[selfbuilt]

4Sevens Quark RGB Review: RUNTIMES, BEAMSHOTS and more!


Reviewer's Note: The Quark RGB was provided for review by 4Sevens.com. Please see their website for more info.

Warning: pic heavy, as usual.

Manufacturer's specifications, condensed from 4Sevens’ website:

• Features a CREE MCE-RGB emitter. The MCE-RGB is a quad-die emitter that features a different color for each corner of the die: white, red, green, and blue
• Available in cool white or neutral-white for the white tint
• UI: When the head is tightened, it is always White, when the head is loosened, it is one of the RGB colors. Cycling between loose and tight will toggle through the three colors.
• Change modes by taping the tail button, just like Quark series (eight modes: moonlight -> low -> medium -> high -> max -> S.O.S -> strobe -> beacon).
• Remember which mode you used in both the tightened and loosened state as well as which color was used in the loosened mode.
• Length: 4.8 in, Diameter: 0.86 in, Weight: 1.8 oz (without batteries)
• Finish: Type-III hard-anodized aircraft-grade aluminum
• Battery Type: 2 CR123A
• OTF Lumens*]Moonlight: 0.4 lm, 650 hours, 1 ma
• Low: 2.8 lm, 130 hours, 10 ma
• Med: 15.0 lm, 25 hours, 50 ma
• High: 58.4 lm, 7.5 hours, 250 ma
• Turbo: 150 lm, 2 hours, 700 ma
• SOS: 22.5 hours
• Strobe: 4 hours
• Beacon: 20 hours
• MSRP ~$91 with CPF discount, also available as just the head (cool white version only) for ~$64 with discount.

The Quark RGB is a special-purpose light. Built on the standard Quark Q123-2 frame, it features a new head with the Cree RGB MC-E emitter (i.e. four separate colour dies: white, red, green, blue). This has required some modification to the UI, but otherwise will be familiar to Quark users. Please see my earlier Quark series review to find out how the basic 4Sevens Quarks compare.



The Quark RGB comes in a similar package to the other Quark series lights. Included with the light is a removable/reversible clip (attached), good quality wrist lanyard, extra o-rings, good quality belt holster, rubber hand-grip, 2 primary CR123A battery (4Sevens brand) and manual.




From left to right: Surefire CR123A, AW Protected 18650, Quark RGB, Quark 123-2.

Quark RGB: Weight 50.5g, Length 123.5mm x Width 22.0mm (bezel)
Quark 123-2: Weight: 47.0g, Length 113.4mm x Width: 22.0mm (bezel)

Overall dimensions are similar (and in fact the battery/body tube and tailcap are identical) - the only difference is the RGB’s head is about a centimeter taller. For all intents and purposes, the hand feel is very much the same.




Again, since the only difference is in the head, I refer you to my 4Sevens Quark series review for a discussion of the general build features of this light.

One feature that has changed from my original review is the tailcap now comes with anodized screw threads, allowing for tailcap lockout. This is standard on all Quark series lights now.




As you can see, the Cree RGB MC-E emitter is very distinctive! I apologize for the reflections in the picture obscuring the red die (top left corner) - in reality it looks the same as the other three. In case you were curious, the upper right die is the green one, and the lower right die is the blue (as expected, the lower left die is the white one).

Something else to note here is the extremely shallow reflector. Coupled with the fact that each individual die is not centered within the reflector, I would expect to see a very distorted “floody” beam pattern for each color.

And now for the requisite white wall hunting … each light is on Turbo on AW 17460, about 0.5 meters from a white wall.





Do NOT get bent out of shape about the distorted beam pattern above! I had to do these beam shots really close to the wall, in order to capture everything in the camera frame. In real life, I do not expect you'll be using full flood at 0.5 m very often!

This will give you a much better idea of what to expect - taken at 1.5 m from the same wall.



While the distortions around the center beam are still detectable, they are not really obtrusive at this distance. And of course, they become even less noticeable at greater distances.

