I’m not quite sure what the appeal is of a Firefly mode where you can barely see the emitter compared to a moonlight mode.
A firefly that dim seems pretty useless. Probably not even enough to light the way from the bed to the toilet in the middle of the night. What do people use it for? Inspecting the surface of the LED for defects?
If I understand correctly, the main questions are:
I don’t know how feasible it is to add, but Sofirn tends to have a much more general audience than a lot of other BLF-related brands. They sell quite a bit to regular folks on Amazon, for example. So it needs to ship in a configuration which is suitable for that market.
If the hardware can do it, and if we can get flashing adapters made, what about adding super-low levels in a special aftermarket firmware?
I’m not obsessiing over the 0.01lm concept but I do generally like and support the idea if it doesn’t impose too many other constraints on the flashlight. Back in CPF’s glory days (admittedly before my time), one of the favorite lights there was the CMG Infinity, which had a 5mm led run at ~ 10 mA, which was about 1 lumen for the white led version. I have several Infinities and found that covering it 90% with my finger so that around 0.1 lumen got out still produced plenty of light for navigating a dark room. I don’t know about 0.01lm and I think it would be interesting to try some experiments at that level. I’d even try 0.001 lm.
Anyway, from my perspective, it’s of some interest whether the SP10 Pro hardware can support super low levels; but if so, it’s fine with me to have to reflash the software to reach them.
Based on old CPF discussions, these very low modes might be of most interest with a cyan led (505nm) rather than a white one. That’s the wavelength where the eye has the most sensitivity, among other things.
Jon_Slider, the light you really want might be a Rigel Skylite. I might someday DIY something like it, with a cyan led.
@ToyKeeper
Somewhat related question:
How is the ramp floor actually calculated? Is level 1 1/150th (0.6667%) of max brightness or max draw?
ok
sorry I mentioned it, and explained it, and justified it, and quantified it
and still its like… no… we dont need no stinking Firefly
ok… really sorry I asked
I’ve been improving Anduril’s options for low-light and low-power modes, like here for example… but I don’t know what will be feasible to do on this particular light.
If it can be done without sacrificing anything important, it probably will be. But I don’t know yet if it can… and I’m not deeply involved in this project, so it’s not up to me.
About the RRT-01 though, someone was asking me what the fuss is about, so I tried to explain it to them… and they didn’t really “get” it, so I took some pictures.
jon_slider knows all about this, but for everyone else…
The RRT-01 is a very unusual light, because it can go ridiculously low. I mean, if you think it’s dark outside at night, that’s nothing. If you think it’s dark in the middle of nowhere during a new moon, that’s still not dark enough to compare. It can go so low it’s literally below the ability of a human eye to see it.
Here is a picture showing a tritium vial next to a RRT-01. The tritium vial is 1mm wide, and about 6 years old, so it’s nowhere near as bright as it was when it was new. The RRT-01 is on, and making light:
Wait, you can’t see it? How about we crank up the image gamma. Here’s how it looks after some post-processing. The surface of the LED is darker than the reflection of the tritium vial… approximately the same brightness as the camera’s noise floor:
And that isn’t even the RRT-01’s lowest setting. It’s just the lowest level the camera could pick up.
To make it easier to see, I took another picture. This time, the RRT-01 has been turned up several levels, and I allowed some outside light into the room.
It’s still well below 0.01 lumens in this photo.
Runtime in this mode is about 1.5 days. I measured the power usage of the lowest visible level at 19.1 mA. It is frequently mistaken for being off at these low levels though, so it’s prone to emptying its battery without anyone ever knowing it was on. It’s easy to turn on by accident, so when I’ve carried mine, I ended up compulsively double-checking it to make sure it’s off.
I would love firefly capability, but it takes commitment from others of their skill and time. I really appreciate the work that has gone into this light so far. If this isn’t a feature any of the contributors is prepared to pursue, I’m happy regardless.
If the light does ship with programming pads, that also leaves open the possibility of accomplishing firefly via a future update.
I notice that page 183 of the t1616 data sheet says something about 16 bit PWM resolution, while I remember (maybe inaccurately) 10 bit resolution on the t1634 used in the d4v2. So maybe that can help with super low output, though it is pwm and might be flickery.
According to what I can follow in these last few pages, I don’t think anyone is being strictly anti-firefly. From what I understand with my limited grasp of the technical side of this type of thing, some are simply concerned that implementing such a mode may add more work/complication/difficulties for those directly involved - and others question if it’s really a feature that’s wanted or needed by the demographic of people who would buy this light.
Personally, I find ultra-sublumen levels to be a novelty more than anything, quite like aux lights. Cool and pretty, but ultimately I could take it or leave it. But that’s just me.
As TK mentioned, Sofirn’s intended demographic tends to be more general as opposed to enthusiasts who might really appreciate the ultra-dim mode. I think it’s just a matter of it being a little unlikely to happen, based on how Sofirn operates, especially if it adds more time or cost to the development of the light.
That said, Sofirn are showing that they do listen. This thread and others like it are evidence of that much. Who knows, maybe we will get those sublumens 
I don’t need it, but I would welcome the option if it was already there.
I find useful and I want have true Firefly mode.
Associated with a super warm LED 18k-20k in the middle of the night,when example you get out of bed to go in bath avoiding turn on lights house and consequently losing sleep.
For save battery firefly mode would be good on smaller cell like button or AAA.But there is electronic limitation,often the driver eat more current than Led.
This isn’t a problem,I can recharge or change battery
Is there any merit to a configurable moonlight mode?
