The LED SAE J583 Fog Pod & Fog Light Review

The Christmas lights are not compliant unless in white/yellow/amber!

 

SS30 bar and S2 ditch lights up top. Taking a picture of them at night is tough though!

 

I went from the Pro to the Max also and found that due to the tighter beam control / better defined cutoff, you can definitely raise them up a bit. (I had to run the Pros aimed all the way down.) Your pictures also clearly show that you can come up a bit.

 

Followed the Dan Stern lighting guide as close as possible and the first thing I learned was my driveway isn’t as flat as I thought.

Alignment was set at 15 feet.

Do you guys adjust to right when any amount of light reaches the points or the hotter spots?

As you can see in the images I did the hot spot for the low beams while fogs were done at the tip of the pattern

Found using the camera on the phone makes it much easier to distinguish the outlines

 

Looks like the driver bulb is angled a tiny bit low and the passenger is very high. Not the actual housing but the way the bulb sits in the housing.

 

These are great responses, thank you. I’d like to know why BD doesn’t publish photometric plots for each product showing distance and lux. This is what lighting science means to me (and probably many others). Not intending to sound like a jerk, but frankly, your “science of lighting” page is mostly marketing. Please give us the science.

 

Installed Diode Dynamics SS3 Fogs. Max version in Selective Yellow.

Top quality kit and super easy install.

Performance is incredible. Wanted them for slow speed driving in fog/snow/rain.

Spent <$20 for H9s in the low beams, but $400+ (Xmas sale) on the fogs. Go figure.

Loving both and like that snow should not build up on either. Thanks Crash!

 

Will my best to explain our take on this. We are working on ways that we can get around or better answer some of the hurdles below, but it has been a long process of comparing loads of different types of data and readings.

We feel these plots are misrepresentative for a few reasons. A big factor is that most of the data and standards that exist measure on a 2-dimensional plane, when in reality lighting in the field is a 3-dimensional experience. Losing that critical 3D data does not give a true representation of what these lights look like behind the wheel.


The scale tends to be too small for the distances (up to hundreds of meters) that are being measured, which allows room for manipulation. Take two wide cornering lights (BD Squadron Pro WC and Brand “X” for comparison). While the BD light produces a 10° x 42° pattern and will look good on paper, all Brand X needs to do is create a 5° x 42° beam pattern with the same lumens. On paper, this light will look amazing compared to the Squadron Pro, but provides a far less usable pattern. The same goes with a spot, the charts are so small that it is incredibly difficult to see a difference in shape from a 5° to a 8° beam angle. This could be misleading, as one is meant to be used as a dedicated distance light, and the other can be used as a driving/distance light.

Further, we consider the measured/displayed data in these photometrics plots to be incomplete. They do not take into account the overall spread and smooth dispersion of the pattern, or again, the light illuminating the vertical axis. There is a large amount of peripheral spill light that would be 100% usable in the field, but does not register in these tests. Instead the data treats this light as if it does not exist. This free/comfort/spill light makes a huge difference in the field and as such, we take care to ensure it is present in all of our LED lights.

Until there exists data and standards that tell the entire story, we will continue to focus on building lights to perform in the field, rather than to have the best on paper readings.

 

The industry standard plots used by every major lighting manufacture are misrepresentative? Yes it is a 2-dimensional plot, with the contour variable being candela which directly translates into the 3rd dimension of projection distance. The 3rd dimension is not lost at all and is accurately portrayed in the plots via the candela contour variable, as higher candela values project further.

As to BD producing a 10° x 42° pattern and Brand X producing 5° x 42° pattern, on paper Brand X would not look amazing as you claim. It would show Brand X illuminates half the area. This would be immediately noticeable if you actually publish output pattern specs, like patten spread dimensions, like all other major brands. Don't focus solely on peak candela/lux. What does make Brand X look great is when you only publish raw lumen values, not telling your consumers anything about actual lamp output pattern/performance, as is Baja's current practice.

As to your claim about photometric tests being incomplete, this is another bizarre claim counter to the industry standard. The smooth dispersion of the pattern is easily captured in a contour plot and the light spill does register on the tests. I have the raw test data proving it so. The lighting plots supplied by Intertek and other light measuring software is pretty basic at best and does a poor job of showing light spill due to the limited capacity of the plot. If you use actual commercial grade engineering software as I have below you can produce commercial grade plots showing your light spill with added contour levels if that is feature you wish to highlight in far greater detail (I have not added the additional light spill contour levels in the plots below).

