Emisar d4s runtime

What is the intended use?

I use a mule, ceiling bounced, when I want to take photos indoors with natural-looking lighting and no shadows. I also use light bulbs or panels sometimes for indoor lighting. Otherwise, I find pure flood is almost never what I want, especially in a torch.

For daily use, I find a beam of roughly 5 to 15 cd/lm is usually about right. Or outdoors, anywhere from 10 to 1000 cd/lm. A mule is generally 1 cd/lm or less.

1 Thank

Intended use is almost everything I do with a flashlight.

If I’m walking my dog I wanna be able to light up as much stuff on the side and floor in front of me , less than 20 feet ahead of us.

VERY RARELY do I look at something more than 20 feet in front me

This is one reason I don’t care for throwers, plus when I’m using a high candela light for walking especially in the house I cant stand the fact that a ton of my capacity is being dissipated into the stupid hot spot that i have no use for plus frankly it’s annoying as hell to. I’d rather it go toward runtime.

Even when I’m walking outside I hate seeing that hot spot

IMy other issue is that I don’t find anything under 300 lumens to be good bright enough, thats why I said my preferred range is somewhere between 300-1000, I’m not picky whether it’s 325 or 650

When I’m working on my car from the top I want pure flood

When I’m working under my car I want pure flood,

when I’m looking for tools in a bag in the tool shed I don’t want to ceiling bounce, I want to point the light at the tool bag and be able to see around the bag and on the wall

When I’m fixing something in the attic I want very bloody

If I had a basement, I’d want a bloody beam again.

Not necessarily 120 degrees but at least 80 degrees.

I know a lot of lights have a 80 degree spill like zebra lights but it’s not a very bright spill

That’s why Im looking for something that has a wide spill angle, can not only do but maintain over 300 lumens for many hours, without dimming

Tons of light can do 5000 lumens but for 10 seconds, tons of them are very pretty and are made from some super cool materials like mokuti or zurcuti, and there there are more than 2 dozen lights that can maintain 500 lumens for hours but they are huge and not necessarily floody

If I was a machinist the easiest thing to do would be to make my own light, but I’m not a machinist and the few people I’ve contacted that can make lights have given me $1000 and up quotes

Not gonna happen

It looks like I’m just gonna have to go with good enough, since I can’t find perfect

Since none of these lights are as floody and smooth as I want them Im thinking about getting that luminit filter in 80 degree angle, if i can buy some without having to buy a square yard of it.

And then install it on a light

I haven’t made up my mind yet, but I’m looking at nextorch ta30 which gets 6 hours on medium, ta30 max at 5 hours, zebralight sc700, emisar d4sv2, acebeam e75, and some others

Fenix c7 can also do 6.5 hours but Fenix, and nextorch are a little too long.

I’m still deciding.

1 Thank

Ah! I’ve just realised where I’m going wrong, I’m an idiot.

Obviously these lights have 4X 519a, so the current is lower, and therefore the efficiency is higher!

Apologies all :sweat_smile:

It helps to quantify the desired cd/lm value, basically the beam shape… because it can make a bigger difference than the lumen value. For example, if you use a mule at 600 lm, it might only be 0.5 cd/lm or an effective brightness of 300 lux at 1 meter. But to get the same brightness from a very floody torch at 3 cd/lm, it would only need to produce 100 lm to reach 300 lux… and the runtime would be about 6X as long at the same brightness.

It also usually helps to have the beam focused at least a little, if you’re walking the dog or something, because of how distance works. To get the path ahead illuminated evenly while pointing a light forward, the center of the beam needs to be more intense than the edges, because it’s going farther before it hits anything. Otherwise, with full flood like a light bulb or lantern, the nearby objects end up so much brighter than everything else that it becomes hard to see past them.

I personally like about 5 cd/lm, but it sounds like you might want something a bit lower… which you can get by taking a ~5 cd/lm light and giving it a frosted optic. Perhaps you might like an Emisar D4K-boost with “additional floody optic”? (10623 frosted) I find it a bit too floody, but the beam is very smooth and illuminates a 180 degree area, with the middle brighter than the edges but no distinct hotspot. The beam is as wide as a mule but it’s quite a bit more efficient. And if it’s still not floody enough, you could add some DC-Fix on the lens to spread the beam out more, or just remove the optic entirely to make it a mule.

It’s common to see 100 lm/W OTF even from simple 7135 drivers. Relatively efficient torches often reach 150 lm/W. Some bare LEDs are spec’d at over 200 lm/W. So I don’t find 116 lm/W to be particularly impressive.

However, that said, selfbuilt is using the same Zebralight-based lumen scale he used ten years ago, before the community really started getting stuff calibrated correctly. It’s expected that his numbers will be relatively high, because Zebralight’s scale was inflated.

OTOH, the 4.5 hour figure earlier didn’t count all the light emitted after the initial step-down… so it underestimates the actual efficiency. I suspect that the inflated lumen scale plus the omitted post-stepdown output probably cancel each other out somewhat, meaning 116 lm/W might be about right.

If runtime is prioritized over beam quality though, the efficiency can be a lot higher. Instead of a high-CRI neutral white rosy LED like selfbuilt tested, a cool white medium-CRI LED would generally get quite a bit higher efficiency. I wouldn’t be surprised to see 150 lm/W from a 6500K SST-20 version of the same light. It’s too bad Hank doesn’t offer XP-L2 or SST-40 or something similar, since that would get even more lumens per Watt and the beam issues wouldn’t matter much for some types of use.

2 Thanks

This is going to be a big problem with mules because a great percentage of the emission is directed way off to the side, even outside peripheral vision. With mules you need a ton of light to make something well-lit even at short distances, which introduces one more problem: when walking you will be blinded by the spill immediately under your feet and thus even more unable to see objects a short/mid distance away in front of you.

I think what you really want is not a mule but a multi-emitter floody TIR light. Something like a Convoy S21D with the “60 degree” optic. It’s more intense than a mule and thus takes fewer lumens to make something well-lit, also the beam to darkness dropoff is continuous, which makes things more evenly lit IRL.

2 Thanks

I have a convoy s2+ already I’m not impressed with the runtime at all , it’s not terrible but it’s not great plus it gets way too hot even on medium

I’m mostly looking at the zebra light sc700 with luminit filter, and emisar d4s with 26650 6400 mah cell from vapcell.

I also like the 300 lumen mode for 6 hours on the nextorch ta30 but it’s a bit too long kinda

I think hink I might have to get both and compare them but I’m not not exactly rich, I wish I could get a good beamshot comparison between the zebralight on 300-500 setting and emisar with the floody lens that hank offers

I’ve been looking all day today but haven’t been able to see a 300-500 on both . I see beamshots and runtime but everyone always tests the turbo mode esp when they do beamshots

I don’t use turbo almost ever

But anyway

Convoy M3-C with a 70.2 is floody and runs a long time .

I still think the Convoy S21D fits your needs better than the zebra for a few reasons:

  • Wider and much smoother beam. No distinction between hotspot and spill and no sudden dropoff, just a continuous wall of light.
  • Does not require a diffusing filter, which loses a significant amount of output and reduces efficiency.
  • Way cheaper.
  • The S21D uses a 21700 cell and buck driver, and is available with XPL emitters for efficiency or 519A for high CRI. It will outperform S2+ in runtime due to (1) larger capacity cell, (2) more thermal mass to dissipate heat, (3) more efficient driver, and (4) quad emitter is more efficient than single emitter given same output.