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

cetary35

Newly Enlightened
Joined
Aug 29, 2017
Messages
56
We're not talking about blue-enriched white light, whatever that's supposed to mean.
My reference to blue-enriched is just a reference towards deliberately blue shifted white light. Don't be so sly. There are no studies showing any definitve improvement in real world driving conditions of higher CCT LEDs like 5000K over lower CCT light such as 3000K LED. Furthermore, this study by UMTRI showed no improvement in driver seeing ability with higher color temperature lights. Additionally, research conducted at the CLTC at UC Davis showed the same thing.

During EPIC-sponsored CLTC laboratory activities, a broad range of products between 2,200 K and 6,500 K has demonstrated that similar color rendering, discrimination, and visual acuity can be achieved.

With the research I linked above, higher color temperature lights do not provide any significant improvement in real world driving conditions. However, they do increase glare.

Now here's where it gets interesting.
No, no it doesn't. The study you linked compared two extremes, spectrally deficient 2100K HPS and white MH. It did not show or quantify any improvement in seeing ability when comparing lower color temperature white light. In fact, as far as I can see, it didn't even quantify the "improvement" in seeing ability of MH. It also didn't mention the color temperature or SPD of the MH lamp.

The study you linked to appears to be designed to support what was already a foregone conclusion,
No, San Diego adopted 4000K induction, and they still install 4000K LED to this day along major streets.

based on my personal experience
Not supporting studies but ok.

We have written about the debate over LED street-light color temperature or CCT. I remain convinced that cooler CCTs, perhaps in the 4000K range, are optimum because I believe we see better at night under such conditions.
And I remain convinced smoking in public spaces isn't a bad thing, but ultimately that and this aren't backed up by research. Obviously, I'm being sarcastic, but the point stands.
 
Last edited:

cetary35

Newly Enlightened
Joined
Aug 29, 2017
Messages
56
The 4300K LEDs are great. Unfortunately, they changed the lights on some side streets to 3000K for misguided reasons
It has been well documented by the AMA that high color temperature LED street lights worsen glare particularly for older drivers.

some surveys showed most liked the 4300K better.
Ah, more mis-leading information. Time and time again when 4000K LED and 2200K/2700K/3000K LED is tested in cities, the public consistently prefers the warmer color temperatures. This was shown in Riverside, CA; Davis, CA; Phoenix, AZ, and Pepperill, MA.
 
Last edited:

jtr1962

Flashaholic
Joined
Nov 22, 2003
Messages
7,331
Location
Flushing, NY
My reference to blue-enriched is just a reference towards deliberately blue shifted white light. Don't be so sly. There are no studies showing any definitve improvement in real world driving conditions of higher CCT LEDs like 5000K over lower CCT light such as 3000K LED. Furthermore, this study by UMTRI showed no improvement in driver seeing ability with higher color temperature lights. Additionally, research conducted at the CLTC at UC Davis showed the same thing.
The first study failed to adjust for intensity. The blue tinted bulbs definitely had much lower output than the untinted bulb. Given that, I'm honestly surprised that the blue-tinted bulbs didn't perform worse than the untinted ones. When you start comparing different CCT light sources, you need to have similar intensity AND CRI or the comparison is meaningless. Not sure what the CRI of the tinted bulbs was, but it was certainly well under the ~100 of the untinted one. I wonder how the blue-tinted bulbs would have performed without the handicaps of lower intensity and lower CRI.

The second study mentions the International Dark Sky Association and the American Medical Association. The former wants lower CCT and even possibly monochromatic outdoor lighting to benefit astronomers. It doesn't care whether or not this lighting is inferior for actually seeing things. The latter is concerned about blue light effects on the sleep cycle, except this is entirely irrelevant for outdoor lighting because people are typically exposed to greater intensity blue-light sources once they get home, and they go to sleep at least several hours after being exposed to streetlighting. By then any effects of blue light exposure have worn off.