It really feels like you guys are trying to reinvent the wheel here. All these lighting metrics and performance measurements have been laid out long ago and are adhered to by every major lighting corporation, both for on road and for off road racing. But while every other major lighting company produces measured data on the performance of their products, Baja seems to only produce excuses as to why it somehow shouldn't apply to them. This looks poor on the Baja brand, yet the brand seems to keep doubling down on the strategy. If you want to call yourselves the "Scientists of lighting" you should produce actual measured output facts, like every other lighting company. In all seriousness though, I want everyone to do well as I believe competition is better for the common good, I am just at a constant disbelief on how far off Baja seems to be on the target.





This delta plot shows the losses from the revised optic expressed in positive values

 

Editor’s note:

I wrote this reply before crash posted his response while working this evening but didn’t finish it. Like most of my posts it’s not particularly coherent nor overly informed, I’m learning as I go here. I am posting it only because it seems potentially interesting as a layman’s perspective on Baja’s response which stood out to me for some reason, much like the Tundras repost a few pages back.

—————-

So, @Baja Designs, what methods or process or algorithm or standards or data _do_ you use to evaluate lighting in the field? It's entirely subjective analysis of raw data and in person road testing? Or there is some weight given to performance metrics such as how fast a driver completes a lap of an off road course at night? From reading what you write it feels like something similar to tossing spaghetti on the ceiling to see if it sticks? Someone reviews a ton of data and makes a gut call? You hold a committee vote? Or have you built a massive BI analytics engine and proprietary algorithm which analyses 3d plots of photoreceptors over time?

I understand your argument that lighting is more than what can be measured inside a controlled laboratory environment but that's similar to saying consumer behavior is more than what economics can model, it's an anti-academic cop-out of sorts. Without some sort of repeatable process approaching a scientific method, how can you have confidence that you're improving upon prior work?

Say you have 3 different versions of the same light with slightly different characteristics, how would you pick one for production?

One thing crash does extensively is L + R photos so there's absolutely no ambiguity about how A compares to B. The problem I have is that my eyes are untrained and while I think I can see some subtlety I once missed, without measurements of intensity my eyes and brain sometimes can't really tell which is brighter.

I fully agree that 2d images don't tell the full story but I do think 2d can do a pretty decent job of showing the pattern and a few intensity measurements can give a rough objective measurement of output within the pattern and that's a baseline. If you claim your light is significantly better and your independent 3rd party confirms this claim then I would expect it should be possible to demonstrate that with some sort of published reference standard. At the very least if all manufacturers did this then someone like crash, with reasonably sensitive equipment, could reproduce and confirm his numbers are in the same ballpark. Peer review is the basis of scientific method, reproducing experimental results is critical to accepting a hypothesis as provable fact.

TLDR, I'm fine with you considering 2d plots incomplete and subject to manipulation (which has overtones of VW and the EPA test) but if you don't have a better method I don’t see the downside.

BTW, your responses to my previous questions were timely, comprehensive, and well worded. Thank you for taking the time to participate and share.

 

While I disagree with your assessment that lighting plots are misrepresentative, I appreciate your response.

 

Here is my 5 minute attempt to plot light output in 3 dimensions using candela as the depth variable. This is Baja's original SAE light. To be fair, the test results only cover 40 degrees left and right of center (80 degrees total), so spill beyond 40 degrees to the sides is not captured in these results.

 

Wow, I've never seen anything like that! I am not sure what the z axis represents, but intensity is inversely proportional to the square of the distance... and if you can figure out how to plot that, you should license it. Though at this point, I think you may just be re-inventing CGI or something.

 

Yeah, I was thinking that about distance as soon as I posted it. It is easy enough to create a new variable via an equation operated on an existing variable (candela) to correctly calculate the distance projection for a more accurate representation of depth.

 

Candela = 1/d^2
So then Distance = Sqrt(1/Candela)

Calculating distance based off candela and updating the Z-axis to use calculated distance. Might help to use a slight magnification factor on distance for better visual.

Edit: this still doesn't quite seem quite right for distance.

 

amazing. crash, you are a wizard. would love to see a visual representation like this in your headlight thread, but i assume the effort would be monumental given the different beam patterns

 

Once I create the first plot, subsequent plots can be mostly automated making it quite easy. The most time consuming part is just making some minor edits in the data format to get it into the application. To make the plots though requires raw measured test data from a testing facility, which I don't have for the headlights.