Convoy S21D w/ legendary Nichia 219b :) - Review & comparison w/ Nichia 519a, E21a & other lights with Nichia LEDs incl. Emisar D4V2, Convoy S21F. (Summary & measurements on P. 1)

This is an excellent point! My lumens analogy with duv is not an accurate one. Subjectively, however, I still find that the same duv difference is more pronounced when a source is close to neutral. I've played with the color space calculators from Waveform Lighting before, but thank you for introducing it anyways; it's an excellent resource.

After playing more with the calculators, I've created the following experiment for everyone to decide for themselves. It's not perfect, as the computer monitor does a poor job of recreating colors, but hopefully there is something to be observed. In the image below, all 4 sources are at 4500K CCT. Compare the top two squares, and then the bottom two. Try to decide: which pair has the greater difference in duv?

The duvs differences are about equal. (0.0098 for top, 0.01 for bottom)

@QR, I appreciate that you put a lot of thought (and “work”) into every post :+1: .

For anyone new to this and wondering what on earth we are talking about, if you like to try, Waveform Lighting calculator web site is HERE. It would get you “deeper” into the hobby and believe or not, A LOT OF FUN. Guaranteed lol.

The numbers come from this inexpensive device “Opple Light Master Version 3 (aka Pro)”. It’s the more expensive $40 item - do NOT get the cheaper Version 2.

Just purchased the Opple Lightmaster 3 Pro. Look what you just made me do!

@stephenk lol coming from you I’ll take it as a compliment :+1: :slight_smile: .

stephenk has over 2000 posts...

and I just noticed his username.

Are you a fan of Stephen King perchance?

Beamshot of Convoy AR Green vs UCL AR Purple. Picture shot in RAW format with DSLR. Manual setting & WB @ 5000K. Both light and camera on tripod to minimize fluctuation. I noticed something that’s not frequently mentioned, that changing lenses not only affects Duv, but also shifts CCT about 50-100.

I hope you could see the UCL purple lens makes the scene more yellowish (attention to the white wall @ middle of the picture) but please keep in mind photography tends to exaggerate difference IMHO/experience. In real life use, I could see this difference against a white wall but it’s not as severe as this picture would indicate, and much harder to see when shining on darker subjects with colors instead of a white wall.

Even though this is comparison against a UCL AR Purple, not original Convoy AR Purple lens (I couldn’t find one to buy), I am glad that all my Convoy lenses have green reflection. :innocent: It does magically remove yellow tint from a scene.

Thank you for the wall beamshots, and for the enlightening discussion! The difference between the two seems pretty remarkable. Also very interesting observation about how CCT drops with increased duv--I suspect it's due to the fact that in the CIE 1931 color space, isotherms are not orthogonal to lines of constant duv. It's very strange that the CCT assigned to a point is NOT the closest point on the BBL by Euclidean distance--could someone explain why they made it so?

I also noticed how difficult it is for white to look white at warmer temps...neutral looks yellow-green, while -duv looks peachy or brown or skin-toned, there is just no white.

The nicest warm white I've ever seen came from some fridge bulbs, which are incandescent, heavily-frosted, and noticeably magenta instead of neutral--I'm not sure what filter they added to make them that way, but the tint is just pure awesome, with extremely good reds that beat any unfiltered incandescent and even my dedomed 519As. And it actually makes white things look white even at 3500ish K.

Has anyone seen the wide range of options available (219b, 519a, XPL HD, 8 emitter colours, 8 x different degree options) at

I came across it after watching a Weerapat Kiatdumrong review on a different light, but the size comparison was with the 21D which caught my interest.

There is a note re the driver differences:

“12A FET driver ,max output 12A,It is greatly affected by the battery voltage and internal resistance, and the current drops rapidly.

8A CC driver, constant current buck driver, As long as the battery voltage is greater than the LED VF value, the driver still maintains a high current output.In addition, the new 8A driver uses springs instead of copper posts.”

I would go for the lower amp driver anyway for less heat. It also sounds as though with the 8A driver uses a spring at the head, instead of a post - this also appeals to me; I’m considering something like this for a bike light / edc.

If I could ask some questions here, I’m guessing at some things without actually knowing:

1. I’m not sure I understand the “degree bead” options. This looks like a wide angle setup - what option would give the most throw here?
2. Being a flashlight Philistine I’m just after the best light to light things up; which would be the best emitter?
3. If I wasn’t a flashlight Philistine, which would be the favored 519a color (4500?)?