With the research I linked above, higher color temperature lights do not provide any significant improvement in real world driving conditions. However, they do increase glare.
Glare is more a function of fixture or headlight design/aiming than anything else. The basic premise is to put the light where you need it, and cut it off where it isn't. Spectral content in the blue area is a secondary effect, but as I said you can have higher CCT with lower blue light content by going with higher CRI.
No, no it doesn't. The study you linked compared two extremes, spectrally deficient 2100K HPS and white MH. It did not show or quantify any improvement in seeing ability when comparing lower color temperature white light. In fact, as far as I can see, it didn't even quantify the "improvement" in seeing ability of MH. It also didn't mention the color temperature or SPD of the MH lamp.
You missed everything else in the study. Yes, in this case they compared HPS and MH simply because that's all that was available in terms of outdoor lighting when the study was originally done in the early 2000s. The major takeaways are as follows:

1) Higher CCT has higher apparent brightness for any given light intensity level
2) Higher CCT makes it easier to discern small details (relevant for driving if you want to see objects several blocks away)
3) Higher CCT results in greater peripheral vision (from the study: However, in a simulated roadway application (no vehicles to run over the subjects!) where they tested peripheral vision, they found that the subjects had faster reaction times under MH than under HPS, all other conditions being equal.)
During EPIC-sponsored CLTC laboratory activities, a broad range of products between 2,200 K and 6,500 K has demonstrated that similar color rendering, discrimination, and visual acuity can be achieved.
Per what I wrote above that is only true only if the intensities are adjusted to compensate. I found this chart on my hard drive:

1664836030102.png

So in theory you can match a light source with lower blue content to give equivalent seeing by increasing the intensity, but by the time you do that the overall blue light flux will likely be the same as a higher CCT source (and you'll be wasting several times the power). If you keep intensities the same, you end up with lower apparent brightness, less ability to discern details, and most importantly far less peripheral seeing using a lower CCT light source.

I think your entire reason for starting this thread was a misinterpretation of JW Speaker's position. It sounds more to me like they're advocating against excessively blue, high CCT which provide no or negative improvement in seeing, rather than advocating going to 3000K or less for everything as you appear to want them to. The studies I've seen (and I've read lots of them) mostly point to 4000K to 5000K being the optimal point for headlights and streetlights. JW Speaker appears to be rallying against sources much over 5000K, and especially those which don't even appear white. There's a lot of junk on the market with an angry purple appearance. I agree these lights are more a fashion statement, and actually make seeing/glare worse. But the pure white ones which might be 4100K, 4500K, etc. are pretty much the sweet spot. They also happen to maximize energy efficiency, although that's a secondary concern for headlights.

And I remain convinced smoking in public spaces isn't a bad thing, but ultimately that and this aren't backed up by research. Obviously, I'm being sarcastic, but the point stands.
It sounds more like you're being dismissive of any research not supporting your desired conclusion, like the way you dismissed the entire part about peripheral vision favoring higher CCT sources.

I'm somewhat at odds with the outdoor lighting industry for different reasons, mostly having to do with their reluctance to even consider high-CRI lighting. However, if studies showed my reasoning on high-CRI outdoor lighting to be baseless, then I wouldn't pursue the matter any more.
 
Last edited:

bykfixer

Flashaholic
Joined
Aug 9, 2015
Messages
19,573
Location
John 3:16
Dirty/scratched glass can play a big role too.
An old car I had with a quarter million miles on it had a (seemingly) million tiny scratches from living in a state with a lot of snow in winter for 10 years, meaning sand specks causing a spiderweb look to the glass at night.

Dirty glass is pretty common too and that can make a big difference in how oncoming light appears.

I doubt computer generated charts and graphs incorparate a lot of real world stuff like that.
 

och

Newly Enlightened
Joined
Sep 15, 2011
Messages
128
The 4300K LEDs are great. Unfortunately, they changed the lights on some side streets to 3000K for misguided reasons. People were associating valid complaints like light trespass due to poor fixture design, or excessive intensity, with "whiter" light. A minority of complainers forced the change on the rest of us, even though some surveys showed most liked the 4300K better. Thankfully they at least kept them on the arterials. Now I'm trying to get them to switch them back on the side streets, my reason being public safety. I can't see as well with the 3000K lights, both because they're lower wattage and lower color temperature.

I'm not sure about your location, but here in NYC I've never seen a warm white LED street light. They started changing to LED around 2013-2014, and pretty much all of them are right around 4300k. Prior to that they had these terrible high pressure sodium HID street lamps, they are still present on some highways and airports.

I'm barely old enough to remember the mercury lamps. Those got switched out in the early 1970s. Almost everyone immediately complained about how awful and dim the high-pressure sodium were.

