1. Welcome to Tacoma World!

    You are currently viewing as a guest! To get full-access, you need to register for a FREE account.

    As a registered member, you’ll be able to:
    • Participate in all Tacoma discussion topics
    • Communicate privately with other Tacoma owners from around the world
    • Post your own photos in our Members Gallery
    • Access all special features of the site

The 921 LED Reverse Light Bulb Study

Discussion in 'Lighting' started by crashnburn80, May 6, 2017.

  1. May 8, 2017 at 10:14 PM
    #21
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!
  2. May 8, 2017 at 10:17 PM
    #22
    Somegreatwood

    Somegreatwood Member

    Joined:
    Mar 8, 2017
    Member:
    #212693
    Messages:
    5
    Gender:
    Male
    Ontario, Canada
    Vehicle:
    2017 Tacoma TRD Offroad
    Amazing write-up. Thanks for all your work, seriously helpful!
     
    DVexile and crashnburn80[OP] like this.
  3. May 8, 2017 at 11:15 PM
    #23
    DVexile

    DVexile Exiled to the East

    Joined:
    Dec 17, 2014
    Member:
    #144469
    Messages:
    2,656
    Gender:
    Male
    First Name:
    Ken
    Vehicle:
    2015 DCSB V6 TRD OR 4X4
    I'd agree to a large extent except there is also the problem of spillover. Which is quite clear in both my side by side photo and yours. So doing a mixture of LEDs in the same shot is already misleading as it makes a brighter LED with a fuller field actually appear dimmer relative to stock than a dimmer LED with a tighter field simply because the tighter field LED doesn't spill into the stock side. But then purchasing twice as many LEDs and swapping twice as many bulbs and back to how to present the comparison best in separate images.

    Without making the testing methodology and presentation onerous there are only different imperfect compromises. I think you've chosen a very good compromise for your test. I can't think of a better one really without actually swapping both bulbs. Given you have to test multiple LEDs the work load is already really high (not to mention doubling the cost of test items). Having done lots and lots of image comparisons myself over the years I understand how time consuming it gets. You've put an enormous amount of effort, expense and time into this (and all the other excellent threads). Hell just my stupid shots turned a 10 minute bulb swap into a 45 minute exercise.

    As I pointed out my shots are not a controlled scene such as yours - just taken during installation for reference since someone else was interested in the bulbs too. They don't have a uniform field and thus are best presented differently than your shots which do. On the other hand I was able to avoid the spillover problem since I had two bulbs and if you look at the three shots can see clearly how misleading the side by side shot is due to the spillover - it makes the stock look brighter than it really is even to the far left. You can see the same effect in some of your images. A careful viewer can use a switching view between the full stock image you provided and the LED under test shot to largely mitigate that but I doubt many will. Note the spillover in many of your shots is actually of a higher magnitude than the tone quantization errors in the GIF presentation you criticized.

    For what it's worth if you are going to make claims about "engineering muster" I'd encourage you to take some time to understand human vision and image processing more thuroughly. You aren't the only engineer with experience in imaging data here and it would be wise to consider that even using a DLSR and fixing exposure that there are already multiple processes in play that are "picking and choosing what parts of [the] test are accurate" without your conscious selection. Ones of magnitude larger than 8 bit indexed color quantization. Whether you are aware of said effects or not they are occurring and affecting the results and their presentation.

    I apologize to be nitpicking - your tests here are hugely valuable and very well executed. I wish I had the time to contribute anything so thorough as you have. I have a lot less time than in decades past. And I really can't think of a better presentation than what you did without making things even more burdensome. Truly a tremendous contribution to the whole TW community you have provided - multiple times now.

    And if one of my expensive LEDs proves to be less durable than hoped I'll definitely be picking up the HIRs - excellent find and research!
     
  4. May 9, 2017 at 6:51 PM
    #24
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!
    Many excellent points.

    My photos are definitely not perfect and spillover is certainly an issue with the way I have chosen to document the project. While not perfect, it seems like the most controlled method with the lowest overhead for taking photos to me, especially when comparing 7 different bulbs. As you know taking accurate lighting photos is very challenging. I am aware that human vision and image processing is not nearly as reliable as people tend to believe, people are just not good judges of what they see. Scrolling static photos of differing light output patterns is a sucky way to compare, which is why I put the highest output LEDs and HIR side by side in the end. It is also why I really like the lux meter, so that results are quantitative (as I like numbers) but that too is imperfect. I actually considered running my output through a design optimization software package, but it is really intended for simulator output and I bumped the tripod once or twice so the pictures are not 100% identically perfect, meaning the software package would identify massive differences making the results worthless. And while my environment is controlled, it is not as ideal as I'd like it to be. I'd ideally like it on a dark road, but there are street lights on all the streets around here, and I need somewhere that I can spend hours swapping bulbs and taking lux readings and taking photos, so my side yard is a compromise. All the ways to compare are a compromise in some way.

