There’s no glue. If it seems stuck try running it on turbo for a minute and try again.
Wait, is there a possibility the SST-20 65CRI models seem overly green not because of their own tint, but because of the AR lens?
To be clear, my 4000k emitters are green and the poorly coated lens makes them appear much greener. Swapping the lens and using a minus green filter makes them much better (no green tint and is slightly rosy).
Just tested mine - quite a bit less green without the glass. I can’t compare it directly now but much closer (if not exactly the same) to what SST-20 in my PL47 looks like which I was much happier tint wise.
Interesting. I’m not noticing it in the 4000k emitters in my D4. I’ll have to dig out some other lights to compare.
I’ve posted this pic before, but… AR coating can have a big effect on the tint, if it’s done badly. I have never personally seen this on an Emisar light though… all my Emisar lights have been good.
This is a stock Olight S10 lens on the left, next to a plain piece of glass on the right:
The Olight S10’s lens made its beam sickeningly green… but it was a very nice pure white after swapping in a plain lens.
The other worst offender I’ve seen was an EagleTac D25A Ti. It has a very nice 219B emitter inside, but the badly AR-coated lens makes the beam ugly. The center is green and it has a very distinct blue halo around the edge. In this picture, I set the white balance to match the hotspot so it doesn’t look green, but the blue part is quite clearly visible:
This kind of demonstrates how many AR coatings work. They don’t change the tint, exactly… they bend different frequencies at different angles, similar to a prism. In this case, the blue frequencies (and some red) were bent to the side while the green frequencies were allowed to pass straight through.
Unfortunately, I haven’t fixed that EagleTac light, because the head was one piece and everything was glued into it, so I haven’t been able to get the lens out or replace it.
Now I’m really going on a tangent, but lenses aren’t the only thing which can act like a prism. The latest Cree emitters (XP-G3, XP-L2) also have a pronounced rainbow effect, and this happens regardless of the lens or reflector. It’s in the LED itself, or presumably, in the LED’s dome.
One of the best pictures I’ve found for this effect was posted by Hogo of HDS a couple months ago. On the right is a Cree XP-G2 6200K, I think. It had some rainbow effect but not too bad. On the left is the newer XP-G3 5000K, showing a rather distinct rainbow effect. When the entire beam is integrated, like what would happen when a ceiling light panel shines through frosted glass, it makes a nice white. But in a flashlight it produces a pretty ugly-looking rainbow instead.
The outer parts of the rainbow are too dim to really see clearly though. So I ran the image through a filter to blur it and then maximize the brightness of each pixel as far as possible without changing the actual tint. This makes it a lot easier to see tints. Here’s that photo again after maximizing brightness:
… and that’s why we don’t use XP-G3 or XP-L2 in flashlights. At least, not without a heavily frosted lens or something.
But that got pretty off-topic. In general, Emisar’s lights have been good — both about emitter type and lens quality. So I’m hoping Hank will get back to good lenses instead of the green ones some people here reported.
^ what she said
Is AR really the culprit? Or maybe they just use low purity glass?
Not that I know anything about it, I just notice that it would give exactly the same effect as complained about - turn the beam green.
Either or, sometimes both.
Okay, so where do we order good quality glass for D4S? It doesn’t really have to be coated. I don’t care for the lumen drop.
Or is the one from KD confirmed to be decent without any additional greens? 
I received my second D4S yesterday (mixed 3000k & 4000k SST-20’s; cyan aux). While the light itself is wonderful, I was disappointed to see that it, like those in recent posts, has different glass that definitely colors the beam a little green.
I noticed the AR coating on the new one is very strongly purple. The coating on my original is a little more subtle, and is more of a pinkish orange.
The order of the lenses is flipped in the photos. The greener lens is from the light on the left.
Daneel, did you ever hear back from Hank?
So the first D4S I got was xpl hi 4000. It functions as expected, ramping up with a flash then a second flash. Then a 2x click to bring it to Turbo. The next 3 D4Ss I purchased do not increase in brightness after that second flash in the ramp and a subsequent 2x click that should bring it to Turbo. These are a 3000k SST20, 5000K xpl hi, and 5000k HD. Strangely, after some use the sst 20 verson has a very slight increase now that it didn’t before some use. What’s going on here? I’ve ruled out being a battery issue.
Since nobody wants my horribly tinted XP-L Hi 1A D4S, I decided to make it my own and live with it. May still reflow some new emitters. Haven’t decided yet. May keep it and order a different one.
Anyway I wanted it to be more floody. Had a sheet of that thick DCFix diffuser in the drawer. Cut a piece that fits and put it in between the bezel and the glass lens. Wow, super floody now and almost perfect. Well, ya know, besides that horrible tint thing. A cool byproduct of the DCFix was how it makes the aux led’s look now. Kinda cool looking. I didn’t really care for the aux led’s at first, now I can almost live with them.
So, is anyone going to make a board for the D4S to take the XHP50.2 3v LEDs so we can easily swap in at home 
It won’t fit.
And besides, even if it fit, you’d smoke everything in the light, if the springs don’t fail first.
Really? I have 3 XHP-50.2’s in a Titanium X6 making 10,200+ lumens and nothing’s smoked yet…
The dome doesn’t fit in the Angie optic opening but a slice and dice will take care of that, no problem. Sure would be interesting…
Only 3x XHP50.2s though, not 4x XHP50.2s.
And the Emisar D4S doesn’t have as good heatsinking as yours.
Hmmmm, mine is using the 6V emitters and running on 2 18350’s… even though it’s a copper pill it’s inside the Titanium head.
No idea what the 3V variants will do but I’ve been reading about it and went ahead and ordered some so I will find out soon enough. I’ll put 4 in my Q8 and see what happens. Ordered 20 of em to, uh, play with! 
Nice Dale, interested in seeing those results 
Is the tint rosy now?