【New Sofirn HS21 announcement】The first Sofirn headlamp with gesture control function

And the last question: is the left tail cap magnetic? If not, would you know which Sofirn (or Wurkkos) tail cap with a magnet would fit and would the magnet there be strong enough to hold this light? Cheers!

And this one is really the last one: can the bezel be unscrewed easily - I was hoping to stick some GITD tape inside if it can.

Here is the UI seen in the manual (looks like a standard template word document lol) provided:

No magnet. I don’t have many Wurkkos / Sofirn lights. WK03 Tailcap does not fit.

No, I was not able to do so. Seems glued, could be very difficult unless you have the proper tool for opening. :frowning:

Yeah, this could be a good thing, but only if you are using the light exactly this way. I am using lights not like this, I want to switch between light modes while doing something, so this mode memory behavior does not fit my usecase.

Cheers! The the IR sensor doesn’t have memory, I take it? i.e. when you turn the light off, and then on again it needs tripple-click to activate it every time?

As soon as the light has been switched off with IR, you must switch it on again with the button. You cannot switch the HS21 on via IR. (It is not possible anyway, accidental activation would be inevitable since the sensor cannot recognize between human hand and other objects.)

After switching off with IR, the last light level is saved and can be reached by turning on the light with the button.

You always have to switch IR on again with 3x click, this is not saved. If you switch the lamp on with the button, IR is generally not active.

So, to be sure: if the light in on IR and you turn it off using the switch (not the IR), when you turn it back again with the switch, the IR is deactivated and needs to be turned on again by triple click? Thanks!

p.s. Sorry, I missed the ‘always’ part :⁠-⁠)

In this case (turning off via button while IR sensing is on) the IR mode will be saved.
After turning on via Button you don’t have to activate this IR sensing mode again.

But now here comes the strange part: this only applies for several secons the light is turned off. After around 10 seconds the memory of IR is not active anymore and you have to activate IR again with triple-click from on. This behavior is not explained in the manual at all.

Only if you turn the light off via IR, you have always activate the IR sensing mode again.

1 Thank

One more, please: if I click and hold from off, a moonlight level is activated. If I turn it off, the moonlight is memorised? What happens if I turn it on in moonlight (either by click-hold or as a memorised mode) and then start changing levels either by holding the button or waving at IR? Does moonlight becomes part of the circle of levels or vanishes as in: Moon-Low-Med-High-Turbo-Low-Med…?

Yes.

With button and IR: moonlight disappear from mode order. Exactly like you wrote: “Moon-Low-Med-High-Turbo-Low-Med-…”

You can activate IR while in moonlight though.

1 Thank

Forgot this - yes, the modes can be cycled via IR pretty fast - sometimes the sensor does not pick up a gesture, so this could be somewhat annoying if immediate change of brightness is necessary.

While running on highest mode for runtime measurement (will follow in the next days) the sticker around the charging port came loose:

This looks ugly, and it was very easy then to remove it completely. Seems that the adhesive used is not strong enough if heat (while running on highest modes) is applied.

I also saw in the user manual that this headlamp is rated for IP 66, not 68! (there is explicitely mentioned “not for diving”) - so it should not submerged. I am not sure why, maybe this is due to the design of the gesture sensors?

1 Thank

Not sure if that has been implemented in some other lights, but maybe being able to turn the memory off globally by some clicking combination coul satisfy everybody? I use a headlamp without memory that always turns on in moonlight in all modes and I kind of like this predictability and no surprising blinding property.

p.s. A radical idea that I think could appeal to me (but then again, I’m still fascinated with rotary phones where you can actually ‘dial’ somebody rather than pushing their buttons):

How about a headlamp with no buttons, just 2 dials? One dial just like the one on HS21 for modes and the other one for intensities from off (similar to Wurkkos HD16 dial sans central button, unless IR is also added in which case the button will be turning the IR either on or off)?

Locking done with the the battery cap, or the intensity dial having an indent at the Off position and requiring pulling + turning to go to the lowest/moonlight level.

I think this could be simple, useful, original, and above all, satisfying.

Here is a runtime chart for all three light modes for the HS21. Since it is clearly not recommended to submerge the light, I tested the runtime without cooling.

The modes are regulated fairly well. After roughly calculating the capacity taken from the battery (including losses due to the optics for the LED current), I arrive at around 2920 mAh for the Throw LED, which indicates an efficient buck driver.

The longest runtime after step-down is achieved with the flood LEDs.

Other measurements:

Turbo:

Throw: 1632 lm @ 15100 lx 1m
Flood: 1282 lm @ 1240 lx
Combined: 2007 lm @ 10700 lx

Moonlight:

Throw: 2 lm @ 19 lx (calculated, 1 m)
Flood: 2.6 lm @ 2.5 lx
Combined: 2.8 lm @ 14.9 lx

The values stated by the manufacturer are unfortunately not met. Even if I consider measurement errors in my setup, there is a really huge deviation in result for the spotlight LED. Also the step-down to 1000 lm stated in the manual is not met, since it stepped down to around 700 lm. But to be fair, there are no testing conditions stated in the manual, so maybe these are the reason for these values. On the other hand, the runtimes given in the manual are also veeeeery optimistic:

5 Thanks

Do the runtimes start in turbo level (the 4th level or the highest intensity) in each of the 3 white emitter modes?

Yes, on turbo. I updated the chart to include this information.

1 Thank

Would it be safe to assume that High would follow similar runtimes except for the first few ‘turbo’ minutes?

I think so, anything else would at least surprise me greatly.

1 Thank

Am I getting it about right: it looks like the ‘throw factor’ in cd/lm is about 9 for the Spot and about 1 for the Flood?

1 Thank

Have you tested the IR switching with gloves on? I ask because winter is coming around here…
I also am a bit curious about how low ambient temperature will affect the IR switching.