Stochiastic Vs Smooth reflectors in LED lights

[kakster]

In my never ending quest for the perfect beam, a thought occured to me whilst considering photon management. Now we all know that a stochiastic (orange peel) reflector surface is required to coax a smooth, even beam from an incandecent filament. Im just wondering how necessary this is for a Luxeon light source, because the light from an LED should be far more homogenous and even (guess on my part..am i wrong here?). Even if the LED light source is not totally even, it *should* be less chaotic than a heated piece of curly wire. This would mean less of an aggresive orange peel surface is needed, and would result more photons being fired out the front of the bezel.
However, looking at Surefire's latest KL bezel reflectors leads me to believe that im wrong on this, but i had to ask anyway.

 

[Otokoyama]

Once we get LEDs with perfectly spherical, perfectly centered sources, and mount them in perfectly centered, perfectly focused reflectors or optics, the "orange peel" may become a thing of the past.

 

[kakster]

So thats a "no" then /ubbthreads/images/graemlins/smile.gif
It would still be interesting to see how much more of an even light source a LED is compared to an incandecent bulb. It seems to me that a more aggressive orange peel will induce greater losses in output than a less rippled surface. I am assuming that the amount of stippling needed to smooth out a beam is proportional to the eveness of the lightsource. Of course, theres every chance that this line of reasoning is total rubbish, this is just a hobby for me.
I know im beating a dead horse, but im looking to see if theres a way to eek out every last lumen from a LED.

 

[Otokoyama]

I don't think of the orange peel as "losing" output. It simply has a different distribution, and specifically doesn't attempt to put every photon into the hotspot.

Come to think of it, some people (like myself) who don't need the ultimate in throw might actually prefer the gradually falling off beam of an orange peel even if everything else in the beam production mechanism was perfect.

In my original post I forgot to also state "perfectly uniform" since a nice reflector will project a nice barbeque grill with some LEDs.

 

[PeterB]

My experience with Hotbeams Mr-X (5W LED X3T): The beam is already very good with the smooth MAG reflector (way better than a standard MAG). But there is still a darker region in the middle. But with Otokoyama's orange peel reflector it is simply perfect!
If you are interested I can post a beam shot comparison.

 

[rodfran]

Hello PeterB,

I would be interested in seeing that beamshot comparison. I have been following the Mr-X posts and the Otokoyama posts with interest. I have one of the orange peel reflectors and it is a fine piece of work.

 

[Kiessling]

when I compare 1-Watters in a Pelicam Reflector, I get a beam of a very much higher quality with the older style having a more aggressive orange peel surface than with the newer, less aggressive orange peel reflectors. Apparently the Luxeons are not that even, and the reflectors ain't either.
bernhard

 

[PeterB]

Hello Rodfran,

sorry for the late answer!
This is Mr-X with the smooth reflector

And here with the orange peel reflector.

 

[CM]

I think the smooth non-stochastic reflector is better for LED's. I'm speaking from experience here since I've seen a reflector similar to the KL4's but smooth like the M@g. It showed absolutely no artifacts. That's both with a 1W and a 5W LED. The picture above confirms this but mine eyes like the smooth reflector better since I believe it would result in a bit more efficiency. The orange peel is necessary because the filament on an incandescent will absolutely show up if the beam was not randomized in some fashion. Randomizing does result in a loss in efficiency.

CM