Note: This article is applicable to most cameras accepting two or four AA batteries, although the tests were performed on Olympus Camedia C-5050Z and E-20.

Digital camera batteries

Most digital cameras nowadays use one of two common battery solutions:

• Batteries in the standard AA size, usually of the rechargeable NiMH type;
• Proprietary, rechargeable Lithium-Ion batteries.

Both solutions have their pros and cons.

Rechargeable AAs based on NiMH chemistry have one great advantage: they are a common standard, used in may cameras and other electronic devices; in an emergency they can be replaced with AA alkalines, available at any drugstore or newsstand. They are also relatively inexpensive ($10 or less for a set of four).

Rechargeable Li-Ion batteries usually deliver more energy from the same battery size; they also do not suffer from high self-discharge rate plaguing NiMH ones.

Most cameras accepting AA batteries can also use three-volt, non-rechargeable CR-V3 lithium batteries. A CR-V3 looks like two AAs side by side, and the battery capacity (more exactly: energy stored) is very high, number of times higher than that of a pair of NiMH AAs. The CR-V3 also has excellent shelf life; it can be stored even five years without losing a significant fraction of juice.

Technological progress — or just corporate greed?

Unfortunately, the marketing wizards at major camera-making corporations decided to have a separate standard for each brand. While this could be understandable for batteries differing in size or shape, most medium- and large-size cameras from Nikon, Canon, Olympus and some others use what is basically the same battery (same size, shape, chemistry, voltage), just with customized terminal (contact) placement, to make each brand incompatible with others.

To me, this seems a case of corporate greed prevailing over common sense and over technical considerations. As a result, a Li-Ion rechargeable comparable in performance to four AAs sells for $70 or so (branded with camera manufacturer's logo) or $30 (third party, often made in the same Far-East factory). Did I mention greed?

The excellent (out of production) C-5050Z from Olympus used the first battery solution: four NiMH, AA batteries. Its follow-up, the C-5060WZ, uses a proprietary Li-Ion battery, and I was not too happy to see that.

It so happens that I am using both cameras, with those two different battery systems, and, after initial resistance to Li-Ion packs, I am not sure now which I like more. True, I like the replaceability of AA's, but these Li-Ions keep going forever, and they keep charge when the camera sits in the closet. On the other hand, when an Olympus-only Li_Ion dies in the middle of the Grand Canyon, I have to call it a day, or to buy a disposable film camera, neither of which is a thrilling option.

Rechargeable RCR-V3 — the best of both worlds?

A few months ago a Taiwanese manufacturer announced what, if proven, could be the best of both worlds: the RCR-V3 rechargeable Lithium-Ion battery. The makers and distributors claim excellent in-camera performance (measured in number of shots per charge) and low self-discharge. Adding the compatibility with most AA-based cameras made in the last four years or so, this sounds almost too good to be true.

There are many skeptical voices about these, as the Li-Ion chemistry provides a different voltage than that of two NiMH units (2x1.2V) or of the lithium ones (3V). To be interchangeable with AA's, such a battery would have to include some circuitry to reduce the voltage. This is just one of the technical hurdles to overcome, I'm not going into details.

More, the nominal charge of a rechargeable RCR-V3 is way below that of a decent NiMH in the AA form factor: 1200 mAh versus 1700 to 2300 mAh, respectively. Let us compare the charge values (exactly speaking: energy stored, measured in watt-hours or joules) between the two:

• A single RCR-V3: 3.6 Wh (computed as 1200mAh*3V);
• A pair of NiMH AA's: from 4.1 Wh (1700 mAh) to 5.5 Wh (2300 mAh).

From these numbers one would expect a camera driven by NiMH batteries to last 50% longer than the same model powered by RCR-V3's.

Note that these are nominal values. The actual energy delivered by a battery should be measured as an integral of power over time, and it will be always somewhat less. For example, a set of four Kodak 1700 mAh batteries gives a nominal value of 8.2 Wh (1700mAh×1.2V×4) , while the actual one, as measured by Dave Etchells (see sources at the end), is 8.0 Wh. For the top-rated Powerex 2200, the numbers are 10.6 Wh and 10.1 Wh (the latter quoted after the same source). Actually, I was surprised to see the difference being this small.

