High CRI emitter isn't so High CRI if its output spectrum isn't fully collimated

I thought I was going nuts with a new drop-in because colors weren't coming out as expected. Things didn't seem as "neutral" as it could be. So white wall hunting I notice the center of the beam appeared too yellow and the spill was tinted too rose-purple. You know, the kind of tint shifting seen in a million other LED lights.

I switch optics to test something more throwy and you know what? That rose-purple went almost completely away, the yellow was much reduced and the emitter now seemed true to its "neutral" description.

So I did a little research and found this graphic:



I guess this explains a lot about what's wrong with our industry and why the tint headaches many of us go through.

so the same issue that plagued sir Isaac newton

I think if your optics are diverging the emitters output, then reds greens and blues that where balanced enough to spec an emitter as "neutral" (red+green+blue = white) isn't happening. You'll get lopsided color extremes.

For instance if I shine a light that was out whack like my drop-in above, I'll see those unnatural tints on whatever I'm observing. If I "ceiling bounce" that same light under a white sheet of paper, then I'm seeing the totality of that emitters spectrum mixed more thoroughly. Things underneath the bounced light appear more neutral.

Round emitter domes are part of the problem.

I like broad flood reflectors/optics, this often has the handy side effect of keeping the spectrum evenly distributed in the beam.

Other than wall hunting, how does this affect color rendition in the real world, inside or outside. I have two types of Hi Cri lights, one using the Seoul P4 and others using the Nichia. My test medium is a pair of dark brown socks, socks that appear grey using non Hi CRI LED lights, and appear to be very brown using the Seoul and the Nichia. How would a tightly collimated Hi CRI flashlight make any difference in color rendition, compared to a floody Hi CRI, when both lights are driven to the same overall lux?

Bill

Let me get the tripod out. Photos are best.

Camera doesn't pick up the tinted spill as good as I hoped but it's there. Beams where centered along the left edge of the photo.

About 40 lumens here (out of 1,700 )





Animated:



The wider beam shows the rose-purple at its fringe and a yellower hot spot. The tighter beam (tighter collumation) shows an elimination of the rose-purple fringe and the yellow part of the emitters spectrum is more attenuated across the entire diameter of the beam angle, spill and all. This translates to a more balanced neutral beam with less yellow and rose-purple concentrated regions. The rose-purple that was shifted back into the spectral yield makes a bit of a difference in the overall neutrality and "whiteness" of the beam.

I wonder if this pic has any relevance to this:

I'm assuming you are using reflector based optics?

If so then the effect is caused by the fact that widest angle light from the LED has a shifted spectral distribution (often to green or yellow). Whilst the spillbeam consists of light going straight forwards from the emitter and OTF without touching the reflector, the central hotspot is a combination of this plus all of the output at a wide enough angle to hit the reflector and be re-directed to the centre of the beam, as a proportion of that is the widest angle output there will be a slight tint shift in that spot. This is most exaggerated in the corona of the hotspot as that is made up proportionally more greatly of the widest angle output of the LED.

You can observe this by simply looking into the front of the reflector with one eye at different angles with the emitter off or at a moonlight level.

Acrylic lens. Multiple emitter arrays and such.

This is all new to me, but great info. So this can happen with reflector lights too, right (just a clear glass window-no optic)? My Four Sevens X10 is more of a thrower with a deep reflector. It has very noticeable purple shifting in the spill beam. Could this be a collimation issue? Is the deep reflector helping out with throw, but sacrificing on tint shift?

It's really an optics design issue and making sure the optic/reflector is matched correctly to the light source's emission angles. There's lots of complicated math to it, but it happens any time there's a design mismatch, both with throwers and flood lights.

I'd like to see optics that can cancel out any chromatic aberration that occurs when light passes through a concave lens. In this case the emitter dome.



http://en.wikipedia.org/wiki/Chromatic_aberration

Before



After


^^^
I wonder if this principle is happening with the LED because it appears the 10621 Carclo tight optic corrects chromatic distortion better than the 10622 spotlight optic resulting in a more neutral (whiter) beam.

And for anyone interested the beam shots are from a quad XP-G2 [R4-5C] bin.

About 40 lumens.





Similar to halogen



Nichia 219 [SW45B10]





1,100 lumens on the Nichia, but you can still get a sense of the color differences.

Presumably a mule light (with no optics) will provide excellent CRI at close quarters.
I believe Don (McGizmo) first came up with that idea.

I am in over my head a bit here, but would that cause a highly Orange Peel reflector for a given LED to have a better CRI index, then the same LED in a smooth reflector. Due to a more blended (or to better state it; less organized) beam pattern? Or am I reading to much into this?

whiteoakjoe said:

I am in over my head a bit here, but would that cause a highly Orange Peel reflector for a given LED to have a better CRI index, then the same LED in a smooth reflector. Due to a more blended (or to better state it; less organized) beam pattern? Or am I reading to much into this?

Click to expand...

Yes, I would imagine that anything that blends the beam more evenly would achieve this.

easilyled said:

Presumably a mule light (with no optics) will provide excellent CRI at close quarters.
I believe Don (McGizmo) first came up with that idea.

Click to expand...

Even a bare LED (especially one with a high dome like the XR-E) has a tint distribution from side to center. My non-domed (Rebel?) In a mini AAA mag lite has very little.

Here's a classic post from awhile back showing the Cree Q5 with different optics.


http://www.candlepowerforums.com/vb/showthread.php?13090

I'm keen to guess things appear most neutral with the optic on the bottom right.

easilyled said:

Presumably a mule light (with no optics) will provide excellent CRI at close quarters.
I believe Don (McGizmo) first came up with that idea.

Click to expand...

Yes, I believe that is correct. I recall his writing about the development of the Mule but don't see it right off hand. See http://www.candlepowerforums.com/vb/showthread.php?298316-Titanium-Mule-Wave

I don't normally photograph walls, but I do photograph people against walls. Here's an example using a McGizmo Sundrop with 083B emitter. The Sundrop has an aspheric lens with a set focus. I've never seen any other flashlight that delivers such a high quality light for photographing people at night.


D4A_2507.jpg by Dean Fletcher - Photographer - Portland, on Flickr

Rachael is just one example of the colorful nightlife here in Portland. Portlandia has nothin' on the real thing.

DTF said:

Yes, I believe that is correct. I recall his writing about the development of the Mule but don't see it right off hand. See http://www.candlepowerforums.com/vb/showthread.php?298316-Titanium-Mule-Wave

I don't normally photograph walls, but I do photograph people against walls. Here's an example using a McGizmo Sundrop with 083B emitter. The Sundrop has an aspheric lens with a set focus. I've never seen any other flashlight that delivers such a high quality light for photographing people at night.


D4A_2507.jpg by Dean Fletcher - Photographer - Portland, on Flickr

Rachael is just one example of the colorful nightlife here in Portland. Portlandia has nothin' on the real thing.

Click to expand...

Nice picture and no shortage of reds popping out, just to prove the point. When I first saw it, for a few seconds I thought Rachael was a porcelain figurine.