To compare the color beam profiles, I have gone with an intermediate distance (a little under 1 m). Again, focus on the relative color and output, not the beam distortions.






The camera has captured what I *seem* to see by eye fairly well - but for a proper discussion of relative outputs, scroll down to the testing section of this review.

One thing general point is that this is MUCH better than what you would see with a collection of colored filters over a white LED light. White LEDs are particularly deficient in the red and green wavelengths of the color spectrum.

User Interface

The user interface of the Quark RGB had to be adapted somewhat from the standard Quark series.

The head tightened/head loosened switch has now been co-opted for the various color modes. With the head tight, you get white light. With the head loose, you get one of the color modes. Each of the three RGB color modes are accessed in sequence by doing a tighten/loosen switch of the head.

So, for an example, let's say right out of the box you want green. Starting from white (head tightend), you would do a loosen-tighten-loosen switch (i.e. go from white to red to white to green). The light will remember what color you are in when you turn it off and back on (i.e. with the head remaining loosened). But if you go back to white (head tightened) at any point when on, the next time you loosen the head you will advance to the next color (i.e. blue in this case).

In practice, this is not as confusing as it may sound - you quickly get used to it, but just realized you always will be cycling back and forth through white to change the color.

The various output modes are now all accessed in sequence from a tailcap press. The sequence is Moonlight > Lo > Med > Hi > Turbo> SOS> Strobe > Beacon. There is a memory mode, so if you turn off the light, it will always come back on in the mode you last left it.

There are in fact two memory modes - one for the head tightened (i.e. white) state and one for the head loosened (i.e. color). So, for example, you could memorize the light in Turbo for white, and Moonlight for color. Note that you cannot assign different memory states for each color - the memory mode applies equally to all three colors (i.e. head loosened).



Strobe is a fairly tactical (and certainly annoying) 12.6 Hz.

Testing Method: All my output numbers are relative for my home-made light box setup, a la Quickbeam's flashlightreviews.com method. You can directly compare all my relative output values from different reviews - i.e. an output value of "10" in one graph is the same as "10" in another. All runtimes are done under a cooling fan, except for the extended run Lo/Min modes (i.e. >12 hours) which are done without cooling.

Throw values are the square-root of lux measurements taken at 1 meter from the lens, using a light meter.

Throw/Output Summary Chart:



As expected, the throw on this light is greatly reduced.

But what should we make of the differing output levels of the colored emitters? Note in the table above that both my home-made lightbox sensor and my stand-alone Lux meter (used for ceiling bounces) report that there is a slight reduction in green output compared to white, with a more noticeable reduction in red output, and a huge reduction in blue light output.

At first I thought this might be due to a varying sensitivity of my sensor setups for the various wavelengths. Well, if it is, then Cree has the same issue - here are the Luminous Flux estimate @350mA for the from the Cree spec sheet for the MC-E RGB emitter.

Red: 30.6 lm
Green: 67.2 lm
Blue: 8.2 lm
White: 95 lm (80 lm for the “Neutral White” version)

In terms of relative differences, that matches pretty closely with what my ceiling bounce measures are telling me on Turbo. I'm not sure why the luminous flux is so much lower on the blue die when driven at the same current - perhaps some of the experts on these matters could chime in here.

Again, let me emphasize that visually, this is a LOT better than what you would see with colored filters over a white LED. I just tested the light with the Fenix red filter over the white LED, and got less than a third of the output of the dedicated red LED.

Output/Runtime Comparison:

Note: Effective January 2010, all CR123A runtimes are now performed solely on Titanium Innovations batteries sponsored by BatteryJunction.com. You can compare the generally excellent performance of these CR123A cells relative to the Duracell/Surefire cells used in all my earlier reviews here. I have marked all the new runtimes of lights with Titanium Innovations CR123As on the graphs with an "*".