What is not helping, is the price of the RCR-V3. To power, say, an Olympus C-5050Z one needs either four NiMH AAs ($10-$12) or two RCR-V3's ($50-$60, as these start at $25 or more a pop). No wonder the market seems quite slow in adapting to the RCR-V3.

Checking it out

I'm getting quite a number of questions of the "what do you think of the RCR-V3" type. I wasn't able to provide a meaningful answer without having a personal experience with these batteries. Plus, I was quite willing to try the new solution because of the low self-discharge rate. Alternating between multiple cameras, I often have one of them sitting idle for a month, after which the NiMH batteries, even if not used for picture-taking, need to be topped off to be useful.

I was disappointed to see that the charger accepts only one battery at a time. As the charging process takes about 3.5 hours (six hours if the batteries are charged the first time), this makes it very impractical for users of cameras which need four AA's (or two RCR-V3's). When such a camera uses its batteries up, I will need 7 hours to recharge my batteries on a single charger, and this only if I'm on hand when the first one is done. Charging overnight? Set your alarm clock at 2 AM.

This takes away lots of the convenience of the new battery type, unless I'm willing to buy and use two such chargers at a time, carrying them both on travel. Inconvenient again.

Unfortunately, I wasn't able to find any maker offering a double-battery charger. The moment I find one, the NEXcell charger goes to the angels. A single-battery charger is fit only for users of pocket cameras, driven by a pair of AA's, and these are intended for the entry-level market, less likely to show any interest in the new technology (or even to know about it). Come on, people, even without a degree in marketing you should figure this out.

The batteries are externally identical to the non-rechargeable CR-V3's, and quite light: one of these weighs less than a pair of NiMH AA's. Surprisingly, there is no NEXcell logo anywhere, or any other identification of the manufacturer except for "Made in Taiwan". This makes me suspect that there may be just one generic source of RCR-V3's, selling them to any battery company willing to resell them further. Oh, well.

All this aside, my first goal was to get an idea about the RCR-V3 performance. While I have read quite a lot of enthusiastic voices on that subject, I will not believe them until I check the numbers by myself: on the Internet the ballot boxes are easy to stuff (and I do have some suspicions there), and even honest opinions may be not always voiced by people with technical background.

The RCR-V3 performance is the subject of the rest of this article.

Battery performance testing: facts and lore

At the risk of sounding opinionated (which I may be) and shocking (which is not my goal), let me start from a frank statement.

Much of the camera/battery performance comparison data seen on the Web is misleading, biased, largely random, and almost meaningless.

What a nice statement to start with, isn't it? Unfortunately, this is true. Here is why.

To compare battery/camera performance between various models, one needs some kind of a measurements, i.e., a controlled experiment, with reasonably reproducible results.

Many such comparisons seem to meet these conditions, providing information similar to this:

We perform the testing by taking a picture every 10 seconds for three minutes, with 50% of frames shot with flash use, monitor on, pausing for one minute, and then repeating the sequence until the camera runs out of power. Based on this, the CrapShot 9900 beats the FoolPix 8400 by 8%, giving up after 216 frames, as compared to 200, both running on the batteries included.

So what's wrong here? The description seems to be adequate for reproducing the results at home, and certainly the CrapShot wins the match here.

Not really.

Let me start from the biggest uncertainty factor: the flash. The engineers of modern cameras of all brands are very smart people, and they've spent lots of time trying to reduce the power usage by the camera's electronics. The last two years brought enormous progress here. Except for the flash, where a given amount of energy is necessary to provide sufficient illumination, period. The engineers can optimize the technology, but they can't change the laws of physics.

As a result, in typical applications most (sometimes up to 90%) of energy used by a modern digital camera goes into the flash. The exact amount depends very strongly on shooting conditions.

• The shooting distance. The flash tries to be energy-efficient: as soon as the camera's metering system decides it has seen enough light, the flash burst is stopped. The energy needed to provide a given illumination grows as the square of the distance. If the CrapShot test shots were at the distance of 1 meter, and those for FoolPix at 3 meters, this translates into a factor of nine. By pointing the camera in another direction, the reviewer biased his results (or, more exactly, the part depending on the flash) by that factor. In fact, the FoolPix camera/battery combination may be a much more efficient.

