Mountain Bike Reviews Forum banner

1 - 20 of 29 Posts

·
Registered
Joined
·
44 Posts
Discussion Starter #1
Haven't been on here in a while. I've been riding more lately, but it's been more commuter than anything. I've been trying to take it easy on my wrist since I broke it about a year ago and because I foolishly didn't strictly adhere to the the rehab.

Anyway, lighting where I have to ride isn't very good, I often have to go through construction zones at night and also wanting to try some light trail riding at night since I've never done it. So I decided to look for a good headlight. I like the setup MTBR uses for testing lights and they're consistent, so I used their numbers.

Attached is the excel file in a zipped folder I did all calculations in. There are four lights that are not included in the calculations, but are present in the table. Three didn't include price (which I guess I could look up, lol), and one does not have a measurement for Lux. If you sort any of the columns, it will change the results in the Max, Min, and Averages below the table (has to do with how the calculation is setup to not include the headlights with missing data points).

This is a scatter plot of the lights from the 2012 Bike Lights Shootout.



The first thing I calculated was the price per unit Lux. I was interested in the spread of this ratio to see if you get more output with higher cost. The red fields just indicate no data available.

Lower is better:
Max:........$15.45/Lux (Light and Motion vis360)
3rd Min:....$2.75/Lux (Niterider Pro 3000)
2nd Min:...$2.47/Lux (Dinotte XML-3)
Min:.........$1.81/Lux (Magicshine MJ-872)
Avg:.........$4.79/Lux


(in the following two tables, green means max and brown means min)

I then calculated the distance from each data point to the line of linear regression (not perpendicular, so the vertical difference) and found the max and min of the set. The max being the greatest positive distance from the linear regression line and same for the min. This means that the greater the distance for a given price yields a better output to price ratio.

Positive distance is better, negative distance is worse:
Max:........105.636 (Niterider Pro 3000)
2nd Max:...51.5675 (Magicshine MJ-872)
Min:.........-37.364 (Light and Motion Seca 1400)
Avg:............ .0183


The next calculation was a ratio of the distance mentioned above to price. Positive values indicate higher than average output vs price and negative values are a low output vs price.

Higher is better:
Max:........ .2787 (Magicshine MJ-872)
2nd Max:.. .1558 (Dinotte XML-3)
3rd Max:... .1509 (Niterider Pro 3000)
Min:......... -.4395 (Niterider Mako 1)
Avg:......... -.0212


So we see that the top two headlights based on these stats is the:
Niterider Pro 3000
Magicshine MJ-872

If you take a look at the trail High End and backyard High End shots, you can easily see how they demonstrate exactly what the numbers say. The Niterider Pro 3000 has the best light output out of the bunch. However, the Magicshine MJ-872 has the best bang per buck. Not to mention, the price difference between the two is a whopping $515.

Trail:

Magicshine MJ-872


Niterider Pro 3000


Backyard:

Magicshine MJ-872


Niterider Pro 3000



So out of all of these, my conclusion is the Magicshine MJ-872 is the one to go with, without breaking the bank. I know I shouldn't limit my headlight options to just the ones MTBR tested, but comparing test results from another site will not be an apples to apples comparison. Hope someone found this helpful/beneficial.
 

Attachments

·
Mtbr Founder
Joined
·
34,659 Posts
Nice man. I will send you my spreadsheets at the end of this week.

What I've come to realize is that my Measured Mtbr Lux setup is very good. We have some very good actual lumens data from the Lezyne lab and they corroborate my test lux data.

Measuring actual lumens is very difficult when the light is mounted outside the sphere because all the light has to go in the sphere and no light can leak out. With so many different light head shapes and sizes, this is difficult.

Putting the light inside the sphere is difficult too. The light head is usually black and will absorb some of the light in the sphere and corrupt the data. Ventilation is a problem too inside a sphere. One of the solutions is to use a big sphere (10 feet diameter) so the surface area of the light head is tiny in comparison. But these big spheres cost $200k plus.

Anyway, long story short, the Mtbr Lux readings you're using are good.

