I'm pretty sure this is not true. If the light hits the reflector before hitting the aspheric, it is still transmitted through the lens, but it ends up being spill and does not get collimated into the main beam and enhancing the throw/lux of the hotspot.
I want to know about those tests that MRGMan did. I'm guessing it was aspheric only versus reflector only. Why not test an aspheric WITH a reflector? I have two aspheric lights with the reflectors in place and I did not notice a large drop in output - only a reduction in spill for an increase in throw. My aspheric MC-E still has tons of spill because of light hitting the reflector before it gets to the back of the lens.
Correctomundo, give that man a star. Most of the flashlights that have an after market aspheric dropped in have the reflector removed because they don't both fit. If you put the aspheric in front of the reflector its too far away from the LED die and doesn't focus.
the problem is not that the aspheric loses light (its a transparent solid typically loses 4% per surface unless it has AR coatings of whatever light actually reaches it), the problem is the light that doesn't get directed to the aspheric because there is no reflector to collect it and send it all forward from an LED whose output is not going straight out the front.
Most of the light that is lost is because its not directed to and collected by the aspheric system in the first place. The DEFT gets major points on this one because its a 2 lense system. The first smaller lens very close to the die collects a lot of the light to send to the large aspheric lense rather than losing it but still some gets lost.
In the systems I looked at and measured that simply had an aspheric replacing the reflector all of the light that would have bounced off the reflector into spill got lost so typically there was a 40% loss of light.
This was not light that got lost from the aspheric but never got to the aspheric to go through it in the first place. That is why the best lens systems for a LED flashlight is not an aspheric in the first place but a Total Internal Reflection optic that mounts directly over the die and collects all the light from all angles off the surface and sends it through the optic out the front and only the normal air gap to lens surface reflection % is lost.
Inova and Surefire both have or had TIR optics that worked well in collecting and transmitting lights, neither were designed with an ultra tight beam pattern such as the DEFT, but they could and would be far more efficient.
Those lights that use a reflector of some sort to collect light and send it to the aspheric have brighter spill and not so great of a loss of light but it can't all be collimated into the tight beam as you said.
The only other issues with aspherics are the chromatic aberrations as was already stated. And as was already said to do this right requires big bucks for high quality low dispersion glass and A/R coating on top of that wouldn't hurt but simply add more $$.
I am surprised that no one has ripped the TIR lens out of an old Inova T5 and put it in front of a Cree CR-E R2 to see what that looks like.