  Actually, you may send two people, both equipped with the same camera/battery combination to take pictures at a party, and easily see a 100% difference in their estimates of the combination performance. All it takes is one photographer tending to crop his subjects more tightly: a 40% in average picture-taking distance will lead to a factor of two in flash energy used (yes: 1.4 squared gives about 2!). In addition, if the first photographer uses a -0.3 EV compensation to avoid overexposed highlights, this will increase his score by another 26%: cubic root of 2 minus one, but this is yet another story). Anyway, while the tight-crop shooter may say that his FoolPix is a battery champ, the other one will call his a dog. Same camera, same battery, same party.

• The average brightness of the subject. The camera's automation tries to bring the image brightness level close to a standard 18% gray regardless of what the subject brightness is. This means that if the CrapShot was aimed at brighter parts of the reviewer's office (as compared to FoolPix), the results will be heavily biased in its favor.

  Even relatively small changes in how the camera is aimed, or in the zoom angle, may bring significant changes here. The differences may be, again, up to a factor of five or ten. No small change.

• ISO setting (CCD gain). If one of the cameras automatically moves that setting up when shooting with a flash, it will reduce energy use at the expense of image quality. True, in many cameras you can set the ISO value manually, but most reviewers do not consider this relevant to the battery performance testing. Wrong.

  If the CrapShot uses ISO 400 with flash, while FoolPix stays at ISO 50, you get a factor of eight in energy use.

• Aperture used. The light output needed is proportional to the F-number squared. The latter is chosen by the camera automation, and limited by the lens maximum aperture at the given zoom setting.

  If the CrapShot "tests" (note the quotes) were shot at the zoom's short end, with maximum aperture, say, F/2.4, and those of FoolPix at the long end, where you can get only up to F/4.8, this adds another factor of four in the pro-CrapShot bias.

• Ambient light. Its effect may depend not just on intensity, but also on how the camera balances both sources (whether automatically or controlled by the photographer).

I've chosen the flash factor as the first and foremost one to discuss, as it is, I believe, the strongest and least-controlable one, and also because anyone who paid attention at high school should be able to figure it out. (I'm assuming that most of people who write about digital cameras graduated from high school.)

I would think the other factors are easier to control, and also bring less of an effect.

One of these, deserving special attention, is the condition of batteries used in the process. This is especially important in the case of non-proprietary ones (read: NiMH AA's), as their performance may vary widely in individual cases. I have recently discovered that one of my 1500 mAh sets of the 2000 vintage (the second one I ever bought) seems to last longer than a six-months old set rated at 2000 mAh by a respectable manufacturer. Still, these effects range in tens of percent, not in orders of magnitude.

One may argue how much may the power consumption depend on the brightness setting of the LCD monitor (which is the second-heaviest energy user here), what is the effect of the compression ratio settings, etc., but whatever these factors bring, it will not be as dramatic as the effects of flash.

While many of the factors may be averaging out in a process of longer camera use, the statement quoted in the opening of this chapter should be read like this by a knowledgeable Reader:

We perform the testing by controlling the least important aspects of the experiment, while leaving the decisive ones to a chance. Based on this, the CrapShot 9900 delivered 216 frames in a single run we've performed, while the FoolPix 8400 — 200, both with an unknown margin error, which may range anywhere from tens to hundreds of percent.

Not a very precise statement but at least a honest one.

So, what do we do?

The most obvious thing would be to curl in a fetal position and cry. That's not something real men do, though; we have to find some way out of that mess.

The second most obvious thing is to run the measurements without use of flash, thus eliminating the most significant sources of possible bias and randomness. If someone cries out "But I want to know the real-life performance with flash!", have them read the story of two party photographers above, or tell them that it may be anything from 50 to 500 pictures, and this will be true.

In either case, what really counts is the relative performance a camera/battery combination as compared to another one, not the absolute ones.

Unfortunately, running a no-flash measurement is a long and tedious process. I hope you will find the following story educational, or maybe even entertaining.

A '5050 which wouldn't let go

Preparing data for this article, I loaded a set of brand-new, freshly-charged NiMH AA's (1700 mAh, Kodak brand) into an Olympus C-5050Z, set up a stopwatch and a notepad, and started clicking away, expecting the experiment to last a half hour or so.

