Zebralight H51w is too dim. Is the H51 cool white better for trail running?

[bbb74]

This isn't particularly scientific but I compared the brightness of my sc51w (172 lumens claimed), against a sc51 (200 lumens claimed), a fenix LD25 R4 neutral white (180 ansi lumens claimed), a Quark AA2 R4 neutral white (190 lumens claimed), a H502c (140 lumens claimed), and a Quark AA2 R5 turbo cool white (206 lumens claimed). I just used a digital SLR light meter so with 1/3rd stop increments its not a very fine measurement. I bounced the light off a matte white wall (trying to keep the beam about the same size on the wall by adjusting distance) and then metered an opposite wall 4 meters away.

Anyway ...

The SC51w with imedion is the baseline exposure, and the others are relative to that.

0 - SC51w imedion battery
0 - H502c eneloop
+0.33 - SC51w eneloop (with the occasional flicker up to +0.66 stops)
+0.66 - SC51 eneloop (note this is the cool version)
+1 - LD25
+1 - Quark AA2 nw
+1 - Quark AA2 turbo cw

Visually the ld25 and 4sevens lights were definitely brighter.

Are there any reviews by the same reviewer that have put the SC51 and/or SC51w up against a common fenix or 4sevens light for comparison?

[srfreddy]

This isn't particularly scientific but I compared the brightness of my sc51w (172 lumens claimed), against a sc51 (200 lumens claimed), a fenix LD25 R4 neutral white (180 ansi lumens claimed), a Quark AA2 R4 neutral white (190 lumens claimed), a H502c (140 lumens claimed), and a Quark AA2 R5 turbo cool white (206 lumens claimed). I just used a digital SLR light meter so with 1/3rd stop increments its not a very fine measurement. I bounced the light off a matte white wall (trying to keep the beam about the same size on the wall by adjusting distance) and then metered an opposite wall 4 meters away.

Anyway ...

The SC51w with imedion is the baseline exposure, and the others are relative to that.

0 - SC51w imedion battery
0 - H502c eneloop
+0.33 - SC51w eneloop (with the occasional flicker up to +0.66 stops)
+0.66 - SC51 eneloop (note this is the cool version)
+1 - LD25
+1 - Quark AA2 nw
+1 - Quark AA2 turbo cw

Visually the ld25 and 4sevens lights were definitely brighter.

Are there any reviews by the same reviewer that have put the SC51 and/or SC51w up against a common fenix or 4sevens light for comparison?

From selfbuilt's reviews:
Quark AA ceiling bounce: 4.6
Zebralight SC51: 10.2
Quark AA^2: 8.0 (This is for the old XRE model)

[moldyoldy]

This isn't particularly scientific but I compared the brightness of my sc51w (172 lumens claimed), against a sc51 (200 lumens claimed), a fenix LD25 R4 neutral white (180 ansi lumens claimed), a Quark AA2 R4 neutral white (190 lumens claimed), a H502c (140 lumens claimed), and a Quark AA2 R5 turbo cool white (206 lumens claimed). I just used a digital SLR light meter so with 1/3rd stop increments its not a very fine measurement. I bounced the light off a matte white wall (trying to keep the beam about the same size on the wall by adjusting distance) and then metered an opposite wall 4 meters away.

Anyway ...

The SC51w with imedion is the baseline exposure, and the others are relative to that.

0 - SC51w imedion battery
0 - H502c eneloop
+0.33 - SC51w eneloop (with the occasional flicker up to +0.66 stops)
+0.66 - SC51 eneloop (note this is the cool version)
+1 - LD25
+1 - Quark AA2 nw
+1 - Quark AA2 turbo cw

Visually the ld25 and 4sevens lights were definitely brighter.

Are there any reviews by the same reviewer that have put the SC51 and/or SC51w up against a common fenix or 4sevens light for comparison?

bb74: Do you recall which mode the SLR light meter was in - matrix, center-weighted, spot? or whichever other modes are available?

[Mr Floppy]

0 - SC51w imedion battery
+0.33 - SC51w eneloop (with the occasional flicker up to +0.66 stops)

Interesting, which Imedion? I know some of my 2400mAh ones have quite a bit of a voltage sag around 2A.

[bbb74]

bb74: Do you recall which mode the SLR light meter was in - matrix, center-weighted, spot? or whichever other modes are available?

