Sunwayman V10A (1xAA, XP-G R5, Continuously-Variable) Review: RUNTIMES, BEAMSHOTS+

Originally posted: February, 2011
Last revised: February 25, 2011

Warning: pic heavy, as usual. :whistle:

Specifications:


This is the first V-series Sunwayman light I have reviewed. Although the build looks very similar to the recent M-series light, the control ring now features a continuously-variable output selection feature (as opposed to the earlier defined levels and detents).

Packaging hasn’t changed much. The light comes in a typical cardboard box with the usual extras - manual, warranty card, promotional insert, good quality wrist strap, pocket clip, extra o-rings and boot cover.


From left to right: Duracell AA, Sunwayman V10A, Xeno E03, Fenix LD10-R4, 4Sevens Quark AA, Crelant 7G1, Zebralight SC51, Sunwayman L10A

V10A:: Weight: 58.1g (no battery), Length 100.6mm x Width 23.1mm (bezel)

The overall dimensions are pretty standard for a typical 1xAA light, if a bit on the tall side.



Build of the V10A remains excellent overall.

Although virtually indistinguishable from the recent M-series lights, the control ring is now smooth over the range (i.e. no detents, as no defined levels). Max range of the ring is unchanged at around 1/3 the total circumference of the light.

The rest of build hasn’t changed. The light can still tailstand (with forward clicky switch). Square-cut screw threads remain anodized for lockout (but still only have a limited number of turns).

The included pocket clip is serviceable, but these sorts of clip-on clips will never be as study or stable as ones that goes around the body tube.

Fit and finish is perfect on my sample, in the classic natural color common to Sunwayman lights. Lettering is bright white and clear. As before, there is no real knurling, so gripability is toward the low side (unless you have the clip attached).

Like the more recent M-series lights, the V10A comes with the Cree XP-G R5 Cool White. The reflector is what I would consider a light texturing/orange peel (LOP).

Which brings us to the requisite white wall hunting ;). All lights are on Hi on Sanyo Eneloop, about ~0.75 meter from a white wall (with the camera ~1.25 meters back from the wall). Automatic white balance on the camera, to minimize tint differences.




As you can see, the beam profile of the V10A is in keeping with its class. Up close, I notice some faint rings in the beam, but these are not so great as to be disturbing (MOP may have helped here).

Scroll down to my Summary Tables and runtimes for more specifics on output.

User Interface

Like with the M-series lights, you turn the light on by the tailcap switch. Half-press the tailcap for momentary-on, click for locked-on

Mode switching is controlled by the magnetic control ring. The V10A features a continuously-variable interface – you control the output level by twisting the ring. You can select your desired mode while the light is off.

And that's it - no blinking modes, no standby modes. :)

Ramping Pattern

EDIT: this section has been updated and revised from my initial post.

The big question is whether the output selection of the V10A is "visually linear". Generally, here on the forums, that has been taken to mean a logarithmic ramp as opposed to an actual linear ramp of outputs. The reason for this is that we perceive brightness in a non-linear way (actually, not just brightness - most of our sensory perceptions are non-linear). This is why twice the lumens doesn't appear twice as bright to us - lumens are an objective (linear) measure of output, and our subjective perceptions are not linear.

A logarithmic adjustment has long been used to adjust for our relative visual perceptions (e.g. the stops of camera are logarithmic). And like most here, up until now I have found that all continuously-variable lights with a logarithmic correction of output generally look "visually-linear". And then the V10 came along. :whistle:

To start, here is what the V10 looks like in my lightbox. To measure this, I slowly turned the ring at as close to a constant rate as I could manage, over ~25 secs or so. My lightbox collected output readings every second, and I then plotted the relative lightbox output against the estimated degree shift of the ring (i.e. with 360 degrees being a complete turn). Note the ring only turns about 1/3 the circumference of the light, or about 120 degrees.

I have blown-up the first third of the ramp in the inset graph, to show you that output does indeed increase over the whole ring (albeit seemingly slowly at first). Now, you can certainly argue that the output trace looks like it could be logarithmic. Indeed, if you plot it on a log scale, you get something approximating a straight line. However, the V10A does NOT subjectively appear to me to be visually-linear when handling (e.g. it does indeed spend a lot of time at the very low outputs over the first third of the ring).

Upon reviewing the scientific literature, I see that relative power relationships have superseded simple logarithmic corrections for linearizing our sensory perceptions. For perceived brightness, the currently accepted linearization method is actually a cube root of output. For a full discussion of this - including detailed graphs and primary literature references - please see my post #28 below.