    Please nitpick away. I welcome criticism, you are not going to hurt my feelings. Educated criticism to further understanding is great, I will use it to do better on the next thread and others can learn from it too. I know I certainly dish out enough nitpicking and can also be terse at times. I purchased an higher end LED lux meter because it was pointed out to me that some of my results were not as accurate as they seemed since I was using a standard lux meter, a criticism I received. I learned something and adjusted. While some people view me as a subject mater expert, my background is not in lighting. I do not claim to know everything.

    And I appreciate the effort recognition as I have put a ridiculous amount of time into this thread. And as a working parent, that time really only comes at the expense of sleep. Funny story. I did all the lux readings and photos doing the bulb swaps. And then read the lux meter manual and it mentions the light source must be manually specified, which I did not do as I just used the default. So I had to do them all again.
     
    GHOST SHIP and DVexile[QUOTED] like this.
  5. May 9, 2017 at 10:06 PM
    #25
    RysiuM

    RysiuM Well-Known Member

    Joined:
    Oct 18, 2015
    Member:
    #167004
    Messages:
    2,692
    Gender:
    Male
    First Name:
    Rysiu
    Was Golden State, now Poland EU
    Vehicle:
    1995 4x4 LX Ext Cab, I4 2.7, MT, 335K miles
    DD Deck+backup camera, LED DRL, All LED except H4 Hella
    I have a feeling that the conclusion (suggestion) is that the best is HIR bulb. I would argue that all depends on what criteria you use. For cost+light output+power efficiency I would vote for Woolf bulbs. Output and pattern equvalent to stock for 1/5 power. HIR has the most light output but for 1.5 times more power than stock. Other bulbs are either insane expensive or not good in lighting.

    I would argue that for the power of hir you can install quality flush bumper LED light that would outperform hir many times. For stock light output Woolf bulbs would do quite well without paying arm and legs.

    The benefit of LED lights was always the efficiency not power. LED bulb must run cold what prevents them from using the same power in the same size as incandescent bulb. But LED bulb size of incandescent bulb can deliver the same amount of light for the fraction of power taken.

    Finally Lux meter is not a good tool to measure the amount of light. It is good for measuring the brightness of the light source. So it is a tool to measure the visibility of light source important for bulbs like DRL, blinkers, stop or markers. For headlights or backup lights the lumen and pattern is the factor.

    My 5 cents :)
     
  6. May 9, 2017 at 10:28 PM
    #26
    Hoonatic

    Hoonatic Well-Known Member

    Joined:
    Jan 21, 2017
    Member:
    #208222
    Messages:
    114
    Gender:
    Male
    PDX
    Vehicle:
    2017 TRD OR DCLB
    Thanks so much for all of the work! This was exactly what I've been looking for and is super well presented and thorough.
     
    crashnburn80[OP] likes this.
  7. May 10, 2017 at 6:41 AM
    #27
    DVexile

    DVexile Exiled to the East

    Joined:
    Dec 17, 2014
    Member:
    #144469
    Messages:
    2,656
    Gender:
    Male
    First Name:
    Ken
    Vehicle:
    2015 DCSB V6 TRD OR 4X4
    I honestly think your presentation is probably the most efficient way to show the comparisons - both in the time it takes you to do the testing and in how people are most likely to view the results. You've got excellent text to accompany the results and I think that's how you can best handle the various anomalies that will always come up no matter what choices are made in the testing or presentation. If you can see the spill over is making the comparison a bit misleading just mention that in the text. If there is such a large difference in brightness that the tone curve used by the camera/software is making the visual presentation not match the real world results also worth a mention.

    The best reviews and tests always include a good write up by the reviewer/tester explaining the results. Just throwing up the images or the data without explanation or context usually leads to people drawing the wrong conclusions. That's what I like most about all your threads - thorough explanation of what is going on and leading the reader along through the results.

    This is another aspect of what makes your presentation excellent. Providing more than one metric or comparison. There isn't any single number, image or plot that will ever tell the full story for even something as deceivingly simple as "brightest tail light". Presenting the comparisons in multiple ways is both more informative but also is a great way to get a lay reader to understand the surprising complexity of some comparisons.