1. 10 is the narrowest beam, 60 is the widest. Even with the 10, it’s not exactly a throwy flashlight, though.
2. 519A has great brightness, tint, and CRI. Most other emitters only have two of the three.
3. That depends on your taste. 4500K is what I like, although 5000K is all right indoors too. You might like something way different, but 4500K is probably a safe starting point for a lot of people.

Thanks for that nicodimus22

I’m not sure if you have look around, short list and choose a S21D-a light with TIR for a bike light.

Or at some point, you are confuse while reading this thread.

Imho, take a look at the S12 triple emitter with 519A-8A driver. It’s more appropriate then the S21D for a bike light.

I will, thank you

I was thinking the same thing. Bike light probably has specific requirements such as certain brightness and not shining on eyes of people in opposite direction, etc., and the S21D with Nichia (lowish brightness and throw, no easy control of direction) is probably not the best candidate, as nutty as I am about it :slight_smile: .

BTW the picture shows why I feel confident about Duv numbers from my Convoy AR Green Tint lens’s comparison. IMHO, the Opple’s number one “enemy” is change of position and angle between light and sensor. While there is nothing wrong with hand holding for quick reads, for me the only way to prevent fluctuation and improve repeatability is tripod use.

It turns out one weakness of my test, I couldn’t find the original “bad” Convoy AR Purple to buy and have to use the larger UCL Purple, is the same reason it is consistent, down to 4th significant figure of X,Y coordinates. I just hold the lens in front and record numbers. Moving it off shows instantaneous change, without touching the light. Where other testers may record one number, I could do 10 instant A:B comparisons (the best kind of comparison IMHO) without much effort.

To me the Opple is a garbage-in garbage-out type of device, if you want accurate comparison test, nothing beats repetition and nutty OCD. I have a nerdy notebook with pages of data to prove. :+1: :slight_smile:

For me photography is the final arbitrator - to answer the question “is there a difference.” I was happy that the beamshots support Opple measurements so closely.

We’ve seen Convoy AC Green vs UCL AC Purple and the rather remarkable effect of a 0.002 Duv change. I was also curious about Convoy AC Green vs NO LENS. Nearly identical Duv. No point to post BBL graph because they occupy basically the same spot.

But what about beamshot - as expected by the nearly identical Duv, the beamshot of Green vs No Lens are very, very close, but not quite the same to my eyes. But… I’ll let you decide for yourself. Is Convoy AC Green that magical, “better” than no lens? :slight_smile:

Repost of Convoy Green vs UCL AC Purple for comparison (much larger Duv change here)

Thank you so much for posting the comparison wallshots! The green coating really does make a difference over no coating, to my eyes at least. If you had shown me just the centers of the two beams, I probably won't be able to tell. But with the edges of the beams also in frame, the reduced tint shift from green coating is more obvious.

@QReciprocity42 thanks for the nice comment. The comparison test took a lot of time and effort so it’s good someone besides me is enjoying it :slight_smile: . The change in color of the beamshots is subtle yet still took me by surprise, especially Green versus NO LENS where Duv are so close I didn’t expect to see a difference. BTW I took 2 shots with each lens and did two rounds. The second round shows the same extremely subtle but noticeable change.

Another interesting note is UCL Purple was clearly worst wrt Duv, but best with light transmission. It consistently has highest brightness among the three TS30S lenses.

In earlier post above you were discussing change of Duv of a LED far from the BBL versus one close to the BBL. I happen to have a light that would be a good candidate to do that comparison with actual beamshots, my Wurkkos with Nichia 219c 5000k. This LED is much closer to the BBL than the very negative 219b.

I am not surprised that the purple AR has the greatest lumen count, since for the same radiometric amount of light, purple contributes a lot less to the lumen count than green, as the luminosity function on the visible spectrum peaks at green and decays quickly at the tails, which are blue and red.

I would be extremely interested to see the duv change comparison around or far from the BBL, and agree that your 219C sample is a good candidate. Though I do recognize how much work and effort it takes to make these comparison shots.

nice comparison :wink:

OT: I can’t help but wonder if optics is your field of expertise? You don’t have to answer - I’m just curious :+1: :slight_smile: . I am just glad that my observation is no BS & there’s an explanation.

The testing of these lenses itself is quick. It’s all the setting-up and little things I have to watch out for: battery full, light at warm temperature, pointing at exact same location, consistent camera setting, not mixing up the RAW files, etc., that’s tedious and a pain, if I don’t want any error.

I am running one more round of Duv testing of the 219c 5000k this evening then will post beamshots. Strange hobby, could only work at night.
Thanks Jon for the crop. Interesting isn’t it?