When plotting with a cube root transformation of my lightbox's output scale, you get the following graph:

This graph MUCH better matches what I see by eye for the V10A, compared to a logarithmic plot. :thumbsup: I have added the LiteFlux LF5XT to the graph, as it does both a linear ramp and a logarithmic step pattern. Max output on the LF5XT on 14500 is also pretty close to the V10A on Eneloop, which facilitates comparisons. The linear and logarithmic ramps of the LF5XT also pretty closely match my relative perceptions.

The point to the above is to show that the V10A goes to much lower outputs than any other linear or logarithmic continuously-variable light I've seen before. It also betters matches what I subjectively see across the whole range of outputs on these lights. So for now on, I will also include this type of plot (along with standard linear plots) for all the continuously-variable lights I review.

All that aside, I actually consider this dynamic range control of the V10A to be superior to a purely “visually-linear” ramp across the whole range. The logarithmic correction allows you to exquisitely fine-tune your low output selection. :thumbsup: And once you get into brighter outputs, the proportional increase seems quite linear.

Current Draw

You would hope that this wide range of low modes translates into a range of super-long runtimes. Unfortunately, the circuit overhead for this level of control is considerable, and Sunwayman estimates max runtime to be only 4 days.

I have measured the battery current draw at the lowest output as 50mA on 1xNiMH, which for a 2000mAh Eneloop would translate into 40 hours runtime. On 1x14500 (750mAh), I measure 13mA, which would translate into just under 58 hours. So, basically 1.5-2.5 days is all you can reasonably expect for most batteries :shrug:

PWM

I presume the light uses PWM for the variable outputs, but I was unable to detect the frequency with my setup (which means it must be in the high kHz range). It is certainly not detectable visually. :thumbsup:

At the highest output levels (i.e. over the last quarter turn of the ring), I was able to detect a weak signal in the near ~900Hz range on Eneloop and low 3kHz range on 14500. This seems to be some sort of circuit artifact – it definitely isn’t the PWM freq, as you could easily spot those levels with an oscillating fan.

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 any extended run Lo/Min modes (i.e. >12 hours) which are done without cooling.

I have recently devised a method for converting my lightbox relative output values (ROV) to estimated Lumens. See my How to convert Selfbuilt's Lighbox values to Lumens thread for more info.

Throw/Output Summary Chart:

Effective November 2010, I have revised my summary tables to match with the current ANSI FL-1 standard for flashlight testing. Please see http://www.sliderule.ca/FL1.htm for a description of the terms used in these tables.

As you can see, the V10A’s max output and throw are very consistent for this class of emitter and light, on all batteries (i.e. it is much brighter on 14500). The lowest output mode is incredibly dim - I can barely measure it in my lightbox. Simply put, you can stare into the illuminated emitter quite comfortably at the lowest levels. :thumbsup:

Output/Runtime Comparison:






Note for the runs below that I set the control ring based on Eneloop output levels, so “~70%” and “~40%” don’t really apply to the 14500 runs. Rest assured, the light can go quite low on 14500 – almost as low as on standard batteries.


Overall efficiency seems reasonably good across the range of levels tested - especially considering the continuously-variable interface and (presumed high frequency) PWM-control.

Of course, a current-controlled light with a limited number of defined (and optimized) output levels would be expected to outperform at comparable outputs. But the V10A seems at least consistent with typical defined-level PWM-based XP-G R5 lights.

Potential Issues

Some of the first batch of lights had an issue with flickering at lower output levels. These were recalled, and all currently shipping lights should be flicker-free.

Clip is basic, and may not adequately hold the light if sudden force is applied. Recommend you use a holster or the included wrist lanyard to secure the light.

Light body is fairly smooth (i.e. low in grip).

Relatively few screw threads hold the head onto the body.

Estimated runtime is not as great as you might expect on the ultra-low levels.

Preliminary Observations

I generally like the overall design and build of most of the Sunwayman lights, and the V10A is no exception. It shares most of the same design elements as the M10A and M10R, so it will seem very familiar to CPF users. The one exception is the control ring, which no longer has defined level detents, but is smooth across a continuously-variable range of outputs.