    When I did my shots I was living in an apartment in the LA area. It is nearly impossible to find any place remotely "dark" and I didn't feel like waiting to do the swap and shots during a camping trip! Like you I've got a young kid around and so time is very precious. The best I could do was find a nearby modestly dark patch where I wouldn't be in traffic or otherwise harassed. I was acutely aware that each minute I spent fussing with the camera was one minute less sleep I was going to get that night.

    Very few people understand how much time even a "simple" comparison takes until they try to do it themselves! Again, kudos for all the effort in producing these really excellent lighting threads.

    OMFG - this always happens to me in some form. I pretty much expect it these days and plan on having to do any comparison test (for me it is usually photographic optics) more than once no matter how much careful planning I do. For large comparisons I've now tried a few times just doing a small test (say three lenses) as a trial first so I can screw that up once or twice without it taking as much time as redoing a ten lens test multiple times.

    Again, thanks for the excellent reviews/tests you have been providing!
     
  8. May 10, 2017 at 9:16 PM
    #28
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!
    I purposefully added no such conclusion.

    HIRs are indisputably the brightest of the bunch in terms of output. However it is not without compromise, they do not lock into the 921 connector like the OEM and LED bulbs. They have a yellowish halogen color, which some do not like. They consume more power and they are also not available currently in the USA.

    The point of the thread was mearly to show how differing designs of directional light sources affect the output and to help people understand and make informed decisions on their modifications. For someone looking for lower power consumption, whiter light and near stock performance at a lower price point, I agree the Flashwolves would be an excellent choice. Personally I love power and performance, which is why I chose the function of HIR despite their lack of sexy appearance.

    LED efficiency is great and true they can deliver more while consuming less, but saving a few watts while in reverse isn't a great reason for LED in an inherently inefficient internal combustion vehicle, unless you are already doing something crazy to utilize the power from the circuit. The reverse circuit is rated at 10A, combined with the tow circuit. Going from OEM to HIR raises the current use from 1.5A to 2.5A per bulb leaving 5A in the circuit for a trailer. True you can certainly add aux lighting that would be brighter with flush bumper lights, it was something I considered. But I personally prefer my truck to look as stock as possible but perform drastically better, you will notice that theme in all my threads. If I can swap bulbs that put out nearly 3x the light for $30 and have it look stock, I'd rather do that than hack my bumper and spend $100+ on flush mount LEDs and then have to tap them into the reverse circuit and have it clearly look modified.

    The lux meter is just one tool in the tool chest. I intentionally provided several methods of measurement as to not bias to any single measurement form. Lux is just a measurement of lumens/area. Agreed that pattern is critical, LEDs can have a terrible headlight pattern due to their lack of uniformity for example but score well on a lux meter due to their directional intensity. Omnidirectional light sources will have better pattern and uniformity than directional ones, but the lux meter will only measure a small part of the beam.

    For interest of curiosity I added some lux 'napkin math' accuracy measurements near the end of the thread.
     
    DubfromGA and DVexile like this.
  9. May 10, 2017 at 10:10 PM
    #29
    RysiuM

    RysiuM Well-Known Member

    Joined:
    Oct 18, 2015
    Member:
    #167004
    Messages:
    2,692
    Gender:
    Male
    First Name:
    Rysiu
    Was Golden State, now Poland EU
    Vehicle:
    1995 4x4 LX Ext Cab, I4 2.7, MT, 335K miles
    DD Deck+backup camera, LED DRL, All LED except H4 Hella
    I'm sorry I did no wanted to offend you, and I do appreciate the effort. If I did, I apologize from the bottom of my heart.

    I guess we have different point of view and that makes me look at different way of skin this cat :). I also prefer stock (or "factory like") look, so I used 1156 LED bulb (about 5W) in my 95.5 Tacoma (my old-old truck have no option for HIR) to get the similar output to stock 20W. Because of Polish legal requirements I had to replace left backup light with red "rear fog light", now I am more in need to get some decent backup light (lavinf one backup light is legal but really sucks). I recently bought a set of quite good quality flush mounted bumper lights (House Tuning- 20W LED Lamps with measured power around 15W) for 50 dollars a set (including harness, relay switch and quality waterproof connectors). It will be some work to get them installed and connected but the light output will be hard to beat. I am going to use relay and new wiring for this 2.5A circuit.

