Sanity check: hotspot angle

Goal is to find 18650 or 21700 light with very pleasing beam pattern, with regard to hotspot angle, spill width, and ratio of output in spill/hotspot. I like W1 emitter in P60 reflector except that the spill is rather narrow, which I believe is due to the depth of the reflector.

Comparison of beam shots in reviews shows that Emisar D1 has one of the widest spills and is capable of tight hotspot.

I used the calculator here: My assumption is that relative comparison of beam divergence angle indicates relative angle of hotspot. I input reflector diameters of P60 and D1, and played with focal length until the reflector depth matched the measured dimensions (found online in reviews and product specs). Is this the right way to do it?

I input LED diameter as diagonal length of die, and came up with the following beam divergence half angles in deg.

W1 7.7
W2 10.6
519a DD 15.1
SFT40 15.5
B35AM 23.4

W1 10.5

This suggests that W2 in D1 host will have similar hotspot angle as W1 in P60.

Is this sound logic?

What I don’t understand is why the calculated beam divergence seems so off. For example, Zebralight specs SC64c LE as 12 deg hotspot, and that is a very floody light. Surely way floodier than 519a DD in D1, which was calculated as 15.1 deg.

Am I misunderstanding the meaning of beam divergence? Or is the problem with the focal length?

Inputs are:
D1 reflector: OD 30.5 mm, center hole 7 mm, focal length 2.55 gives approx. reflector length of just over 20 mm (matching TG reviews measurements).

P60 reflector: OD 26 mm, center hole 7 mm, focal length 1.85 mm gives approx. reflector length of 22 mm (matching specs on Kaidomain product page).

Thank you.

No formula will tell you how a beam will look in real life. you may get some idea on a hotspot, but even at that very vague idea, but you wont know anything about spill. look for beamshots to compare, but again if shots are taken with camera in auto mode, it wont show you true picture, there is a site they have beamshots of many different lights, made in a dozen or two different locations, taken by a camera in manual mode, so you can reliably compare them. they have many lights compared but not every lights in existence.

In theory, this is the right way. In practice, some of the parameters (such as reflector focal length) are extremely sensitive to even minor errors in measurement--it's almost impossible to get values that are precise enough to give you even a ballpark estimate. The resulting D1 numbers are definitely way off, I'd say it overestimates divergence by at least a factor of 2.

There is a much simpler and more practical way to estimate divergence, as long as you have one reference light X, with LED diameter D, reflector diameter R, and known divergence angle A. If you want to estimate the divergence angle of light Y with LED diameter d and reflector diameter r, you can get a decent approximation via A*(d/D)*(R/r). This approximation is fairly accurate assuming similar reflector shapes and relatively throwy beams.

Appreciate the feedback.

I ended up getting a Sofirn IF22a after becoming enamored with beam shots from reviews showing a wide mule like flood and fairly tight hotpot. Some people said the spill was too dim to be useful but for me it lives up to the hype. It’s a very pleasing beam to use in my backyard. The UI is a little clunky but I don’t mind the mode memory since medium is about all that I use. The button is so hard to find though. I wouldn’t mind getting a second one to try and install a better button plus a better driver with Anduril. Possibly, would be even better to ever so slightly diffuse the beam, to get just a little bit brighter spill.