My point is that if the mirasope is actually a proper parabola, what djozz found by measuring with his luxmeter (that blocking the outside didn't affect the lux that much) would be normal.
It should not look like that picture above of the tests he did blocking out parts of that other flashlight's reflector.
There is hope ;)
OK, I guess I don’t understand parabolic reflectors well enough to say whether you are right or wrong regarding how collimated beams are, but I don’t think you are understanding everything, either. Regardless of whether the reflector is perfectly parabolic or not, the relation I=LA still holds. This means in order to get maximum throw (at some point far away from the reflector), the entire reflector must appear to reflect the LED, so all parts of the reflector will reflect the light towards the beam center. This condition seems to be in disagreement with your understanding/definition of the reflector being focused, right?
Again, I=LA means that a larger reflector provides more beam intensity since we focus our flashlights to maximize the throw at some far distance. I mean, this is the whole motivation for making a large reflectored light, to make it throw farther.
Just not getting my hopes up high.
I think he’s thinking parallel beams and you’re thinking tiny hotspot (like slightly convergent)
In real life it is the same thing.
Right, focused to infinity is as close as an optical system will get to actual parallel rays. But if one is not careful one will predict incorrect observables, like saying that a larger reflector doesn’t give more throw.
I’m with EasyB in that since our light source is not a point source but a plane, part of the light source is not in the focal point and therefore we are not producing a parallel set of rays to infinitely, but an actual image of the die that at distance will be way larger than the diameter of the reflector.
In a theoretically ideal parabolic reflector the image is a perfect copy of the source, and every part of the reflector produces exactly that same perfect image in the same spot. I think that a good telescope mirror is close to that: blocking any part of the mirror will cause dimming of the image but it stays the same sharp image. (a mirror telescope is in principle a reversed recoil thrower with an extra enlargement step by the eye-piece)
But in recoil flashlights we have, as with conventional flashlight reflectors, a focal point very close to the reflector and here comes trouble: different parts of the reflector produce a different image, caused by varying distances to the light source, and the shape of the light source varies too as seen from different sections of the reflector. Things are going on that I do not fully comprehend but do deviate vastly from the simple ideal lens imaging that we learned in secondary school, that only is approximated when using lenses with long focal lengths.
But actually parallel rays are not focussed, they’re parallel, so they never meet.
But in practice it’s the same.
Theoretically a larger reflector will give a larger projection though.
If all works out eventually, i want to make it focussable by moving the reflector.
Make it a zoomy 
Ok here we go, several things. First of all, I would like to point out that this is why djozz's images showed that spotlight with the center blocked out projected a concentrated point instead of a ring of light:
This is how you get a high lux reading, not high throw, past that point at the left the rays diverge and you end up with no good beam.
Next, this is the light rays coming from the focal point of the parabolic lens. This is how you get maximum throw, all light beams collimated. Please ignore the half dozen other reflectors in the bacground, those are inactive.
This is just the 9" mirascope reflector (green) being used. 9" diameter, 1.5" depth.
Now, this is how I test beam divergence. The origin is offset 2mm from focal point, so this is the light coming from the LED at the greatest possible offset angle (in this case it would be an XHP70 which has about 4mm side length)
Ok, now let's take a look at what happens with small vs large reflectors. First the 9" mirascope, then a larger 35" reflector, similar to what is used in the old military searchlights.
As you can see, the larger reflector has much smaller "skew" due to the offset. This means the light that is not coming from the middle of the LED, eg. the edge, still gets very well collimated.
This is because even though the two mirrors have the same curvature (and collect the same 120 degrees of light) the larger lens is farther away (because of a longer focal distance)
So the light rays coming from the center of the LED, vs the edge of the LED, have a much smaller difference angle when they hit the reflector, and are bounced back also at a much smaller difference angle.
This is why for long range lights, the size of the light source matters a lot. Instead of using a huge reflector, you can use a smaller light source and get also a huge amount of throw.
This is why the maxabeam uses a 0.25mm short arc lamp. Even though it only makes a few thousand lumens, it can reach many kilometres with only a 8" or 6" reflector or whatever it uses (i forget).
Lesson is, the light needs to be going straight forward to be collimated, not like the first image, that is bad. That is what you use for burning paper and matches, not for lighting up kilometres away :D
Interesting, looks like a neat program. Thanks for posting the pics; this has aided in my understanding of the subject.
One thing that I’ve been thinking and now can put better into words: for smallish reflectors, focusing at 5 or 10m makes a beam that is practically close to collimated. For example, consider a 40mm reflector with an XPL HI that is focused at 10m. The divergence angle from the focus not being actually at infinity is 0.002rad [inverse tan(.04m/(2*10m)]. This would cause the beam to diverge an extra 20cm at 100m, which is a small amount. In particular it is small compared to the beam divergence that exists as a result of the LED size, which is approximately 0.06rad.
So while what you have been saying about reflector focus is true, for most lights that aren’t huge, focusing at 5 or 10m results in a practically collimated beam.
Yeah, this kinda only applies to large reflectors like the ones in searchlights, and also to an extent the 6 or 9" reflectors of mirascope size, not the kind of flashlight you use to light a path or carry in your pocket.
If you're interested and have a few days to learn a new program, the optical ray tracer is free to use :) http://arachnoid.com/OpticalRayTracer/
The hardest part is taking manufacturer given height and diameter and putting those in the program by adjusting the parabola radius, since focal point of parabolic reflectors is not given often unfortunately.
But yeah, this program helped me find a good lens setup for my thrower flashlights and is pretty awesome!
Hey guys,
Just got my mirascope in, and as a mirascope it’s not impressive, at all…
The mirror surface is a bit messy unfortunately…
But, a quick test with an XP-E2 light looks kind of hopeful none the less.
This evening when dark outside i will see if it’s worth working with or just a waste of time and money.
Another benefit of large reflectors is that imperfections/aberrations have less effect on the beam ;)
Like, look how ugly this is, yet it still throws several kilometres xD
Meanwhile it got dark outside, so i can do some beamshots.
I placed a ‘screen’ in the garden, will be projecting out the window.
Nice setup!
Oh, I am SO looking forward to this! 
Well, i’m eating now, i tried some stuff, and well, it works, but…
Beamshots will follow later…
I have to admit, I didn’t think this idea would be taken to this level. Great to see people really putting the effort into something new!
Okay, some beamshots then:
And compared with Jacob A6:
I should have used a tripod…
So with the same power it seems to throw more light in the right direction than an A60.
This was only 10 meters distance to the screen, so i’m still not sure what it will do over longer distances.
Quite a bunch of anomalies around the hotspot, which seem to come from the outer perimeter of the recoil reflector.
Will do some more pics later to investigate further.
What bin XR-E and XP-E are you using? I’m wondering how much of the difference is from the emitters themselves. Or, possibly, the difference in emitters could be working the other way, depending on bin, so the difference from the reflector set-up might even be more than shows in the pics!
I have no idea what bin those LEDs are.
The XP-E is from a SingFire 348, so probably nothing special, The XR-E is how i bought the Jacob.
I was assuming it would be similar enough.