    I may add that the backup light is not "rated" at 10A but it is fused by "10A Gauge Fuse". It simply means that the circuits (wires) are protected by 10A fuse from being destroyed by possible short. Keep in mind it does not mean that the circuit can handle constant load of 10A and especially connectors and contacts in 20+ years old truck may start overheating under such load. For the reverse light it is not that big deal as they are used for short period of time but still the fact is worth mentioning. It is not about the "electricity saving" but about "saving electrical circuit".

    This is the main reason I replaced all bulbs in my truck (minus headlights) with LED bulbs of similar light output or brightness - to save all those old switches, relays, connectors from destructive power of electricity. The last thing I want in my 22 years old truck is electrical fire.

    Regarding measuring and comparing different lamps or bulbs used for "lighting purpose" such as headlights and backup lights (the purpose of all other exterior lights in this truck is to be visible) is to show the picture of objects or surrounding lighted by these lights. This is what you really care about. And yes, you did it. Great job on that. Thank you. In my response I just mentioned, that "lux numbers" or picture of the lamp itself does not mean anything and actually can be missleading. I did similar comparison for different bulbs used in puddle lights in my 5th gen 4Runner ( ) where I did not do lux measurement (I do not have that tool) but quality SLR with manual setting. Therefore I know how much time consuming such review is, and I really do appreciate your effort. Just wanted to insert a different point of view from which other readers may benefit some.
     
    crashnburn80[OP] likes this.
  10. May 10, 2017 at 10:42 PM
    #30
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!
    Seriously you are not going to offend me. Bring health discussion and challenge what you do not agree with or think can be improved.

    Good point on the fused circuit, you are correct it is fused at 10A. Though your concerns about a 20+ year old truck do not necessarily translate to a 1-2 year old truck. If you check out my Gy6.35 thread, I left the truck in reverse for 20+ minutes and took IR temperature readings. I'm comfortable for my use case on the circuit load, but that is for each user to validate on their own. If you use the truck on a trail I think "reverse is only for a short time" mentality can get you in trouble. I've had to back out of some long spots on the trail and would hate to be sweating about my reverse circuit.

    Nice shots on the SLR, I often think I need to go back to one of those for more control of the variables in my lighting photos.
     
  11. May 10, 2017 at 11:24 PM
    #31
    RysiuM

    RysiuM Well-Known Member

    Joined:
    Oct 18, 2015
    Member:
    #167004
    Messages:
    2,692
    Gender:
    Male
    First Name:
    Rysiu
    Was Golden State, now Poland EU
    Vehicle:
    1995 4x4 LX Ext Cab, I4 2.7, MT, 335K miles
    DD Deck+backup camera, LED DRL, All LED except H4 Hella
    Oops, I did not notice that the link to this tread jump out of 1st gen forum I was reading.:facepalm:

    And you are right about long reverse run, I just did not think about it as I don't do trails per se :burnrubber:.

    And finally about photos, SLR is is designed to have more controlled photos, not only because of the fixed exposure but also fixed white balancing. Removing all these "point and shoot" automations makes possible taking a good comparison photos. And to get better comparison often it is better to take underexposed pictures, so there are no overexposed pixels or areas for which the sensor chip might try to be "smart" and compensate for it.
     
  12. May 10, 2017 at 11:33 PM
    #32
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!
    I've done 3 years of photography in school, so I am reasonably well versed but that was back in that day before DSLRs where we actually spent time in a darkroom. I did add a section to the original post on the effect of automatic aperture settings for separate lighting photos as people do not seem to understand what a profound effect that has on the presentation.
     
  13. May 11, 2017 at 12:40 PM
    #33
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!
    Realized I missed responding to this. I have IR heat gun readings in my gy6.35 thread, including running for extended operating times.

    I've secured them by bending the prongs, I think another member used some high temp RTV, it is in the comments.

    See this thread:
    https://www.tacomaworld.com/threads/gy6-35-hir-921-reverse-light-upgrade-vs-high-power-leds.474996/
     
    mabepossibly[QUOTED] likes this.
  14. May 13, 2017 at 7:29 PM
    #34
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!

    Ordered these to add to the study. We'll see how they compare.
     
