In the context of your personal observations on lack of red content, stating what you have found out about your personal color vision would be interesting. Not everyone shares the same proportions of receptor types.How about finding out more about ECE's rule of 5% 610...780nm requirement on red spectral content? Is there evidence it is too little? Your links do not answer this question.
"Flashlight Etiquette" for Cyclists
- Thread starter zespectre
- Start date
Help Support Candle Power Flashlight Forum
Steve K
Flashlight Enthusiast
So is it better for others to not know what you are, or is it better for them to know?
Or perhaps another way to state it is: how do you think cyclists are treated, and what would be the advantage of others not knowing what you are?
again.. you make it sound like there are people on the road that are hunting for cyclists, and we need to stay hidden.
The technical aspects of transmitting data via the LED are simple. If the other vehicle can sense the light from your bike, it can receive the data and identify you as a bike, and therefore something potentially traveling at a different speed (a good assumption if you are on a road where the speed limit is 55mph). Even an RF system wouldn't require much power to send a response to an interrogation pulse, or just transmit every 100ms or so. Consider how little power a cell phone needs to stay in touch with the tower network periodically.
Flamingtaco
Newly Enlightened
- Joined
- Feb 26, 2008
- Messages
- 31
It's quite the dilemma. I would like for motorists to know exactly what I am when they act appropriately, but in answering the other two questions, we know some motorists do not give cyclists the same care as they afford other drivers, some are downright disrespectful and abusive, and some take advantage of a cyclist's vulnerability to commit assault and theft. I have fallen victim to all but the last, and I feel we should open ourselves up to additional nefarious activities.
In my situation, due to relative safety of the area I live in, confusing drivers works well. Middle to upper income, don't have college kids or hoodlums driving around at 3am, just a few people coming and going to the industrial park I work in. My main threat is afternoon rush hour drivers not paying attention, but the vast majority are headed the opposite direction. I get to use a lot of residential streets,too. For others... inner city commuters... must traverse high crime areas... I don't think it's possible to develop a single solution for all, but I think a solution for most is possible.
I think using light information would be fine under ideal conditions, but what about low sun angle, or reflections? Would this require a complex lens setup to be able to pick out light signals in the direction of the sun's orientation? Would the sun blow out the sensor, rendering light signals invisible? RADAR detectors are impervious to sunlight, and also rain/fog over short distances, and those go as cheap as $30 in the us, maybe that type of technology can be readily converted (change frequency range, use proper transceiver, modify software?)
I think we have gotten well off the topic of etiquette
Maybe roll this off into a "Future Tech" thread?Flamingtaco
Newly Enlightened
- Joined
- Feb 26, 2008
- Messages
- 31
Agreed re: receptors. My personal experience is modern LED's make white text pop loudly, against a reddish but largely dark background, unless I am very close to the sign, or the background isn't red. My wife, who claims to be night blind but I know just has poor night vision, claims no color at all except at certain angles where it seems almost as if you are getting a chromatic aberration or something on the micro level. To me, this appears as the background color peaking suddenly, kinda like when you get the on angle burst from a prismatic reflector. This is with newer, high reflectivity stop signs. I need to get out and take notice of the older signs.
Bicycle reflectors: Can't claim to have spotted a red reflector once in ten years, but the white ones on the spokes work well when you're the cross traffic, and they cannot be missed when reflecting LED light. The red reflectors my boys broke off their bikes years ago react as well to the $5, 2w halogen flashlights (long forgotten in drawers scattered about the house) as they do to an XM-L pushing 5w. This seems to parallel the measurements in the articles I linked, but is purely anecdotal, and not measurable without extensive testing by doctors that care not one bit about the afflictions facing cyclist at the mercy of a society hell-bent on LED's.
Since the US is not following most of the ECE trends, I did not look further into the spec's after it became apparent the NHTSA was not going to follow suit on the glare issue (most pressing discussion at the time). I also got a busier in life, so... you know. Hopefully someone can follow up on that and some of the other questions regarding ECE and the German regs as I haven't been given a reprieve from life, and I think I've blown my research quota for the year in just a few days
There's a few new parts on the bike that need a good wearing in, and I'd like to get a battery pack idea moving along as well.Sorry (everyone) for turning this into a long-winded thread, sometimes I let hope outshine reality a wee bit too much and forget that TL;DR is as bad as NMI.