For example, while in moonlight, click 9 times then keep holding on the 9th. Moonlight will flash once then begin ramping up. Release when you’re at the level you want to set. If you need to step down, do a click+hold and it’ll start ramping down. Release at desired level. Then, either click once to turn off and commit that level to memory, or wait 2 seconds for a single flash—indication of level committed to memory.
Yes, that’s the floor setting.
If you’ve ever had an infant, but needed to find something on the other side of the room at night, and you don’t want to wake it… you appreciate a firefly mode.
If you’ve ever been out in a tent , miles from anything, in the middle of the night, overcast skies, a pitch black that can only be beaten by a sensory deprivation chamber or a cave, you appreciate a firefly mode to find your pee bottle and not sear your own retinas.
If you’ve ever been in a power-out and want to leave a light for your loved ones you leave it on next to them in firefly mode so they can find it but it isn’t so bright that is bothers them. Whereas you, yourself may have a disciplined routine and system so that you know you can reach out and grab your light or whatever else you need night or day, light or no light, the kid may just sleep in their own room still because they appreciate the firefly mode.
Thanks TK for the insight.
PhantomPhoton, maybe you want a GITD tag on the light?
I also appreciate a sub-sublumen mode and am willing to pay extra for it, but I think doing it really properly would require extra analog stuff in the circuit that this light won’t be able to have. It’s something I hope the BLF hardware brain trust can keep in mind for a future light, but that may be the best we can hope for.
Also, as TK indicates, there is always a little bit of variation between individual leds, and between individual drivers. In the narrow region where the led is lit but only barely lit, those differences may make themselves noticed. So doing the best possible sublumens with pure software might require not just reflashing the light, but also sitting around tweaking the settings to get the specific light operating the way you want.
You could always put some kind of cap or filter over the bezel to let just a little bit of light through. A little circle of this stuff in a 3d-printed holder could do a good job. The 16x version would reduce 1 lumen to 1/16th, and two layers of it would give 1/256th.
https://www.aliexpress.com/item/32710187671.html
Something like this could be cut down to use as the holder, or even more awesome would be if it could be made in neutral density gray: https://www.aliexpress.com/item/32991805908.html
Very well said. Basically, the original main objective of SP10 Pro was to outfit SP10S with Andúril - nothing less, nothing more, the first Andúril flashlight with dual-chemistry support for both Li-Ion and Ni-MH/AA batteries. Of course, further improvements on the SP10S concept evolved over the course of time.
If specific sublumen parameters can be achieved by software enhancement, I do not see any issues at all, at least not on Sofirn's side if I may say so. The only question is when ToyKeeper and/or gchart will be able to provide a corresponding hex-file with improved sublumen-support to Sofirn. gchart just created a new hex-file with all latest updates incl. auto-lock working in Simple UI (being active only, if deliberately programmed in Advanced UI in the first place!). AFAIK, Sofirn is really pushing things forward now. They have already started to work on a new prototype which is being sent out to ToyKeeper, gchart and Tom E for in-depth analysis within the next days. If those ultra low lumens could only be achieved by significant hardware changes, I personally doubt that Sofirn is willing to redesign their entire driver design incl. new testing loops. I am sure we will find more nice things we would like to see in SP10 Pro but this probably something for revision 2.0. ;)
How ’bout that pocket clip? 
Any feedback on our feedback?
Update: Barry said the clip will be the same as the one used in SC21. I guess it's too late for any design changes. :((
I wonder what SP10-v30 is? 
I must be odd… the original D4V2 would do 1/150 with dim red aux, also the D4SV2 would as well. Those are both of my extreme dark bedside lights. The new driver Hank uses, the 5A CC + FET will only reliably go down to 3/150 and that was way to bright for deep night uses. So, I caught and released a D4V2 with 219b’s. But. I have serious eye problems, to much output when night adjusted and everything is gone for about 30 seconds for me.
Hmm, that’s unfortunate. Sounds like Sofirn probably made a bunch of hat clips for the SC21 and needs to get rid of them before considering any new designs.
The first two are typically referred to as a moon mode. It illuminates things to roughly the same brightness that the moon does.
The last one is usually handled on other lights by adding a button LED or aux LEDs. These stay on whenever the main LEDs are off, so the light will be easy to find in the dark. They also frequently indicate how much battery charge remains.
In terms of numbers, the boundaries are fuzzy, but in approximate terms, the low output levels are:
Moon modes are useful for seeing the ground in front of you without waking anyone or compromising night vision. Firefly modes are typically too low for that, but are still useful for locating the light.
The button / aux LEDs are generally in the range of 0.01 to 0.1 lm, while the main LEDs on most Anduril lights typically go down to about 0.2 lm. So the entire range is usually covered. But this light has nowhere to put aux LEDs, and it sounds like it might not have a button LED either.
I was bothered by the poor low modes on his linear drivers too… so I finally fixed it. At least, as much as the hardware allows. This involved a few different changes:
On all five of my linear-driver lights, the lowest mode is just as low as the original D4v2, or in some cases even lower… and the bottom end of the ramp is lower and smoother. Plus, it still has the aux LEDs if you want something even lower or a different color.
However, the power regulator is still somewhat sensitive at the bottom few levels, so if you go directly from turbo to moon, it’ll usually turn off for a second or two before it turns on and stabilizes. This is kind of a silly corner case though, because if you go from turbo to moon, you probably won’t be able to see anyway.
Also, the default floor is still set relatively high, because Hank wanted it to be high enough to avoid the turbo-to-moon issue. But this can be lowered by the user.
This is all brand new though, so it may require flashing new firmware yourself. I uploaded new builds yesterday, and am still adding and improving things.
After these changes, I finally got the linear drivers to work as well for low modes as the older FET+1 style… and in some cases, even better than the old ones. But it takes a bit of tweaking for each light to get the settings just right.