Looks like a parabolic reflector to me.
Yeah think your right on that the description was
10pcs 50mm Led Lens Base Housing For 5W 10W 20W High Power Led in Round Shape 45mm Lens
Yeah, sorry parabolic reflectors won’t work, which is what 99.99% of everything on the market is.
only ~5 collars left 
In this german thread is a pic from kenjii
Thank you very much.
The 2013 Wavien patent describes a parabolic collar, which they claim works better than a spherical one, particularly with mutiple LED emitters.
“The parabolic reflecting collar is made by rotating the parabolic curve around the axis which goes through the focus and is parallel to the directrix. The resulting reflector is round with a parabolic surface. Light emitted from the LED placed at the focus is reflected from one side of the parabolic surface, becomes a parallel beam, incidence onto the opposite parabolic surface, and refocused back to the LED itself. One major difference between this parabolic recycling collar and the spherical recycling collar as shown in FIG. 1 is that the light emitted from a point at and near the focus will be reflected by the parabolic reflector twice and back to the same point where the light is emitted. With this property, with used with packages with multiple LEDs and multiple colors, each LED will be able to perform recycling of its own light independently and each LED with its own color will be able to improve the brightness independent of the other LEDs and colors.”
Collar arrived, nothing to use it on yet though.
Yes but that uses two parabolic reflectors, one facing forward and one facing down. You can’t replace a collar with a single parabolic reflector that is upside down.
The regular spherical collar reflects an inverted image which is why you can’t use it with multiple different coloured LEDs.
The spherical collar is better however because it only reflects light once instead of 3 times and it also reflects back at exactly the same distance from where it was emitted, which means that the reflected die image has the same sharpness as the original die, with no divergence.
That looks really good, seems like they did go away from the slice method.
The other side looks sliced for sure FWIW
Well from your pic it seems like there is definitely some thickness to the glass around the hole, so maybe they sliced part of it but not all the way.
This is how the old collars were:
With 0 thickness at the edge which resulted in a not-so-clean circle.
I see. It is as you describe and not like the above picture.
Some measurements:
12.56mm inside opening at top (ID)
31.77mm wide at bottom (OD)
3.63mm bottom wall thickness
Thank you to everyone that showed interest in the collars enough to make this possible <3
I also said thanks to my Marinebeam contact on behalf of us at BLF, here’s their reply:
“My pleasure. I appreciate your help with the communication. I have been trying for more than a year to make something happen, and while they were very supportive, they didn’t quite have a direction on this ironed out, and they surely had much more critical irons in the fire. They are really just regular guys, not so unlike us, that have mortgaged good jobs and their futures to buy these patents. I spent a lot of time at Wavien also, and they too were good, hardworking, and exceptionally gifted people trying their best to get this technology out there.
Contrary to popular opinion, we won’t make some small fortune on this deal. I wish these collars were some trinket with massive volumes that we could sell for a dollar a piece and make everyone happy. When I factor in the tooling, yield fallout, scrap, license fee, and royalties, it is modest deal for us at best. I am happy to share some of the spoils with Meadowstar too. If they can be successful, it is best for all of us.
I did this BLF deal mainly to get some out there in the hands of some creative hobbyists. Because we deal in so many different products, it is hard for me to spend the available time 100% on flashlights, so a group of keen tinkerers trying different hosts, drivers, and emitters can cover much more ground than we can and push the envelope, further validating the concept. We are all on the same team trying to make cool long-throw LED flashlights. Should someone figure a clever concept for making low-cost collars, that is good too. Same with any host that gets well-enough developed that it is easy to drop in a collar. We can have the host produced (or re-produced using the basic concept), drop in a collar at Marinebeam, and sell to the public. We could even have a BLF version and pricing for BLF members.
Have a great weekend, and let me know of any interesting successes.”
Can you measure the thickness for us?
Actually you can, using a Compound Parabolic Concentrator (CPC) which is a type of none linear optic i.e. doesn’t use image forming.
I think what Wavien describe in their recent patent may even be a CPC.
Their description of a rotated parabolic curve is the same way a CPC is constructed, and I see no mention of there being two reflectors. We’d have to see the patent drawings to be sure, if anyone could get hold of a copy ?
A CPC works by concentrating the etendue of the system without the need for “light recycling”, indeed it wouldn’t work if light actually bounced back onto the LED. And is pretty much un-patentable.
See e.g. https://www.youtube.com/watch?v=71c4Yat8jRI
You can buy them here: https://www.edmundoptics.com/optics/optical-lenses/specialty-lenses/compound-parabolic-concentrators-cpcs-3213/
Or here: Electroformed Compound Parabolic Concentrators
One of these, with a collimating optic on the pointy end could be an interesting project, but I agree it’s not much like the conventional spherical collar.
Credit where credits due a big thanks to Enderman for taking the time and making the effort for this to happen for members here and elsewhere and to the members who participated I’m looking forward to see what you do with the collars. 
:+1:
I just got mine today as well,A big thank you to Enderman. You Rock!!!
A CPC has nothing to do with increasing the intensity of an LED light source.
It’s just another type of optic for narrowing down a beam of light into a smaller aperture (at the expense of less collimation)
What I said earlier is that you cannot use a parabolic reflector upside down, it needs to be spherical. The image below only works when the reflector is a sphere because it projects the die image back onto itself.
If it was a parabola instead of a sphere, almost all the light rays would not return to the die.
The “dual-parabolic” thing is similar to this:
If you ignore the 33, 34, and 35 components the combination of two parabolic reflectors effectively does what the spherical reflector is doing, reflecting the image of the light source back at itself to increase intensity.
Thanks guys 
I posted some measurements. Here is my excellent drawing. I assume what you are asking for is the bottom thickness that I have as 3.63mm
The lip on the top Enderman was referring to is prob ~2mm. I dont have an easy way to get my calipers in there and I dont want to scratch it.
I see what you were talking about now, this image with the “rotated parabolic reflector”
“The parabolic reflecting collar is made by rotating the parabolic curve around the axis which goes through the focus and is parallel to the directrix. The resulting reflector is round with a parabolic surface. Light emitted from the LED placed at the focus is reflected from one side of the parabolic surface, becomes a parallel beam, incidence onto the opposite parabolic surface, and refocused back to the LED itself. One major difference between this parabolic recycling collar and the spherical recycling collar as shown in FIG. 1 is that the light emitted from a point at and near the focus will be reflected by the parabolic reflector twice and back to the same point where the light is emitted. With this property, with used with packages with multiple LEDs and multiple colors, each LED will be able to perform recycling of its own light independently and each LED with its own color will be able to improve the brightness independent of the other LEDs and colors.
[0009]
The parabolic reflecting collar is positioned such that higher angle light rays are reflected twice, off opposing wall reflecting portions, back to their point of origin. Preferably the light source is an array of multiple LEDs having different colors and sizes.”
So the thing is that it doesn’t actually work. The lines are not reflected back to the point of origin as descibred in the patent. Would be cool if it did, but nope. It would need to be a different shape than a parabola, and then it would still only work from light rays originating from the center, not any point on an LED. 