Noob flashaholic here and I'm coming to you with a noob question: Where and/or how is the corona formed in a flashlight beam ?
I've seen flashlights with nice round coronas without much artifacts, that you can rotate and you would never notice the beam rotating, coronas with a 4 petal flower shape, easy to notice while rotating the beam, flashlights with big coronas, smaller coronas, sharp or faded coronas, but mostly, it's always there and many a times it's ugly with some sort of artifacts.. and here I can tell that some of these artifacts have to do with the emitter alignment itself, mostly the centering, worst offender being the 4 "petals" in the case of a petal shaped corona being uneven. How come some reflectors with the same emitter would give you that petal shaped corona and others would just give you a mostly round or at least with no noticeable distortions corona ? I don't know if I'm wrong here and how much it has to do with the emitters themselves and how much it has to do with the reflector, but at least let's say for a given reflector, an emitter with a smaller die in comparison with it's dome, eg. XM-L2 would give a much more pronounced petal shape effect in the corona compared to let's say an XP-L, which has basically the same die, just a smaller dome, also cut vertical on the sides, thus not actually a perfect dome..
Best as I can figure it, it’s how far away from the perfect focus it is.
Inside a parabola, there’s only a single point in 3D space which is the focus. And an LED chip isn’t a point-source, but a flat square, a tile. So the farther away you get from that (at the sides of the chip, especially at the corners), the more out-of-focus it is. So the hotspot gets distorted.
That’s why especially in a G3 (flip-chip), which has horrible hideous tint-shift as you go off-axis, the worse the corona, which is light that hits the reflector not from the focus but slightly “off”, and doesn’t get focused parallel but again slightly off-axis, and is tinted that horrible urine-yellow color.
So, basically, an ideal reflector would reflect a point situated in the focal plane at an angle, that angle being the 90 degrees perpendicular to the emitter plane, let's disregard that for now, our emitter it's an ideal point in space.. and it will do that for every point on it's surface area - Now that would result, ideally in a hot-spot the size of the reflector, am I right ? And that would be from a single point, our ideal emitter. So in a real application, where our emitter would be more or less a plane itself, all but one point in the center of the plane would be in perfect focus and give us most of the hot-spot ? Sounds too ineffective.. and the rest of 99.999.. % of the emitter would be reflected as the corona.. ?
What I was thinking was that the emitter surface area would give the shape and size of the hot-spot - let's say the focal point in the emitter's plane it's reflected in a point somewhere in front of the reflector (focal distance ?) and whatever else surrounds it, the rest of the emitter area just gets more and more offset in a linear manner ? Thus giving us a bigger hot-spot for a bigger emitter ? Thing that can be confirmed in practice ? So, from that slide, what I'm taking as the divergent angle/s from the various points on the surface of the reflector it's basically what would form the hot-spot shape - also with distance, the hot-spot gets bigger, which makes sense.
It's just, I can't imagine a single point in the emitter surface to basically give us the bulk of the hot-spot, while 99.999.. % of the rest of the emission area would be in the corona and spill. And I still don't get how the corona is formed.. :))
So, the slides were indeed informative, but from my understanding they didn't really made complete sense. Now I think i got what I was missing.. Basically, it would've been helpful if in those graphs there would've been one more ray represented from the emitter point closest to the reflector - that would've made for one, the biggest difference in incident angle, basically reflecting outside the perpendicular angle, pretty much in the other direction than the incident angles (thus falling under the corona surface area, outside of the hot-spot area) that are already represented in those slides and second, without representing what happens to the rays while reflecting from both diametrically opposed points on the walls of the reflector, you'd never notice that the corona comes from the center towards the reflector side of the emitter's surface area that at more than the intended beam angle only gets reflected not only by one side of the reflector, but also by less and less of the reflector's surface area of that particular side, thus, the progressively reduced brightness, hence the corona.
Hopefully it makes sense.. At least this is what I think I observed by looking at those slides and also confirmed by simply looking down a reflector and tilting it at different angles. It all clicked more or less in place.
EDIT: Also.. figured out that the focal distance that I was blabbering about in a previews post it's only a side effect of the incident angles given by the emitter being a plane, instead of a dot, not the actual reflector's calculated, intended focal distance.. that being only for the ideal dot emitter. This brings me to another point.. disregarding the real life applications imperfections, the focal distance which the reflector curve is calculated is actually tending towards infinity for the given ideal dot emitter..