Of course I used centre weighted aimed at the same point on the wall, but then went waved it around a bit to see whether I could get the exposure to step down or up 1/3rd of a stop. I tried evaluative too, and that gave basically the same results as centre weighted. Spot I felt was too small an area to sample. Note I also had the lens in manual focus and had everything thrown way out of focus.

[moldyoldy]

Interesting, which Imedion? I know some of my 2400mAh ones have quite a bit of a voltage sag around 2A.

My measurements on a 10A range for several of the H51 series provided values easily in the 2.5A+ range. I recall the highest current was ~2.7A. All currents were measured on cells just off the charger and swapped out after a few minutes. However my contacts to the battery & casing have been difficult to maintain, so I do not know what the real current draw would be.

FWIW, I had more problems with high-currents when using the earlier 2100mah Imedion (2000mah min) than with the 2400 mah (2250mah min) in Zebralights. The Fenix MC-10 also pulled over 2.4A+ (now discontinued). Also, I have had current-draw problems with Eneloops as well. Not sure if they were the 1500 cycle version or earlier. nevertheless, modern lights or photographic flash units running AA cells are a challenge for any AA cell design. High current or most cycles or slowest self discharge. Balancing all of those characteristics is difficult.

[moldyoldy]

Of course I used centre weighted aimed at the same point on the wall, but then went waved it around a bit to see whether I could get the exposure to step down or up 1/3rd of a stop. I tried evaluative too, and that gave basically the same results as centre weighted. Spot I felt was too small an area to sample. Note I also had the lens in manual focus and had everything thrown way out of focus.

Good choice and verification! BTW, do you recall how much ambient light was in the room at the time? ie: roughly how many f-stops below the flashlight spot light was the ambient light? sorry for the questions.

[Mr Floppy]

FWIW, I had more problems with high-currents when using the earlier 2100mah Imedion (2000mah min) than with the 2400 mah (2250mah min) in Zebralights. The Fenix MC-10 also pulled over 2.4A+ (now discontinued). Also, I have had current-draw problems with Eneloops as well. Not sure if they were the 1500 cycle version or earlier.

Yes, I've had problems with the MC-10 too and certain AA batteries. It's the one with the Osram LED and it's a bit of a battery killer. Cleaning contacts on it seems to help but it still killed a couple NiZn's and 2 Sanyo 2700 (all with high IR > 2.10V afterwards).

[beamis]

0 - SC51w imedion battery
0 - H502c eneloop
+0.33 - SC51w eneloop (with the occasional flicker up to +0.66 stops)
+0.66 - SC51 eneloop (note this is the cool version)
+1 - LD25
+1 - Quark AA2 nw
+1 - Quark AA2 turbo cw

Visually the ld25 and 4sevens lights were definitely brighter.

Nice test. I think there are two particularly interesting things in this thread: 1) That human eyes, digital camera sensors, camera light meters, and a lux meter all read the the Zebralights as being significantly dimmer than other lights of similar ratings from other manufacturers, and 2) that neutral white lights from other manufacturers do not seem to have the same issue.

While I can't say it's definitive proof that the Zebralight lumen ratings on their neutral white models are a little "optimistic," I think that there's enough subjective and objective observation confirming the initial impressions of many who feel this to be the case to warrant further scrutiny. It looks like the definitive proof is going to have to come from a shootout between some lights in a sphere. Given that I don't have a sphere or many lights, I don't know if this is something I can realistically do. I do have a light meter at the office, but I probably can't take it home.

[bbb74]

Good choice and verification! BTW, do you recall how much ambient light was in the room at the time? ie: roughly how many f-stops below the flashlight spot light was the ambient light? sorry for the questions.

Not much. When I switched the torch off, the exposure indicator went "off scale", which was -3 stops (I had the camera in M mode, normalised so that a correct exposure of +-0 was for the first test and everything else was relative to that).

[bbb74]

Interesting, which Imedion? I know some of my 2400mAh ones have quite a bit of a voltage sag around 2A.

It was a 2400. Its not new (neither was the Eneloop) though. Generally on the C9000 my Imedions always have a noticeably lower voltage during discharges.