Nothing in this document concludes or suggests inadequate red rendering with LED headlamps.
Whoah! What's after your comma does not follow at all from what's before your comma. (In fact, a strong case can be made on human-perception grounds that green, not red, is the optimal color for tail lamps. Tail lamps will always be red, because that decision was made many years ago and switching now, even to a technically superior color, would be unfeasibly expensive and create huge safety nightmares in the changeover.)
True.
There are LEDs that give superb color rendering, but they have not (yet?) been used to make headlamps because everyone's busy chasing bluer and bluer and bluer appearances. All the blue look does is increase discomfort glare and make the advertising copywriter's job easier. I would much rather have LED headlamps made of LEDs like these (they really are every bit as good as claimed).
Compared to LEDs, automotive HID headlamps actually tend to be lower in red RSP. Both HIDs and the LEDs that have so far been used to make headlamps are usually relatively(!) deficient in red compared to tungsten-halogen headlamps. However, that information alone is not sufficient to say that the red output of an HID or LED headlamp is inadequate for the task. For more, see this (note numbered points at the end of the abstract), this, and this.
What do you mean by "required"? The colors of a sign are a part of its message, which is why research such as I just linked is done.
Keep in mind also, all of the emitters in your aforementioned results are 10 years out of date -- that's an even longer time than it sounds like, given the rapid pace of LED evolution. But even if we disregard that or assume today's white LEDs (as used in headlamps) are spectrally comparable to the ones in that paper, it still looks like you're reading conclusions into that paper that aren't actually there. It is silent on the question of whether stop signs and red reflectors are adequately conspicuous -- the UMTRI paper I posted, on the other hand, looks at that question and concludes "probably no problem here".
Road signs are not within the purview of NHTSA's regulatory activities. They are regulated by FHWA.
Whoah, again! Nobody said anything about LEDs "producing no red". This looks like you have pulled something that isn't in the paper you read, and stretched it until it's even further from anything actually measured or assessed in that paper.
Cree is not an LED. It is one of many makers of LEDs. They produce a large line of products, and that line is constantly evolving.
OK...what grade is it?
That's puzzling. All trailers over a particular size have been required to have conspicuity retroreflectors for many years now, even if they were made after the date it was required as original equipment.
It is popular among vehicle lighting enthusiasts to point to the UN ("ECE", "European", "E-code") headlamp standard as superior because of its tighter control of glare towards oncoming traffic. However, that is a difficult proposition to defend based on all the facts. For one thing, it's not an apples/apples comparison because countries that use the UN headlamp standard virtually all specify much lower headlamp aim angles than the US does. A UN headlamp aimed to the US specification (0.7% declination) instead of the common non-US specification (1% to 2.5% declination depending on vehicle characteristics and headlamp mount height) produces oncoming-traffic glare generally within the range of values produced by US headlamps aimed to the US spec. A US headlamp aimed downward to the UN specification produces oncoming-traffic glare generally within the range of values produced by UN headlamps aimed to the UN specification.
Moreover, the downward aim specified in UN-headlamp countries gives the driver much shorter seeing distance than the aim specified in the US. Take a single headlamp designed to meet both the UN and the US standards -- many such headlamps exist. Mount it at 0.59m height, and aim it 1.2% down (near the higher-aim end of the UN spec). On a straight, flat road, the cutoff intersects the road surface (ending the driver's seeing distance) 161 feet in front of the car. Now aim that same lamp, at the same height, 0.7% down (US). On the same road, the cutoff intersects the road surface 277+ feet in front of the car -- an increase of 72% over the seeing distance with UN-spec aim! The shorter seeing distance with UN-spec aim creates a safety deficit that is enormously far out of proportion to the nuisance created by the higher glare level that comes with the US-spec aim.
Yes, I just called US headlamp glare a "nuisance", because the data show us pretty clearly that's what it is. The question has been studied and studied and studied, and the results keep showing the same things: no causal link between glare and collisions (i.e., drivers tend to behave safely and achieve safety in the presence of very disturbing glare levels), but definite causal links showing most drivers don't have nearly as much light or seeing distance as they need, and people are dying because of it -- mostly pedestrians, with bicyclists second on the list. The data (i.e., science) supports greater low beam intensity and the longest practicable seeing distance for a significant improvement in roadway safety, even taking glare into account. The safety effects of glare intensity and seeing light intensity are not equal, they are very lopsided -- in favor of the US approach.