IMO, the dynamic range of the continuously-variable control ring is excellent on the V10A. While not consistently “visually-linear” in the classic sense, it is quite cleverly designed to allow you to access a wide range of low outputs, and then quickly access a roughly visually-linear set of high outputs. This is the first time I’ve seen this particular pattern, and it makes a lot of intuitive sense to me as I turn the ring (see the output ramp graph earlier in this review). :)

The total traverse of the control ring is consistent with the earlier defined-level M-series lights, at around one third the total circumference of the light. Given the continuously-variable interface here, I would have preferred a bit of a wider range, but that’s a minor point.

I like the lack of detectable PWM with the V10A – rare on a continuously-variable light. :thumbsup: For this type of light, output/runtime efficiency seems quite good across the range of outputs directly measured. However, the overhead on the circuit seems fairly high, with an estimated lowest mode runtime of only 1.5-2.5 days, depending on battery source. :shrug:

I also really like the ability to use both standard batteries and 14500, with each battery type maintaining appropriate and reasonable output range control. Sunwayman has definitely gotten that right with the circuit - it provides a nice range of options for the end user. :twothumbs

To get a better feel for how the interface works in practice, I will be carrying this light around as my main EDC for the next little while. I will keep you posted on my experiences!

----

V10A supplied by Sunwayman for review.

----

Post #28

I’ve been going through the data for all the continuously-variable lights in my collection (including so-called logarithmic/“visually-linear” ones), and I think I can revise my earlier position on the V10A (and NiteCore IFE2, which shares the same circuit). I have also decided on a better way to graphically depict the difference, based on the currently accepted model of human perceived brightness.

To explain my original conundrum, here is a direct comparison of the V10A to a light that does both a linear ramp and logarithmic step pattern, the LiteFlux LF5XT. Max output on the LF5XT on 14500 is pretty close to the V10A on Eneloop (i.e. facilitates comparisons). Below is a graph where the y-axis is common (relative output), and the x-axis has been adjusted to show the dynamic range of each ramp (i.e. scaled to put everything in direct context for comparison).

Although not all so-called “visually-linear” ramps exactly match the LF5XT logarithmic steps, I tend to trust that LiteFlux does indeed have a proper logarithmic-adjustment to the outputs. The V10A/IFE2 ramp seemingly looks quite different.

But it occurs to me now that this is not the best way to present the data, as the dynamic range of all the other continuously-variable lights is too narrow (i.e. they don’t go anywhere near as low as the V10A/IFE2). A better way to plot the difference would be to compare the ramps matched for output levels, like so:

At this resolution, it becomes clear the LF5XT’s logarithmic steps correlate very well with the V10A, for the range where the outputs are comparable. So would the LF5XT (and other lights with logarithmic ramps) also look the same as the V10A if they driven to lower outputs? They may very well. So far, on the basis of this comparison, I think you could be justified in calling the V10A logarithmic after all – it is just that that it has a much wider range of low outputs not previously seen.

The question is, are the V10A outputs “visually-linear” over the whole range? Clearly, the V10A is not linear in the linear plot of my lightbox's output readings. So should the lightbox outputs be plotted differently - like on a log scale, as some have suggested? Here's what happens if you do:

Well, that certainly looks pretty linear for the V10A. The problem is, that is NOT how the V10A subjectively appears to me when handling. :shakehead The linear plot is more correct in that the overall output doesn't change much over the first third of the dial. Also, the log plot is not accurate as to how the light seems over the last third of the dial either (i.e. the perceived output increases quite a bit toward the end, but the log graph is flat). Plus, the log plot doesn't represent how a traditional linear ramp looks either (i.e. it makes the regular LF5XT ramp seem like the output barely changes over the vast majority of the time ramp).

So what's the problem here - how come the log scale doesn't accurately represent the perceived brightness over the full dynamic range of these lights? :thinking: Note that the presumption here has always been that logarithmic = “visually-linear”. But up until now, we only had examples with a relatively narrow dynamic range (i.e. the V10A goes much lower than other lights with a logarithmic ramp). Does this relative relationship still apply over this much wider range?

I know it sounds like heresy - but could logarithmic not be "visually-linear" over this wider range? :duck:

Before you all hit the roof, I’ve looked into the scientific literature on visual perception and perceived brightness (that is, the relative perception of varying light output), and found an interesting story. Although generally agreed that perceived brightness is non-linear, the method for “linearizing” it falls into two camps:

1. The Weber-Fechner Law calculates perceived brightness as a logarithmic function. This long-standing method dates back to the late 19th century, and seems to be the prevalent view here on the forums. It also seems to still be popular with amateur astronomers (for low-light viewing of stars through telescopes), and of course many shutter bugs are familiar with it as the basis for photographic light stops. The problem is that while it serves as a good general guideline, it is known to yield wide margins of error across different levels of light output (the limitations of the method are also long-standing - see for example the 1924 paper (ref #4) in my reference list at the end of this post). The general view in the perception field seems to be that a logarithmic-based Weber-Fechner conversion has been superseded by the Stevens' Power Law.