NYC definitely still used many mercury vapor lights in the 90ies, perhaps even early 2000s, before they started to switch to high pressure sodium. They had the same terrible color rendering as HPS, but with a blue/green huge vs orange.

Weird that most cities never used metal halide HID lamps - they have the same efficiency as HPS/Mercury vapor, but with much better CRI and color temperature. Years back I had a HID backyard fixture, and I swapped its factory 100W 4300k lamp for a Philips 3000k lamp, and other than slow start it was great. I believe Philips claimed 90 CRI vs 70 for the 4300K lamp.
 

jtr1962

Flashaholic
Joined
Nov 22, 2003
Messages
7,331
Location
Flushing, NY
It has been well documented by the AMA that high color temperature LED street lights worsen glare particularly for older drivers.
Again, if the concern is with blue light content, you can use high CRI, high CCT lighting, and also possibly decrease the intensity, without compromising seeing relative to low CRI, high CCT lighting.
Ah, more mis-leading information. Time and time again when 4000K LED and 2200K/2700K/3000K LED is tested in cities, the public consistently prefers the warmer color temperatures. This was shown in Riverside, CA; Davis, CA; Phoenix, AZ, and Pepperill, MA.
The problem here is public safety. The Davis, CA study is particularly concerning. The original plan was a choice between 4000K and 5700K at 2115 and 2326 lumens, respectively. The higher apparent brightness of 5700K relative to 4000K, along with the actual higher output, likely made the 5700K too bright, so the choice of 4000K made sense. Had the 5700K choice been perhaps 1800 lumens, which would have given similar apparent brightness, I suspect things would have been more even. However, that's not what concerns me. The final choice of 2700K at 1800 lumens does. Per my chart above, 2700K is only about 1/3 as effective in terms of lumen effectiveness, or perceived brightness. Therefore, if they insisted on using 2700K, they should have went with around 5400 lumens, not 1800. Effectively, they compromised seeing (and safety) by a huge margin.

The Pepperill, MA study seems to be based more on astronomical concerns than what is best for actually seeing. The comparison of the LED retrofit and original HPS says it all. Both seem to be very poor at actually lighting the roadway, which is the primary purpose of streetlighting.

The Riverside, CA study didn't even test any sources higher than 3000K.
 

jtr1962

Flashaholic
Joined
Nov 22, 2003
Messages
7,331
Location
Flushing, NY
Dirty/scratched glass can play a big role too.
An old car I had with a quarter million miles on it had a (seemingly) million tiny scratches from living in a state with a lot of snow in winter for 10 years, meaning sand specks causing a spiderweb look to the glass at night.

Dirty glass is pretty common too and that can make a big difference in how oncoming light appears.

I doubt computer generated charts and graphs incorparate a lot of real world stuff like that.
Yes, thank you a million times. "Glare" is unfortunately a very subjective thing not lending itself readily to studies.

The trains I used to ride to NJ in the 1980s often had very dirty windows. That made the glare of light sources a lot worse. And back then outdoor lighting was mostly HPS, which as you know has little blue content.
 

och

Newly Enlightened
Joined
Sep 15, 2011
Messages
128
Here is an interesting photo I found on Wikipedia - Brooklyn Bridge in 2008. You can see a mixture of blue and orange lights on it, the orange ones are obviously HPS, but the blue one are most likely mercury vapor.

Incredible how fast time flies. It is so weird seeing NYC skyline without the new tower.

Pont_de_Brooklyn_de_nuit_-_Octobre_2008_edit.jpg
 

cetary35

Newly Enlightened
Joined
Aug 29, 2017
Messages
56
The first study failed to adjust for intensity. The blue tinted bulbs by definitely had much lower output than the untinted bulb. Given that, I'm honestly surprised that the blue-tinted bulbs didn't perform worse than the untinted ones.
No, you're wrong again. You didn't read the study through.
A neutral density filter (Lee Filters .15ND 298) was applied to balance the
luminous intensity of the Blue and TH bulbs so that approximately the same photopic
illuminance level was obtained for all three light sources at the observer's eye.
Additionally, page 16 of that study also had a table confirming that illuminance levels were the same for both the blue-tinted halogen lamp and the warmer un-tinted halogen lamp.
It doesn't care whether or not this lighting is inferior for actually seeing things.
Also wrong, see here.
Outdoor lighting fixtures that shield the light source to minimize glare and light trespass help prevent light pollution.
6. Avoid blue lights at night -Blue-rich white light sources are also known to increase glare and compromise human vision, especially in the aging eye. These lights create potential road safety problems for motorists and pedestrians alike.