That doesn't mean the present US way is "correct" or that the present UN/outside-US way is "wrong". The European regulations prioritize glare control instead of seeing distance. Not because they have data showing glare causes a big danger, but because that's the philosophy their regulations are based on. But people are dying because of it; they're recognizing that and beginning to look at how to fix it with glare-free high beams/adaptive driving beams, etc. The question has to be considered in the whole-system context. Traffic lights, vehicle stop lights and turn signals, and many other roadway lights are capped at much lower intensities outside North America, which may allow a greater amount of dark-adaptation, which may countervail (to some degree) the inadequate seeing distance provided by UN headlamps aimed to UN specifications. In this chain of "mays", though, the only certain thing is the "inadequate seeing distance" part -- that's just a matter of simple mathematics, factoring in vehicle speed, headlamp height and aim angle, and driver reaction time.
This is another matter of philosophy. It's worth discussing, but I don't think it's not at the top of the list of things that need fixing in the US vehicle lighting regulations.
It's not a dilemma -- not yet. We don't have the data to be able to answer what this question really means (i.e., "How can we minimize car/bicycle collisions via bicycle conspicuity?") with other than guesses, opinions and assumptions.
Yes, and those problems probably cannot effectively be addressed with lights, reflectors, or other such devices.
That takes us well off any kind of fact/science basis for the discussion, and well into the realm of philosophy based on anecdotes. My philosophy on the matter (which is probably just as useless as yours for making traffic safety decisions, though I am tempted to call mine more thoughtful than yours) is that brinksmanship games of nefarious-activity one-downsmanship cannot lead anywhere good. Let's say you did something nefarious and tricked that disrespectful driver into momentarily not realizing you're a bicyclist...yeah, so? He's still got two tons and 300 horsepower at his finger- and toetips, and U-turns are easy. Do you want to be right, or do you want to be alive?
I'm sure you have anecdotes that reinforce and re-reinforce and re-re-reinforce your belief that confusing drivers "works well". I'm fairly confident the data, if it existed, would refute your belief. In the meantime, pending the availability of that data, I think the precautionary principle would weigh against your belief.
It may not always be practical, but one of the best ways to reduce your likelihood of being hit by a car is to minimize your risk exposure by minimizing your use of car thoroughfares. This is a close-the-drain solution (the best answer to "Water is running into the bathtub at 3 gal/min and out the drain at 1.9 gal/min, how long will it take to fill the tub?").
Having done some closer looking at the SPDs for white LEDs popular in auto headlamps, I'm amending my opinion somewhat to back away from that word I used ("plenty"). The red output is probably adequate to discern a red reflector from a reflector of another color, which is the focus of the research. But The R9 values of most of these LEDs are pretty feeble, so red won't be very vivid and as a result there will likely be a relative loss in conspicuity.
This means I'm also doubling-down on my lamentation that the high-CRI, high-R9 LEDs are being completely ignored for headlamps because they're not blue enough to satisfy the marketers. (facepalm)
As much as I like good color rendering and more consideration of the left part of the mesopic sensitivity curve, was that "blue look issue" the actual reason? Or is it more about the design and cost challenges when trying to compensate headlamp designs for the nicer LEDs' lower optical, efficacy and thermal performance?
swhs
Newly Enlightened
There is no good physical reason to use cool white LEDs at all. Therefore the only possibility is that using them is caused by marketeers and I presume by people wanting them as they 'appear to be brighter'. And thus you get a loop until someone decides to break out and tell people the facts instead of fairy tales. People who will compare lamps will then see this is correct, but most people of course are easily influenced... But it doesn't change that colour rendition with cool white sucks.
High CRI is irrelevant as what you need is simply a good colour rendition of the colours you need to be able to see well. Neutral white at normal CRI does that well. A bit lower output is not much esp. for cars where you have plenty of power, so I don't understand why cool white is used there in LED headlamps.
Better than neutral white might be needed to be able to properly read number plates, but this depends on the colour of such plates.
As to research being needed, I think Scheinwerfermann's attitude is incredibly disrepectful towards what people experience, and saying scientific research is needed is bogus. I or anyone who can objectively analyse can do that without being given a research grant, which, produces stuff that I consider mostly garbage in this and related fields, as being usually of the kind that when I read it, I already knew it from analysing "what if" scenarios, or even is the kind of stuff that makes me think "No poop Sherlock!", but then, it's mostly a psychological field and research therein is often not worth much.