No flash, no zooming, SHQ JPEG compression, LCD always on. Click, click, click, every eight seconds, interrupted only to re-format the memory card.

After three hundred frames I said, "Well, this is not bad. It may go up to four hundred, very respectable!".

Two hours, fifty minutes, and 1095 (one thousand, ninety-five, this is not a typo!) frames later I saw the red low-battery warning in the monitor. Another four minutes and 42 frames, the camera went belly-up. It was 1:30 in the morning, and what I did was just plain stupid, but we Slavic men are stubborn. (In addition to being quite charming when we want to.)

I'm not going to go through the same ordeal with the RCR-V3's (which was the original plan). We are not that stubborn (and possibly not that charming, even when we want to). A plan involving flash, but in the most controlled and reproducible way, had to be devised.

The controlled experiment

The difficulties are much alleviated by the fact that all I wanted to do is to compare two types of batteries used with the same camera. Therefore we need is the following:

1. Set the camera to aperture priority (to avoid any tricks played by the program logic), write down all the settings in case we want to repeat the experiment some other day: aperture, zoom position, CCD gain (ISO), exposure compensation;
2. Mount the camera on a tripod towards my living room blinds, carefully measuring the distance (remember: just 20% of difference in the distance will change the flash energy usage by 44%);
3. Get myself a handy supply of Guinness (I've run out of the excellent Lancaster Milk Stout from Pennsylvania, too bad);
4. Start clicking.

In addition to the C-5050Z (and to do a sanity check of my results), I also used an older model, the E-20, following the same procedure, but without using flash, as the E-20 uses enough energy to run out of it within a reasonable time even without the strobe.

Text conditions and procedures

The batteries used for comparison were:

• In the blue corner: Kodak-branded NiMH AA's, of nominal capacity of 1700 mAh, once before charged and discharged completely, purchased in June, 2004; charged to the top and kept under a trickle charge until tested;
• In the red corner: no-brand, Taiwanese Li-Ion RCR-V3's, marketed by NEXcell, nominal capacity 1200 mAh, once before discharged, then recharged, and topped off again within 2 hours before the experiment;

The comparative performance was measured for two cameras: an Olympus C-5050Z, and an Olympus E-20, which is much less energy-efficient. (For the latter, batteries were recharged once more, and the test was done the next day.)

Test conditions

1. Olympus C-5050Z: firmware Version 558-82, aperture priority at F/5.6 and ISO 64, no exposure compensation, saving as SHQ JPEG, built-in flash and LCD monitor on, zoom at EFL = 35mm. Both series shot from a tripod, with the frame composed identically (beige window blinds at the distance of 2.5m).

   In my rough estimate (not really relevant for the comparison), the flash was firing always at about 75% of full power.

2. Olympus E-20:; firmware version 29-1102, flash off, LCD monitor on (the camera disables it when saving images), zoom at EFL=35 mm. (Other settings, most probably without a meaningful impact, were the same as above.)

Results

Once again, a warning: these results can be only compared against each other, not with any others, published by anyone, including myself. Within these series, the experiment conditions (even those presumably irrelevant) were strictly controlled and identical.

And here are the raw results. For your information, the total uptime (i.e. the time from turning the camera on with fresh batteries to the moment it ran out of battery charge) is also given in the lowest row.

Oh boy, are we in for a surprise!

The middle-of-the-road, 1700 mAh NiMH's easily outperformed the new battery type, providing almost twice as many frames (exactly: 76% more) with the C-5050, and 40% more for the E-20. This is even more of a difference than what one would expect from the nominal ratings: 8.2 Wh (4*NiMH) is just 14% more than 7.2 Wh (2*RCR-V3).

Interestingly, the total uptime of the '5050 was lower for the RCR-V3's, as the flash was recharging significantly faster with these batteries. This may be due to the higher voltage (3V versus 2.4V) and/or lower internal resistance (I have no information about the latter, so this is just a speculation). Actually, in order to avoid any effect of the frame rate, I should have timed it, but this was too much. I doubt if this could have an effect exceeding one or two percent.