Again, if the concern is with blue light content, you can use high CRI, high CCT lighting, and also possibly decrease the intensity, without compromising seeing relative to low CRI, high CCT lighting.
but as I said you can have higher CCT with lower blue light content by going with higher CRI.
Argumentum ad nauseamI The problem with this is that even with a perfect black-body radiator increasing CCT will still increase blue. Also misleading.

Per what I wrote above that is only true only if the intensities are adjusted to compensate. I found this chart on my hard drive:
I don't care what a random chart you found on hard drive says. It's not a study.
 
Last edited:

jtr1962

Flashaholic
Joined
Nov 22, 2003
Messages
7,331
Location
Flushing, NY
I'm not sure about your location, but here in NYC I've never seen a warm white LED street light. They started changing to LED around 2013-2014, and pretty much all of them are right around 4300k. Prior to that they had these terrible high pressure sodium HID street lamps, they are still present on some highways and airports.
I'm in eastern Queens. They swapped out quite a few of the original 4300K LEDs for 3000K, including on my block. Yeah, I still see a few HPS here and there. It's a good reminder for how awful street lighting here used to be.
NYC definitely still used many mercury vapor lights in the 90ies, perhaps even early 2000s, before they started to switch to high pressure sodium. They had the same terrible color rendering as HPS, but with a blue/green huge vs orange.
In Queens those seemed to have disappeared by the 1970s. I didn't know they were still in use elsewhere until the 1990s. I don't get around to the other boroughs all that much.
Weird that most cities never used metal halide HID lamps - they have the same efficiency as HPS/Mercury vapor, but with much better CRI and color temperature. Years back I had a HID backyard fixture, and I swapped its factory 100W 4300k lamp for a Philips 3000k lamp, and other than slow start it was great. I believe Philips claimed 90 CRI vs 70 for the 4300K lamp.
I thought that was strange myself. You can actually decrease the wattage quite a bit by swapping out HPS for HID, and still end up with similar apparent brightness. That and the better CRI/more pleasing color temperature have me wondering why we didn't do mass HID installation in the 1990s. Parking lots got them pretty quickly. I remember thinking why doesn't the city put them in streetlights? Only thing
I can think of is maybe the lifetime wasn't up to par. That would have meant a lot more labor replacing burnt out bulbs.
 

jtr1962

Flashaholic
Joined
Nov 22, 2003
Messages
7,331
Location
Flushing, NY
I don't care what a random chart you found on hard drive says. It's not a study.
The chart was part of a larger study which I can no longer find a link to. Again, you're being dismissive of whatever fails to support your desired conclusion.

We're talking about lighting the roadway effectively, not light pollution. By definition any amount of light of any color will mean light pollution. If you're concerned about light pollution, don't light at all.

Also, how do you explain the fact the blue-rich starlight and moonlight doesn't seem to compromise human vision or create glare? If there are any issues with outdoor lighting, it's intensity more the color temperature. We often light things too brightly outdoors, far more than needed just to see.

BTW, "glare" is mostly a function of source size, which is why complaints about it went way up when LED headlights/streetlights became more common. I don't recall similar complaints when we installed HID lights instead of HPS in parking lots or gas stations because the source size was similar. We could have designed fixtures with a similar emitting pattern as older light sources, but chose not to for styling/cost reasons. So sure, a headlight which is 1/5 the size of a halogen headlight is going to appear more glaring, even if it were 2700K. The problem is people see glare, see a high CCT LED source, and immediately assume the problem is the light color, not the source size.