[moldyoldy]

back to the question of spectral colors: These lights do not represent the same beam spectra which probably affects our perception of the beam intensity. Additionally there appears to be a discrepancy between manufacturers in the terms "warm" or "neutral", as evinced by a casual survey of beamshots in the various reviews posted on CPF, or some surfing on the Internet.

Example: Zebralight rates it's H51w as "warm" and 4200 Kelvin. Fenix rates it's LD40 as "neutral" and about 4000 Kelvin (from the parent Fenix website). I have owned an SC51w and an H51w and find that the beam color is consistent, but far too "warm" for my preference. On the other hand, the single copy of the LD40 that I had (before my nephew absconded with it) seemed no where near as "colored", meaning excessively "warm" as the Zebralight definition of "warm". I could almost like the beam color on the LD40. That represents nearly no relationship to the rated Kelvins of the beam color except that they are not "cool white".

caveat: I have a strong aversion to "warm" or "neutral" colored beams. Such colors remind me far too much of the light from a Maglite with dying batteries - which is what I grew up with. I strongly prefer a cool white beam color. Any light with a neutral or warm beam represents a very narrow niche usage for me, hence I own only one of them.

The point for this thread is related to the listed testing methods to date, none of which take in to account the spectral color of the light, including the Mark I eyeball. All of the sensors described in this thread so far have a spectral sensitivity curve which may or may not have a limited correction back to - which Kelvin color? Unknown. For ad-hoc testing, they are OK to better identify questions regarding low-performing lights. My point is that if the light beam emanating from a tool has an inadequate strength for your purposes, change the tool. That might be via an RMA, or changing tools or manufacturers.

[bbb74]

Moldyoldy. I agree that different tints/spectra may affect our perception of brightness. However these lights have lumen ratings that we are comparing, and lumen measurements take into account the light's spectra. If light "A" and "B" both have 200 lumens, they should appear pretty similar in brightness as the lumen measurement already takes into account most of the spectral difference between them.

[Mr Floppy]

If light "A" and "B" both have 200 lumens, they should appear pretty similar in brightness as the lumen measurement already takes into account most of the spectral difference between them.

I believe that to account for the spectral difference, the light wave either side of the light sensitivity curve of the human eye are weighted up. Thats if the lumens is a photometric reading (i.e based on the visible light). I believe that's the basis for the light meters. It's probably different in the lab, so if a lot of the light of the cree xp-g warm white falls mainly to the red, it is given a relative lumens in the lab, which probably measures the power of the radiant flux, apply the Judd-Voss correction, etc. I don't know, not an expert in the slightest and may have read things wrong.

[moldyoldy]

Moldyoldy. I agree that different tints/spectra may affect our perception of brightness. However these lights have lumen ratings that we are comparing, and lumen measurements take into account the light's spectra. If light "A" and "B" both have 200 lumens, they should appear pretty similar in brightness as the lumen measurement already takes into account most of the spectral difference between them.

As a flashaholic, I agree with you! absolutely!

However my primary military experience with the signal spectrum, some of my former engineering work (as an EE) where I ran an electrical and mechanical calibration lab in a military/industrial plant, and my experience as a serious "hobbyist" photographer (2000+ rolls/yr) attempting to go pro/independent where I worked independently as a sports and wedding photographer for many years, suggests otherwise. In those endeavours I discovered that standards in some areas were very tight, in other areas there were almost no standards except the basic laws of physics. IOW, Any attempt to compare at our equipment level was doomed to failure. Hence we individually or collectively decided on a "standard" and evaluated accordingly. But we knew that we possibly were in "never-never land".

IOW, given the current lack of knowledge of accurate emission spectrums from LEDs and how those emissions were measured, decide on whatever you want with whatever tools you have available, but know that comparisons are fraught with pitfalls and traps.

a moldyoldy...

[moldyoldy]

I believe that to account for the spectral difference, the light wave either side of the light sensitivity curve of the human eye are weighted up. Thats if the lumens is a photometric reading (i.e based on the visible light). I believe that's the basis for the light meters. It's probably different in the lab, so if a lot of the light of the cree xp-g warm white falls mainly to the red, it is given a relative lumens in the lab, which probably measures the power of the radiant flux, apply the Judd-Voss correction, etc. I don't know, not an expert in the slightest and may have read things wrong.

a good thought process. However for an example of a condition where that thought process has difficulty, go in to any large old industrial plant that uses sodium vapor lamps for lighting. These lamps have almost a single-line spectra and drive light metering systems crazy. Why? Because most light meters are expecting a relatively broad spectrum/spectra of incoming light. Hence the reason why GE developed their multi-vapor lights to correct this design deficiency. However some locales near observatories or airline flight paths will still used sodium vapor street lights because the perceived illumination is much lower, both to human eyes or to instruments.