I have experiences that perhaps he considers 'anecdotal' but they are not. What matters is whether a person knows himself, how he thinks, knows and not distorts what the situation was before etc. I grant that many people are not so good at analysing, but just saying "your evidence is anecdotal" about anyone's statements is just BS unless that person has shown that his evidence is not really evidence.
Further:
Position of lamps is governed by what you need to achieve so again this is not an issue, colour is set by laws so is not an issue, shape could be of use if not car manufacturers are doing all kinds of weird shapes so conspicuity is not given with any shape you make unless the shapes of those in cars are mandated to be simple (squarish). Flashing certainly gives conspicuity but attracts attention too much, loses distance estimation which is already hard.
2 taillamps above each other could work for bicyles, as being not-a-car but tests would be needed.
The big problem is not the reach of car headlamps (which in StVZO is max. 70m btw, which is not enough at speed, but if you ever ride on non perfectly flat roads, then you will understand that more than about that range is not feasible, it will blind oncoming traffic too often).
But reach is not that important for accidents: Accidents seem to be mostly caused by 'crossing the streams', which is bad not just when ghost busting (crossing the streams = car going to the right when a cyclist or moped rider or even motorcycle rider, or even another car, to his left wants to go straight on). Here you get the problem of incorrect speed estimating and esp. cutting cyclists off.
For pedestrians, retro-reflective angle-bracelets when walking on a road where cars drive would be useful... This is what I do when running at night.
Taillamp brightness/conspicuity required depends on attitude of drivers, not on what they really see. After all, we've used fairly pathetic 0.3W bulb taillamps in NL, which along with 0.6W were used a lot, give not much light but motorists see them anyway. I'm not aware of any excuses used in court cases "he used a crappy bulb taillamp", and of course not, as they were and still are standard and there were never excessive numbers of accidents because of cyclists supposedly not being visible.
Motorists not 'seeing' ambulances and police cars shows what's going on when it's going wrong: Not paying attention or making wrong estimates. And this is exactly the problem in 'crossing the streams'.
Last edited:
Recalling max. mA, CW vs. NW, in the XR-E data sheet, there is at least one, and it's related to the phosphors. And remember, in our application, we want maximum source luminance, no "let's double size, drive it less hot." Have a look at two more recent Lumileds data sheets: DS64 shows a measurable difference in temperature dependence between "lumen bin equal" cool-white and neutral-white. DS58 prefers to just show a graph of the cool ones, stating, "H/C factor is color dependent; 5000K to 6300K shown." Would this not cause a CW preference when planning temperature and flux controls for automotive designs that need to stay within certain intensity distribution, within color boundaries, within weight and lamp volume constraints?
When you combine knowledge about light sources, about street and off-road objects' spectral properties, and about human vision, you can, to some degree, qualify what some call "anecdotes." However, I can't recollect having seen the following data "scientifically" gathered AND published as pertaining to LED bicycle lighting:
obstacle detection performance; performance estimating street surface properties; level of rider comfort/confidence
This little LED test grid, for one LED family alone,
LED type
_______CCT 4000 5000 6500K
CRI
65/70
80/85
would take 6 identical, output-matched lamps for a set (... if you don't want to field-swap emitters and readjust current.)
Sure you don't want a CPF grant?
Last edited:
Which version of DS64, I have three. [and all are pretty old]
Just use a high CRI 4000-4500K neutral, and call it a day. The efficiency is only going to go up over the next decade, so pick the best right now. The efficiency is already far better than incandescent/halogen.
Just use a high CRI 4000-4500K neutral, and call it a day. The efficiency is only going to go up over the next decade, so pick the best right now. The efficiency is already far better than incandescent/halogen.
markr6
Flashaholic
- Joined
- Jul 16, 2012
- Messages
- 9,258
My god what happened to this thread?
You mean, after this?
markr6
Flashaholic
- Joined
- Jul 16, 2012
- Messages
- 9,258
It's definitely the blue look driving this. The automotive tier-1s are all excited about how they can get a given (subjective) customer reaction rating with less light, by making the light bluer. There exist LEDs with high R9 that produce more than enough light to meet the UN requirement for headlamp LEDs to emit at least 1000 lumens, and most of the cost differential goes away as a function of volume.