And the word of the day is: XML :)

fc
 

·
Registered
Joined
·
44 Posts
Discussion Starter #3
Cool, I look forward to it! :thumbsup:
 

·
Registered
Joined
·
373 Posts
watts wanted ,... wh
since all lights are powered by a battery,
it be pertinent to know, the actual power used.
I use a power meter for that.
initially for first 5 sec, gets you peak.
and after 5-7 min, when it's hot,

on the battery a discharge curve measurement,
can now ballpark the runtime.
and with the light watts,
running through the pct.cree.com calculator,
gets you ballpark number on lumens, with compensation on optic loss, and drivers.

so power meter, watts-up, voltage * current = Wh ,
battery wh / light wh , ~ miles per gallon ,...

----
why so important, because , once you run out of juice, you got Zero (0) lumens !

cheers, Rob
 

·
Registered
Joined
·
2,374 Posts
wow, that's some detailed analysis! I'm guessing that you have more than a fair head for mathmatics :)

I think these price/ performance analyses are great, but price and performance are not the only factors in buying a light (sometimes they're not even the most important). Harder to quantify factors such as beam shape, reliability, customer service back up and warranty are also very important, let alone the weight and runtime that Rob alluded to. There are already enough people reporting problems with both of the top lights to be wary about purchasing them.

For me, the Dinotte, Baja and Designshine lights a little further down the scale would be better options. You'd give up a little in terms out output per dollar but gain a huge amount of peace of mind in exchange.
 

·
Church of the Wheel
Joined
·
749 Posts
wow, that's some detailed analysis! I'm guessing that you have more than a fair head for mathmatics :)

I think these price/ performance analyses are great, but price and performance are not the only factors in buying a light (sometimes they're not even the most important). Harder to quantify factors such as beam shape, reliability, customer service back up and warranty are also very important, let alone the weight and runtime that Rob alluded to. There are already enough people reporting problems with both of the top lights to be wary about purchasing them.

For me, the Dinotte, Baja and Designshine lights a little further down the scale would be better options. You'd give up a little in terms out output per dollar but gain a huge amount of peace of mind in exchange.
+1

This is very useful data, and thank you for organizing it so well. Prospective purchasers would do well to remember though, that this analysis takes into consideration only two variables (albeit the easiest two to quantify). If we bought cars this way (cheapest price for highest performance), we would all be driving Ariel Atoms :p

For me, a quality beam pattern is more important than max output, but that's just my druthers, and a difficulty thing to quantify.
 

·
Registered
Joined
·
463 Posts
And what I notice is that lights are still quite pricey as a whole. I'm in the market for a light and need to stay around the 200 dollar mark (or less!) I notice that for a legitimate MTB light, there's only a couple of options. (Nightrider MiNewt, MagicShines, Nitelights). This also makes you wonder.
Is a $100 MagicShine or Nitelight worth the gamble in terms of reliability and service? You could buy two of them for $200.
My buddy has had a lot of problems with his MagicShine and another had his suddenly quit working recently, and what's the price of being left in the dark during a 24 hour race or an epic night ride??

I've used the MiNewt 600. Its a good little light, and I would hope it brings Nightrider's durability and service with it. But also found it just on the end of acceptable for brightness when night riding at speed. The run time is also a bit low.

thoughts? Advice?
 

·
Village Dirtbag
Joined
·
2,284 Posts
Something else to consider is battery life. A super bright light might cut costs by not adding a bigger battery, leaving you with short run time. Perhaps lux minutes per dollar would be the most useful figure. Weight might also be a consideration.

Magic shine really blows things away for value, but some have had overheating problems on the 872. The older smaller 808 tends not to overheat though....although who knows...maybe it does and just doesn't have a protection mode. I doubt that's the case though because it has a pretty big heat sink for one LED.

@gcavy1: you think lights are expensive now...you should see how much it cost to get this kind of brightness 6 or 7 years ago. I recommend always riding with two lights, so select a price point where you can afford two. This gives you a backup, and having one on the bar and one on the helmet give you better depth perception and ability to see around corners.
 

·
Registered
Joined
·
2,195 Posts
Not sure that trail shot of the 872 does it justice. Almost looks like an LED isn't working on the right side. The backyard shot is much better and looks like my MS 856 (the 872 clone in different housing)

The Nightrider 3000...it's just nutty bright.


Francois....Did you get the Gemini XERA yet? Look forward to your thoughts and test results on this tiny light.
 