[beamis]

a good thought process. However for an example of a condition where that thought process has difficulty, go in to any large old industrial plant that uses sodium vapor lamps for lighting. These lamps have almost a single-line spectra and drive light metering systems crazy. Why? Because most light meters are expecting a relatively broad spectrum/spectra of incoming light. Hence the reason why GE developed their multi-vapor lights to correct this design deficiency. However some locales near observatories or airline flight paths will still used sodium vapor street lights because the perceived illumination is much lower, both to human eyes or to instruments.

Please forgive me if it comes across the wrong way, but are you suggesting that even when a light appears dimmer than another to the naked eye, reads dimmer in camera light meters, photographs dimmer with CMOS sensors, and reads dimmer using a lux meter, it's still possible that our eyes and instruments deceive us and the light is in fact equally bright?

[moldyoldy]

Please forgive me if it comes across the wrong way, but are you suggesting that even when a light appears dimmer than another to the naked eye, reads dimmer in camera light meters, photographs dimmer with CMOS sensors, and reads dimmer using a lux meter, it's still possible that our eyes and instruments deceive us and the light is in fact equally bright?

nope, nothing like that. Only that it is often difficult to quantify the "dimmer" difference. Comparisons w/in a brand lineup are less troublesome. Comparisons between brands become more difficult because of the obviously different standards used. eg: my example above comparing Kelvin numbers and my perception of color between Zebralight and Fenix for a non-cool-white beam color.

Ultimately the user decides whether a flashlight provides sufficient illumination for the tasks at hand, not the specs or the manufacturer. If a flashlight fails to meet your needs, change it!

[moldyoldy]

FYI: There is a very informative series of posts by UnknownVT on the subject of LED beam colors and human perceptions in this CPF thread:

http://www.candlepowerforums.com/vb/...parison-Review

The thread may refer to obsolete flashlights, but the color comparison information is very useful. Just about every post by UnknownVT provides considerable detailed information. Note his use of color desaturation to determine brightness comparisons. Also, the links in posts #5, #7, #17, #19, #29, #36, and #39 are so useful that I bookmarked that entire thread in my browser.

My complements to UnknownVT!

[Mr Floppy]

I have the newer H51w now so I can run some readings. First of all, I have a cheap light meter. Secondly, the light box I built recently is having its first run. It is a 30cm cube (a foot). Unfortunately it is a bit too shiny inside so it does affect the reading depending on where the hotspot is aimed. The H51 was hard to mount as the hole was designed for your regular lights, not angle ones. I had to aim the hot spot as well as I could at the same place.

So, my readings are 77 for the H51w and 94 for the Fenix LD10 R4. Using selfbuilts figues here, here is comparitive result. His LD10 R4 had a lightbox reading of 73, compared to mine of 94. So with this, a comparitive reading is 59 on selfbuilts figures. Which is a bit better than the Nitecore D10 Q5 cool, which has a stated lumens of about 130.

Now another caveat, The LD10 R4 reading I think was a bit too far out. If my readings translate to selfbuilts figures, then LD20 R4 should give me a reading of 104. Well it didn't. It was much lower at 99. I've previously done the same test with a ceiling bounce and it was much closer to selfbuilts figures so I'll have to go back to that I think until I calibrate the light box against selfbuilts figures, (move all the stuff out of the toilet ...)

Anyway, to my naked eye, it looks quite bright, much brighter than the TK20. What I think could throw people off is the hot spot. The TK20 hotspot was visible when shone directly into the middle of the H51w hot spot but it is throwy. It looked as bright as my L2D Q5, which has a comparable hot spot, maybe just a bit smaller. That is rated at 180 lumens and in my light box, it gave a figure close to that of the H51w at 79-81 but at that stage, the holder started to move and I had to hold it with my hand.

Current draw at the tail cap was 2.2A if curious. Tried a few batteries and only the Eneloop managed to get H1 mode.