Not necessarily. There exist cool white light sources with excellent color rendering.
"Neutral white", "high CRI", and "normal CRI" are not quantitative, they're relative and thus not very useful terms.
How are you defining "better than neutral white"?
Er...no. The human visual system is notoriously unreliable. We humans are no good at accurately assessing how well or poorly we can see; objective measurement shows our subjective impressions are virtually always far skewed from the reality. "Science" means knowledge, and that is exactly what's needed. Not what you or I or anyone else think we understand of what we think we're experiencing. This is why lighting technical standards are defined in objective terms, not subjective ones.
Objective analysis is a type of that scientific research you were just dismissing as unnecessary. Moreover, objective analysis requires time, tools, and resources. Those things all cost money, whether from a research grant or another source. I suspect that phrase "objectively analyse" does not mean what you might think it means. What do you think it means?
It's difficult to make sense of this. Please clarify.
When you say something like this, it makes you look like you don't know much about the field of human factors. There are psychological aspects, but they aren't dominant or overriding. Even if they were, there's a great deal of psychological research that's robust, repeatable, and reliable.
Oh, dear. You don't appear to know what "anecdotal" means, either. Your experiences (and mine, and everyone else's) are, by definition, anecdotes. They are not knowledge, and they are not science.
You don't appear to have a clear understanding or firm grasp of how the human mind works. It does not work the way we might wish it did, as a video recorder and analysis computer. For that reason, science and facts trump opinions and guesses, every time, no matter how sure the guesser or opiner is that he's being all objective and analytical and self-knowing and stuff.
Even the best analysts among us are still human.
Anecdotes aren't evidence, which is why nobody's saying "your evidence is anecdotal". That appears to come from your head, as the result of your "analysis" of this discussion; thank you for vividly illustrating and confirming exactly what you set out to refute.
Position, lit area, and colour are all regulated, but regulations are not the eternal unchanging word of God. They are updated and amended from time to time in accordance with our ever-improving state of knowledge, based on research to understand the facts and build the science.
Outside the North American market, there is an example of shape being used to encode a message in the vehicle lighting system: UN Regulation 48 requires that trailers be equipped with triangular red rear retroreflectors. No other shape is permitted on trailers, and no other vehicles may have triangles.
Er...what? You appear to be saying that conspicuity is a function of lamp shape, and that "simple (squarish)" shapes give better conspicuity than other shapes. Where is your evidentiary support for these claims?
And where is your evidentiary support for this claim?
Tests...? You mean like...scientific research? Or some other kind of tests? (Also: there have been plenty of cars over the years with dual over/under taillamp designs)
StVZO does not specify a maximum headlamp beam reach.
This is definitely not necessarily true -- and there's good quality research (dang it, there's that pesky word again!) to prove it.
It's not clear whether you're talking about car headlamps or bicycle headlamps or motorcycle headlamps or what with this comment, but for cars and motorcycles at least, beam reach (seeing distance) is actually one of the aspects of headlamp performance most causally linked to collision likelihood. Again...research versus uninformed opinion, here.
Evidence to support this claim?
No "would be" about it -- retroreflectors on the arms and legs are known (again, by -- wait for it! -- scientific research) to reduce the risk of pedestrians and bicyclists. We even know what configurations tend to work best: retroreflective lines following the leg and arm lines through the knee and elbow bends.
There is no basis in reality for this random guess.
Bicyclists and motorcyclists have long been stuck with grossly inadequate lighting because of technological limitations. Now those limitations are being lifted with the advent and evolution of LEDs. That doesn't mean yesterday's pathetic lighting was adequate -- the death and injury statistics (oops, there's more science!) show they clearly were not.
This, too, is wrong. There are far too many car/bicycle and car/pedestrian crashes. A major reason for that is inadequate conspicuity.
You're guessing -- incorrectly -- that drivers simply aren't paying attention. This really isn't what's going on at all. A fundamental thing to understand is that just because an object is within an observer's visual field, doesn't mean the observer will see it. Here, go read this; you may learn something from it.
Last edited:
Beautifully succinct, completely correct, and supremely apposite. Thank you!
YBCold
Newly Enlightened
I had to ride in pitch black conditions my neighborhood had no streetlights and my headlights could glare oncoming traffic, in general drivers out here dont pay attention to cyclist so for me the brighter the better.