So how do the individual colors compare for runtime? Below is the AW 17670 run:



Despite the varying output levels detected by the lightbox, you'll note the very similar overall runtime for each of the four individual emitters on Turbo (i.e. they do indeed seem to be driven to comparable levels).

Interestingly, the red LED runtime pattern on 3.7V Li-ion is a little different - very tight regulation followed by a quick drop off into a slow strobing mode (instead of into a low moon mode). I note in the Cree spec sheet that the forward voltage is much lower for the red LED compared to the other three, but I don't know if this has anything to do with it. I was able to manually switch the light into the Moonlight and Lo constant output modes in red at this point, so you are not stuck strobing when the battery is almost dead.



Runtime patterns and output levels for the White LED look as I would have expected, on all battery types.

Here's how it compares to the single-die competition:










Again, no surprises here – for all intents and purposes, the RGB’s white emitter performs the same as the standard R2 emitter in the Quark Q123-2.

Potential Issues

Due to the shallow reflector and inability to center the individual dies, the Quark RGB produces is a very floody beam with noticeable distortions up close. At reasonable distances however (>1 m), these are generally negligible, IMO.

Luminous flux (i.e. output) of the various colored dies is not equivalent. The blue LED in particular is relatively low output compared to the others. This is a feature of the Cree MC-E RGB emitter however, and is not the fault of the Quark. In any case, the individual outputs are all considerably greater than what you would get with colored filter on a white LED.

The UI is a little complex, and may not suit all users. However, I can see it was an attempt at a reasonable compromise given the rather unique nature of this emitter.

Preliminary Observations

The Cree MC-E RGB emitter - with its ability to access individual emitters independently - is a rather unique beast, and thus stands in a class all of its own. However, the quirky nature of this emitter platform means that it is virtually impossible to please everyone in a single offering.

Simply put, the colors add an extra dimension of complexity to what is a limited number of control options for a light (i.e. tailcap clicks or head twists to switch modes). Given the general layout of the Quark series, putting the colored modes on the head loosened position makes sense to me (although it is unfortunate that you have to keep switching back to white to change color modes). And the two memory states – one for white and one for all the colors - is also understandable (again, if not ideal for everyone).

But I personally don’t like seeing strobe and SOS in the same sequence as the constant output modes now. Frankly, I don’t see any easy way around this - unless you plan to dispense with strobe/SOS all together.

I should point out that there is no way to run all the emitters concurrently - to keep heat down, you are limited to one die at a time (besides, it would probably look like a pig’s breakfast if you tried! ). But you can drive each of those dies to the max of their ability. In other words, it's like having four separate Cree single-die lights in one body.

But since you cannot focus each die individually, you are left with a general flood light (i.e. too many distortions for anything more focused). I don't see any way around this, unless you designed some sort of complex movable head that allowed you to rotate the reflector over each die in sequence. That would require a major redesign in a much larger light (translation: much more expensive), and likely be prone to breakdown anyway.

But in any case, a floody multi-colored light makes the most sense to me (i.e. you will likely be using it up-close, not illuminating objects at a great distance).

Although on this point, I'm not really sure of the benefit of R/G/B per se. This is not a comment on 4Sevens' light, but rather Cree’s decision to market this particular combination. There are certainly varying opinions about the value of the specific wavelength LEDs, but I would think UV would be more valuable for inspection purposes. How often do you need regular blue? And I would take red over green any day, but I understand there may be applications better suited for green. Anyway, that’s just my little off-topic raving …

One thing we can all agree on is that dedicated color LEDs are a LOT better than using filters on an existing white LED light. Green and red in particular are quite deficient in the white LED spectrum (which is really just made by adding yellow phosphor over a blue LED, after all). If you want a high-power green and red flood light, this MC-E is not a bad way to get it.

At the end of the day, I think 4Sevens has come out with a reasonable attempt to incorporate the Cree MC-E RGB into their existing product line-up. I will leave it to your judgment as to whether this RGB application fits your needs.