The problem with this is that even with a perfect black-body radiator increasing CCT will still increase blue. Also misleading.
No, more like misunderstanding on your part. Higher CCT lighting requires lower intensity to achieve a similar apparent brightness. This was actually well-documented when cities went en masse from MV to HPS streetlights. The photopic lumens of both were similar, the HPS had more lumens per watt, so the general consensus was a 180 watt HPS lamp could replace a 400 watt MV lamp. Unfortunately, people complained immediately at how much dimmer the streets appeared. Getting back to what I wrote, while true that blue light will comprise a higher percentage of a high CCT spectrum, the fact you can reduce intensity to get the same apparent brightness means the overall blue light flux might be similar, perhaps less, than that of a low CCT source. A full study would need to be done. Assuming the overall blue light flux is the same, the higher CCT light will offer better peripheral vision and higher contrast than a low CCT source putting out the same amount of blue light.
 
Last edited:

cetary35

Newly Enlightened
Joined
Aug 29, 2017
Messages
56
The major takeaways are as follows:

1) Higher CCT has higher apparent brightness for any given light intensity level
2) Higher CCT makes it easier to discern small details (relevant for driving if you want to see objects several blocks away)
3) Higher CCT results in greater peripheral vision (from the study
The first point doesn't prove anything good. The second point is not relevant to driving. It specifically talks about high light levels like...
This effect may be helpful for people doing inspection, surgery, sewing, and other detail-oriented visual tasks.
You'll also read the wording may be helpful as well. The third point was not quantified in any meaningful manner. Additionally, no improvement was seen for objects viewed head on.

So in theory you can match a light source with lower blue content to give equivalent seeing by increasing the intensity,
Once again, you're wrong. The tables on page 16 of the UMTRI study completely invalidates any of your pet theories.

It sounds more like you're being dismissive of any research not supporting your desired conclusion
No, you.

like the way you dismissed the entire part about peripheral vision favoring higher CCT sources.
You haven't provided any quantifiable evidence to show this. Ok, neat, you can see slightly better at some far off angles. Now how much better? Is it even relevant given the baggage of blue-rich white light? Without those numbers, this point is not relevant.
 

cetary35

Newly Enlightened
Joined
Aug 29, 2017
Messages
56
The problem here is public safety. The Davis, CA study is particularly concerning. The original plan was a choice between 4000K and 5700K at 2115 and 2326 lumens, respectively. The higher apparent brightness of 5700K relative to 4000K, along with the actual higher output, likely made the 5700K too bright, so the choice of 4000K made sense.
Good lord are you being misleading. Apparent brightness is not target detection. It is not actual brightness. It is not a "fc" of "cd" value.

Therefore, if they insisted on using 2700K, they should have went with around 5400 lumens, not 1800. Effectively, they compromised seeing (and safety) by a huge margin.
Once again, your opinion. Not a quantified study.

Both seem to be very poor at actually lighting the roadway
Nancy Clanton's study shows the superiority of warm white on roadways counters this poorly founded statement.
 

och

Newly Enlightened
Joined
Sep 15, 2011
Messages
128
I'm in eastern Queens. They swapped out quite a few of the original 4300K LEDs for 3000K, including on my block. Yeah, I still see a few HPS here and there. It's a good reminder for how awful street lighting here used to be.

Interesting, I am in Southern Brooklyn, and now that you mentioned it, would be nice if they would replace the 4300K LEDs with 3000K ones on residential streets. Keep the brighter 4300Ks on main roads.

In Queens those seemed to have disappeared by the 1970s. I didn't know they were still in use elsewhere until the 1990s. I don't get around to the other boroughs all that much.

Apparently the necklace lights on all the bridges, including Queensboro, were mercury vapor until 2009.



I thought that was strange myself. You can actually decrease the wattage quite a bit by swapping out HPS for HID, and still end up with similar apparent brightness. That and the better CRI/more pleasing color temperature have me wondering why we didn't do mass HID installation in the 1990s. Parking lots got them pretty quickly. I remember thinking why doesn't the city put them in streetlights? Only thing
I can think of is maybe the lifetime wasn't up to par. That would have meant a lot more labor replacing burnt out bulbs.

I'm surprised the city moved to LED so quickly, usually the city is slow to do anything. But I am pretty positive most NYC street lights in the 90ies and early 2000s were mercury vapor, I remember the lights were blue before they switched to HPS orange.
 

cetary35

Newly Enlightened
Joined
Aug 29, 2017
Messages
56
The chart was part of a larger study which I can no longer find a link to. Again, you're being dismissive of whatever fails to support your desired conclusion.
And without a link, you're being deceitful.

