Managed to get pretty good, although could be better.
I may be remembering wrongly. If so, feel free to help me out. 
My understanding has been that a properly focused parabolic reflector takes aim from one central spot. All the light from that one spot forms the hot spot of the beam, and the rest of the light from the die is out of focus. So, the die shape shouldn’t affect the hot spot at all. If it is a particularly well focused reflector, the corona may show some rounded version of the die shape. But it depends also on how bright the spill is, and how it blends from corona to spill.
Not out of focus, but rather off-center.
A rectangular die (right) would make a slightly oblong shaped spot which is not really noticeable because of corona and also the fact that the CSLPM1.TG is fairly close to square.
Today I putted one 1mm2 white flat in a SMO reflector Convoy L6 with short tube and single 26650. You need to wait a day to measure. I’m at work and I only got a little time at home to done measuring while dark but before work.
Edit: added images.
Well, any point that is “off-center” from the focal point is what I’m calling “out of focus”. Perhaps I should have used another word, like “detuned” instead?
No, the projected images of the dies are rotated as you go around the reflector. Make a pinhole in a piece of Al foil and move it around the reflector to see.
They do not rotate.
They get stretched when the angle between the LED and reflector is large (aka near the center of a forward-facing reflector), but they do not rotate.
Try the pinhole experiment I described in the post above to see how it works. Every little part of the reflector projects an image of the die. The parts near the outside project a smaller image and this forms the “hotspot”, the parts of the reflector near the center project a larger image and this is the “corona”. Additionally as you go around the reflector the images rotate, so this is what I mean when I say it takes a circular average around the center of the die.
They do, we discussed this before. Have you tried it?
Yes, they do not rotate.
This is the same for lenses, record reflectors, and forward-facing reflectors.
I will literally upload a video right now for you to see.
what is the best kind of 18650 battery without blowing up this LED?.
panny 18650B 3400 6.6A? 18650 3500 10A?.
thanks
That looks like a lens light. It rotates with a reflector.
I bring the popcorn.
This is a nitecore R40.
https://flashlight.nitecore.com/product/r40
It should be obvious from the beam pattern……
Seriously, go try it yourself if you don’t believe me.
This is just simple optics.
Same for lenses and recoil reflectors.
It inverts the image (left side of LED is projected on right side, etc) but nothing rotates.
You’re right that’s obviously a reflector beam. But that image of the LED you are referring to is not from the reflector; that is imaged directly from the LED. It’s completely independent of the reflector. It’s like the spill of the flashlight, not the beam. Move the pinhole near the outside edge of the reflector and you will see the images that actually make up the beam. You don’t really see it in your video because you only move the pinhole around close to the center where there is no reflecting surface.
The pinhole was moved from one edge of the reflector all the way to the other edge. It’s only a 36mm diameter reflector.
You can see the hotspot leaking from around the piece of paper near the end of the video, and you can see how the radius of that spot is about the same as the distance from the square to the center.
If it was direct light from the LED itself the square would be moving outside of that radius because it would be exiting at an angle.
Also the brightness of the square would be much dimmer, like the spill brightness, not anywhere close to the brightness of the hotspot.
Seriously, just try it yourself.
I don’t know why you have such a hard time believing it.
Well this is like one of those divisive issues where each party is convinced he’s right despite the other’s argument. Luckily it’s one of those situations where it’s not a matter of opinion and only one of us is in fact right. I’ll give it to you that you are very sure of yourself even when you are wrong. I don’t know if that’s a good or bad thing, but it might serve you well in life.
I don’t think you did actually move the pinhole across the whole reflector. Try it again farther away from the wall and you will see the two different images of the LED that make up the beam and the spill. The image you are talking about that didn’t rotate is just an image from the LED formed by the pinhole camera. When you are farther from the wall you will see that it is in fact the spill and not the beam.
Check out this classic thread. This is where I learned of the pinhole demonstration and there are some nice pics of some demonstrations.
Edit: I admit I don’t really understand the details of why the image rotates. While moving the pinhole 90 degrees (like from the north side to the east side of the reflector) the image of the die rotates 180 degrees, which is not what I expected. On the bright side, this is some more “simple optics” we can learn. 
Well shit, you were right.
I was looking at the light directly from the LED and not from the reflector.
Moving farther away made it clear that there were two die projections, one from the LED and one from the reflector.
And the one from the reflector does indeed rotate.
My bad!
Weirdly enough, it does not happen with my syniosbeam.
Even though it is still a reflector and simply the other half of a parabola, the LED is always right side up.