So basically, you prefer throw instead of flood? Are there any that you feel are the best as a smooth flood instead of throw?
Also it seems like the larger diameter reflectors do better overall. Is there any real test data that suggests the "optimum reflector size" for an XM-L. What I mean to say is that maybe a small reflector is not utilizing the output of a well driven XM-L, because it's too small a surface area to do it justice?
I think as a general rule, the bigger the LED die, the bigger reflector you need to wrangle all its light and focus it somehow. Bigger, and especially wider reflectors produce a more concentrated hotspot that throws better (more lux at the center). Compared to reflectors of equal diameter, deeper reflectors produce a narrower, more concentrated (i.e. brighter) spill. This is especially evident when the depth exceeds the width, like in the C9-T60/HS-802 and the C308.
For me, the cutoff size for a single XM-L in a light is the reflector from the UF C308, which has an inner diameter of 29mm and a depth of 41mm. I've never seen an XM-L in a conventional 26mm P60 sized drop-in, which is considerably smaller. I can imagine based on how perfectly mated they appear to be for the much smaller XP-G LED, that I would consider the output of pairing one with an XM-L as having a hotspot that's too soft and squishy (large and lacking in intensity).
IMO, the best flooders are multiple emitter lights using small optics, like you've seen in some of my Mag mods. For a single LED flooder, I suppose a large wide-angle optic would perform the best.
I've made a couple XM-L Mags, two with incandescent reflectors (tint bins 3C and 5A3) and two with Mag Rebel reflectors (3C and 7C-ish). I'm sure you've taken note that an XM-L has sort-of a tint gradient, so that light radiated at larger angles with respect to the normal being noticeably warmer tinted than the rest. This seems responsible, to some extent, for a warm hotspot compared to spill in the incan reflector.
Paired with the incan reflector, the 3C has a flat yellow [French's mustard but more pale] hotspot and cold spill; in-all, pretty good hotspot tint for outdoors, but not hotwire/Planckian black-body yellow. Casual users/normal people probably not notice anything and be happy with it. At optimum focus the corona looks like a compass rose.
With Mag's LED reflector, the 3C became a entirely different [very much improved] tint. Spill and hotspot tint are considerably more alike. I'd characterize the hotspot as very...um...natural looking white. Imagine how the sun beats down at 9am on a cloudless summer solstice*. I think it may be my favorite tint. *(clouds would mix in more of the blue normally lost to the sky-->higher CCT).
Below: XM-L T4, 5A3 tint bin, ~500-600mA. Mouse over for Rebel reflector. Note that for 2nd shot I had to support the position of the head containing the deep reflector, so focus is not optimum. The tints came out a little exaggerated but it illustrates what I describe above-to my eyes, a 5A3 XM-L in the Rebel evokes impossibly overdriven tungsten
So, could you say that a reflector's depth/diameter ratio has a bearing on this phenomenon, with deeper ones mixing the tint gradient 'better' (<--subjective) than reflectors that lean toward the wider side? What have you observed with all the reflectors you've tested?
Thanks, I look forward to others' observations.
edit: My apologies Techjunkie, this is sort of veering off your original topic.
I concur with your observations. Every time I've tried a shallow reflector with neutral white XMLs, the result was a horrible yellow hotspot. Gotta go deep to blend some of that cooler spill into the spot.
The drop-in is really wedged in there, both at the top collar of the reflector just before the lip, and also on the angled part of the reflector, which sits either on the ledge inside the head, or if I went too aggressively with the dremel, then on the top of the Mag neck which comes up to meet it when the head is turned down as far as it can go. The driver isn't too ambitious and only drives the emitter at an estimated 2.7A. So far, I've run the light on high for 10 min and noticed the head get warm, but not hot, and no visible difference in output between start and finish.
Thank you for taking the time to clarify that TJ. It never occured to me that those batteries might sag very little with the given load. And I surely see that there is little to be won driving it harder. But them again my nature demands it to be driven harder (see name) . Not my logic or lack thereof