We're talking about lighting the roadway effectively
Good golly man! The IDA and IES are entirely about lighting the roadway effectively. That's part of how you minimize light pollution, by minimizing the amount of wasted artificial light. That's why they also advocate for things like tight optical controls to keep the light on the roadway and fully sheilded lights to prevent wasted uplight everywhere outside.

Higher CCT lighting requires lower intensity to achieve a similar apparent brightness.
you can reduce intensity to get the same apparent brightness means the overall blue light flux
More arguementum ad nausium. Not relevant to how well one "sees." Again, you're trying to mislead with the wording "apparent brightness." You aren't using words like target detection, detection distance, or the like.

This was actually well-documented when cities went en masse from MV to HPS streetlights.
If it was so well documented, I'm sure we have a ton of research on it we can pull from.... There probably isn't any research on it, or it was so limited in scope as to not being relevant.
Assuming the overall blue light flux is the same, the higher CCT light will offer better peripheral vision and higher contrast than a low CCT source putting out the same amount of blue light.
What is this? Why would you handicap a lower CCT with higher blue?
 

och

Newly Enlightened
Joined
Sep 15, 2011
Messages
128
I also find that bluer light gets absorbed more by fresh black asphalt vs yellow halogen light.
 

bykfixer

Flashaholic
Joined
Aug 9, 2015
Messages
19,573
Location
John 3:16
There’s always so much passion in the automotive section.
When I first joined here this sort of passion could be found throughout the place.
Most of the nit pickers and hair splitters are long since gone.

Back then a new SureFire thread, any SureFire product would hardly make it past day 1. Maglite too.

Folks would get banned and return with new user names and go right back to hurling insults at one another. Most just organically went away.

This thread here is pretty peaceful compared to those days. Folks debating the post without hassling the post-er has always been what the staff has requested. So far so good.
 

jtr1962

Flashaholic
Joined
Nov 22, 2003
Messages
7,331
Location
Flushing, NY
And without a link, you're being deceitful.
Well, I can't find it. There are related papers though:


Just search for lumen effectiveness multiplier.
Good golly man! The IDA and IES are entirely about lighting the roadway effectively. That's part of how you minimize light pollution, by minimizing the amount of wasted artificial light. That's why they also advocate for things like tight optical controls to keep the light on the roadway and fully sheilded lights to prevent wasted uplight everywhere outside.
And I advocate those things also. In fact, I submit that most of the issues with LED lighting have more to do with light getting where it's not wanted than with the color temperature. The smaller source size is a big problem also.
More arguementum ad nausium. Not relevant to how well one "sees." Again, you're trying to mislead with the wording "apparent brightness." You aren't using words like target detection, detection distance, or the like.
It's pretty much the same thing. Lumen effectiveness multiplier is simply calculating the eye's response in the mesopic region to a given spectrum, as opposed to in the photopic region where it's usually done. Generally, eye response correlates well with reaction times and other visually oriented tasks.
If it was so well documented, I'm sure we have a ton of research on it we can pull from.... There probably isn't any research on it, or it was so limited in scope as to not being relevant.
See above. It's called lumen effectiveness multiplier. A 4000K MH source is nearly 8 times as effective as a HPS of the same photopic intensity.
What is this? Why would you handicap a lower CCT with higher blue?
I'm not. I'm simply saying that the lower CCT will need to have a higher intensity to give the same visual response/reaction time as a higher CCT source. Therefore, even though blue comprises a smaller part of its spectrum, the higher intensity may result in the same or higher blue light flux. Note the operative word "may". A full study would need to be done. That's obviously way beyond my scope (or yours).
Nancy Clanton's study shows the superiority of warm white on roadways counters this poorly founded statement.
So why do my eyes say otherwise? I have a direct A/B comparison of 3000K and 4300K LEDs in my neighborhood. Invariably I fail to see things as quickly on streets lit with 3000K. Why do several others here say they also see better with higher CCT LEDs?

Again, look at the scotopic and mesopic responses. Much of the light from low CCT sources is virtually useless at the intensities typical of streetlights or headlights.
 
Last edited:

jtr1962

Flashaholic
Joined
Nov 22, 2003
Messages
7,331
Location
Flushing, NY
I also find that bluer light gets absorbed more by fresh black asphalt vs yellow halogen light.
Which is actually a good thing because that means more contrast between the road and road markings, as well as anything on the road.
 
Top