·
#1 Latex Salesman
Joined
·
1,869 Posts
And what I notice is that lights are still quite pricey as a whole. I'm in the market for a light and need to stay around the 200 dollar mark (or less!) I notice that for a legitimate MTB light, there's only a couple of options. (Nightrider MiNewt, MagicShines, Nitelights). This also makes you wonder.
Is a $100 MagicShine or Nitelight worth the gamble in terms of reliability and service? You could buy two of them for $200.
My buddy has had a lot of problems with his MagicShine and another had his suddenly quit working recently, and what's the price of being left in the dark during a 24 hour race or an epic night ride??

I've used the MiNewt 600. Its a good little light, and I would hope it brings Nightrider's durability and service with it. But also found it just on the end of acceptable for brightness when night riding at speed. The run time is also a bit low.

thoughts? Advice?
I approached my decision using mostly the same thinking and landed on the DiNotte XML-3. $259. Check it out.
 

·
#1 Latex Salesman
Joined
·
1,869 Posts
Love the analysis piranha. Rather than having every light head to head (let's face it, no one is deciding between the niterider 3000 and the Magicshine), perhaps you could group the lights by claimed lumen, in 200 lumen increments and show the top two in each category.
 

·
Mtbr Founder
Joined
·
34,659 Posts
Not sure that trail shot of the 872 does it justice. Almost looks like an LED isn't working on the right side. The backyard shot is much better and looks like my MS 856 (the 872 clone in different housing)

The Nightrider 3000...it's just nutty bright.


Francois....Did you get the Gemini XERA yet? Look forward to your thoughts and test results on this tiny light.
No Gemini's yet. I haven't heard from them.

I might reshooot the 872. That trail shot is actually deeper on the right side. The foliage on the left side is closer than the right. So lights that are very wide and don't have a lot of throw will look much darker on the right.

fc
 

·
Registered
Joined
·
44 Posts
Discussion Starter #16
Thanks for the replies. Trust me, I do realize this is a limited analysis and these two factors are the starting point for me. I then look into the products themselves and start comparing things with less or no quantitative data; such as charging method, battery pack or stand alone, over heating issues, etc. Beam pattern is also very important, but is difficult to incorporate due to some not providing specs, different methods and thresholds for measurements, etc. I believe you have to do some sort of analysis in order to have a starting point and a kind of guide for what product to look at next.

If someone were to gather any additional specs (such as runtime and weight) I'd be willing to do another analysis. Unless fc's spreadsheets will include that data as well.

As far as grouping the lights into various performance ranges based on Measured Lux, that's why I provided the spreadsheet. :D Not trying to be a smart ass; I'm just not personally interested in that breakdown. Everyone has their set of values and objectives and I think the only way to do analysis and accommodate almost everyone would be to have a database (just in Excel is fine) of all lights and all specs (or as many as possible) and vary the 'query' to get what you're looking for. This is essentially a data mining problem, but I digress.

In regard to the cygolites; unfortunately if MTBR didn't include them, they won't be present in this analysis. I'm sure that it would not be cost effective for them to include every bike light, test every one, and then do analysis on them. Not to mention they change just the next year, lol.

And for anyone wondering, I absolutely love analysis and statistics. I am in my last year for my BS in Mathematics. But believe it or not, I have not taken any statistics or applied math courses; they're just not required and my graduating plan didn't allow time for them. I'm considering some graduate degree dealing with applied statistical methods which entails many professions, but I don't have any specific career goals in mind.
 

·
Registered
Joined
·
43 Posts
Great analysis.

The tough thing to factor in is quality.

I decided to avoid the Magicshine issues and went with a Dinotte XML-3. The quality is better and I expect a longer life.

I know some might think "what the heck" and buy 2 or 3 $100 dollar lights and just use a new one when one breaks, but I prefer one quality light that lasts for years.

Heat management is also an issue. Lots of those cheap lights have lousy heat sink designs and will not last and/or fry their electronics. Again, quality in design is key.
 

·
Registered
Joined
·
587 Posts
So I have a light that puts out well over 2k lumens, cost less than $150 and has been trail tested nearly 100 times without a fail. Beam pattern and tint are perfect. Where would it be on the chart?

Of course the light is a DIY. Previous owner of Lupine, Dinotte, light & motion, and several others. If you want to get the maximum performance at the cheapest price - build your own. All of the lights out there with the exception of the Magicshine are too expensive. However, the Magicshine's quality is below what a regular trail night rider would consider acceptable. I've seen them on the trail and more than once have had to give up my backup light. I understand not everyone can go the DIY route but it isn't as hard as one would think.
 
1 - 20 of 29 Posts
Top