2. The Stevens' Power Law is a revision of the general Weber-Fechner law based on actual measurements over a much wider range of sensations (including vision). It calculates perceived brightness as a power function, with a specific power exponent derived for different types of stimuli. Although this method is relatively newer than the Weber-Fechner Law (i.e. first published in the late 1950s), it has been refined over the years under a variety of conditions, and is now the accepted method for calculating linear brightness more accurately. It is widely used for calibrating modern display devices (such as computer monitors and TVs), as well everything that plays on them (i.e. computer graphics, games, 3D modeling, etc.). The wiki link above shows you the measured power exponents for perceived brightness of different light stimuli (e.g. point source, etc.). For our purposes, the relevant comparison for the output of a flashlight beam is the perceived brightness of a "5-degree target in the dark" (i.e., with a uniformly dark background), which is calculated as the cube root of the total light output (i.e. power exponent of 0.33 in the wiki table). FYI, this stimuli choice and power exponent is also the basis for modern computer color brightness scaling.

Also, I should note that Stevens (and others who followed) were not the first to observe a cube root relationship for perceived brightness - I found one reference going back to 1927. Most of the modern literature reports a Stevens exponent for perceived brightness somewhere between 0.25-0.35, with 0.33 (i.e. cube root) the most common. See reference #1 below for a good summary and references.

So let's see how the lights above look when transformed to a cube root output scale.

Now, that is a lot better, IMO. No, the V10A ramp is not "linear" on the graph, but it much better represents what I see when handling the V10A. It is also a lot better for depicting the relative perception of the LF5XT ramps. :twothumbs

I realize many may not like the implication of the last 50+ years of reserch that a cube root power relationship (i.e. Stevens' Power Law) is more linear for perceived brightness that logarithmic (i.e. the older Weber-Fechner Law). But that is what the literature suggests, and fits with my relative perception of all these ramping lights. Don't shoot the messenger! :shrug:

For those interested, here are some some links to full-text academic research on the subject:

  1. Does Stevens's power law for brightness extend to perceptual brightness averaging? Ben Bauer. The Psychological Record. 2009, Spring.
  2. Perceiving the Intensity of Light. Dale Purves, S. Mark Williams, Surajit Nundy, and R. Beau Lotto. Psychological Review. 2004, Vol. 111, No. 1, 142–158
  3. A probabilistic explanation of brightness scaling. Surajit Nundy and Dale Purves. PNAS. 2002, Vol 29, No. 22, 14482–14487.
  4. The Visual Discrimination of Intensity and the Weber-Fechner Law. Selig Hecht. The Journal of General Physiology. 1924

You can also find a general discussion in the online textbook Sensation and perception
By E. Bruce Goldstein, 2007.

In any case, I am just trying to find the best way to graphically compare the difference between light ramps. I think the cube root scale better resolves the relative perceived difference between ramping patterns, and I will be including it along with linear plots in this and future reviews of ramping lights.

The key point remains – the V10A/IFE2 look very different from other continuously-variable lights, because of their ability to ramp to ultra-low outputs not previously seen.

To follow the online discussions for this review, please see the full review thread at CPF.




Return to the master review list (at flashlightreviews.ca).

For a list of all my CPF flashlight reviews in chronological order by battery type (direct link to CPF), please see here:
Candlepowerforums Threads by Selfbuilt

If you are interested in sending a flashlight to me for review, please see my Terms and Conditions here:
Terms and Conditions

For more information on my testing methodology, please see my Methodology section here:
Testing Methodology

Unfortunately, my flashlights are expensive to feed with all the runtime tests I perform. I don't accept any payment for any of my flashlight reviews, but I will gratefully accept donations to my Paypal battery fund. Your contributions will go toward helping defray the costs of creating all my detailed reviews.

 border=

For cash donations, please use my personal Paypal account (note that ONLY cash transactions are possible on this account).

 border=

For all credit/debit card donations, please use my regular Paypal account.


Site last updated on February 28, 2008 - selfbuilt (at) sliderule (dot) ca (replace the "at" and "dot" labels with the appropriate symbol for e-mail)
All material © 1999, 2011 by Eric Marcotte.