    DVexile[QUOTED] and Norton like this.
  15. May 14, 2017 at 5:36 AM
    #35
    Icepuck72

    Icepuck72 Well-Known Member

    Joined:
    Jul 9, 2011
    Member:
    #59563
    Messages:
    1,809
    Gender:
    Male
    First Name:
    Mark
    Richmond, VA
    Vehicle:
    2006 DC TRD
    OME 885's, Dakars, TRD Trail Edition Wheels, 285x75x16 Goodyear Duratracs, Magnaflow 12256 muffler, Fog Light mod, 2 Baby seats
  16. May 14, 2017 at 8:32 AM
    #36
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!
  17. May 14, 2017 at 9:53 AM
    #37
    Muddinfun

    Muddinfun Well-Known Member

    Joined:
    Apr 17, 2017
    Member:
    #216500
    Messages:
    7,491
    Those do look the same as the DeAuto bulbs. I don't know if they're exactly the same or a copy. That being said, after I read this review, I bought the DeAuto bulbs. When they arrived, I ripped open the box and pulled out the transparent antistatic bag with the bulbs in it. What? Those don't look very impressive. I was expecting something bigger. Before I tore open the package, I came back here to this thread to look at pics. I figured they sent me the wrong bulbs. Nope, those look just like the pics in the OP. Skeptical, I put 1 of them in the truck and compared it to the Sylvania LED from Orielys. Huge difference, even though it was daytime. I put in the other one and anxiously awaited night time. As soon as it got dark, I blocked the wheels, jammed the ebrake, fired it up, put it in reverse and got out to look. Holy crap! Lots of light! And light everywhere! I have never owned a vehicle that put that much light out to the sides. They shine light perpendicular to the truck as well as behind the truck. Very impressive!
     
    crashnburn80[OP] likes this.
  18. May 14, 2017 at 11:33 AM
    #38
    RysiuM

    RysiuM Well-Known Member

    Joined:
    Oct 18, 2015
    Member:
    #167004
    Messages:
    2,692
    Gender:
    Male
    First Name:
    Rysiu
    Was Golden State, now Poland EU
    Vehicle:
    1995 4x4 LX Ext Cab, I4 2.7, MT, 335K miles
    DD Deck+backup camera, LED DRL, All LED except H4 Hella
    99% sellers publish "lumen rating" value, not the actual lumen output. "Lumen rating" or "power rating" (also used in product description) is a simple calculated value from datasheet values of led chips used:

    Lumen rating = absolute maximum lumen output of the led chip times the number of led chips in a bulb.
    Power rating = absolute maximum power of the led chip times the number of led chips in a bulb.

    These numbers are used for marketing purposes and mainly to attract buyers. The actual performance of LED bulb is at least two or more times lower than these numbers as it is impossible to sustain that power in less than ideal laboratory conditions.

    But it is possible to estimate the actual lumen output of the bulb using diagrams published in led chip datasheet, understanding bulb's internal diagram (driver) and total power consumption. Knowing the power lost in the driver (buck or current limiting type), voltage drop on rectifier (diode or bridge) and power consumption by "CANBUS resistor" (if used) you can calculate the total power used by LED chips. Dividing that by the number of chips in the bulb you willget the power used by individual LED chip. Using datasheet diagram you can then read the lumen output at given power. These will be true numbers.

    It requires some research and calculation but then the buyer will know exactly what he/she paid for ;-). All these reviews on Amazon like "OMG so brighter than my stock lamp" just show how ignorant most buyers are. If possible I'm always asking the question about the real numbers such as easy to measure current taken by the bulb.

    I really enjoyed reading the comparison made by OP including numbers he published as this is a good base for people to make conscious decision, not marketing BS you read all over the place.
     
  19. May 14, 2017 at 3:50 PM
    #39
    crashnburn80

    crashnburn80 [OP] Vehicle Design Engineer

    Joined:
    Jun 6, 2015
    Member:
    #156893
    Messages:
    14,562
    Gender:
    Male
    Kirkland, WA
    Vehicle:
    2003 DCSB TRD OR
    Go Hawks!

    Exactly. These 'theoretical' values are not comparable to actual output values published by respectable lighting manufactures, and are usually misleading at best. But it is important to note that not all LED manufactures fall into this category.
     
  20. May 14, 2017 at 5:02 PM
    #40
    Icepuck72

    Icepuck72 Well-Known Member

    Joined:
    Jul 9, 2011
    Member:
    #59563
    Messages:
    1,809
    Gender:
    Male
    First Name:
    Mark
    Richmond, VA
    Vehicle:
    2006 DC TRD
    OME 885's, Dakars, TRD Trail Edition Wheels, 285x75x16 Goodyear Duratracs, Magnaflow 12256 muffler, Fog Light mod, 2 Baby seats
    I agree..I rarely even look at lumens any more. They're all hyped...With that being said...they're much brighter and crisper than the regular 921s.
     
To Top