Another interesting difference in behavior of both battery types is how they cut off the juice.

• When run from the NiMH AA's, the '5050 would detect a drop in voltage and signal it with the low-battery warning after 396 exposures, still allowing me to shoot 20 frames more. In most of these, the monitor would go off while the flash was recharging, to go on again when it was done. This is understandable: drawing the (significant) flash-charging current lowers the voltage on battery terminals if the internal battery resistance is not negligible.
• With the RCR-V3 the camera would work happily up to the moment of shutting off with no warning whatsoever. This clearly confirms that, as the makers say, the RCR-V3 has internal circuitry cutting off the current to prevent the battery from damage by over-discharging.

Also, the NiMH batteries were seemingly warmer when removed from the camera — but this could be expected, as they delivered almost twice as much energy.

The verdict is clear. The guy in the blue corner wins by a knock-out. At least in the freshly-charged battery competition. We didn't even have to call his 2300 mAh team-mate.

At the same time, you may notice vast differences in energy usage by both cameras, without flash, when using NiMH batteries. While the E-20 gets 302 frames from a charge, the C-5050Z — 1095 frames mentioned before, quite a difference! Actually, the '5050 gets more shots with flash than the E-20 without.

Why all this mess?

Having eliminated most (possibly all) factors affecting the relative results, I stand by my numbers. Actually, I trust them more than anyone else's. So why are people often reporting so contradictory results?

I've given the answer in the "Facts and Lore" chapter above. Some people, acting in good faith, are not aware about the effects of (uncontrolled) flash use on their results; some might have been comparing a brand-new RCR-V3 against a well-used set of NiMH's which have seen better times. I think my measurements avoid at least these two pitfalls (and most others).

Following the RCR-V3 threads on a number of discussion groups, I'm also smelling something fishy here, suspecting that a number of individuals are active in "working" the groups, making biased posts, sometimes under assumed names (some of these are given away by the same spelling and grammar errors in messages differently signed; come on, guys, it really pays to learn some English, even if it is your native tongue).

Also, some distributors are making inflated claims on RCR-V3 performance, without giving any information on the data source, or on the measurement procedure.

Even worse, some reviewers are repeating these claims making them sound like their own conclusions. Go to the RCR-V3 "review" (quotes intentional) at Steve's Digicams. Mr. Jenkins provides a comparison of 420 frames delivered by an RCR-V3 on a Kodak DX4530 camera, versus 250 frames with use of the NiMH AA's.

The source of data is not given, so an obvious conclusion is that these are reviewer's own measurements, although the procedure is not described or even mentioned.

The advertising link at the bottom of the page used to lead to the RCR-V3 distributor, batteryx.com; near the bottom of which you could find a table with exactly the same data (also without any information on the procedure or source), with a reference to a third-party review at, you guessed it right, Steve's Digicams! (February 2005: the link is broken and the whole site seems to be down. Out of business? October 2006: it looks like the domain has been acquired by another battery distributor.)

In my previous life I was a scientist, doing full-time research in cosmic-ray physics for sixteen years. If I've done something like this, it would be called a "publication fraud", and it would have cost me my job. The only thing a reviewer has to lose is credibility. It just happened.

Other sources of information

Given the intrinsic difficulties of coming up with objective battery/camera performance comparisons, and the less-than-impartial nature of many sources, my advice is to approach all information with skepticism. This includes my own results (who knows, maybe I have a stock in NiMH manufacturing?). This is especially valid in case of sources financially connected to battery makers (e.g., accepting advertising from them), or voices on discussion groups (ballot-stuffing).

The best way out is to do your own measurements, under as controlled and identical conditions as possible. This, of course, requires shelling up the money for the new batteries up front, so it is not always feasible.

The second-best approach is to ignore any claims not accompanied with an unambiguous description of procedures used in measurements, focusing on those which do.

With all these reservations in mind, let me point you to some RCR-V3 information on the Web which I came across in the last weeks.

• NEXcell, a battery maker and distributor, published a Performance Comparison Chart on their Web site. The information shown there seems reliable.

  They test batteries off-camera, submitting them to a 600 mA drain for 3 seconds, and a 300 mA drain for another 3 seconds, and using this cycle as a measure of one-frame energy consumption by a typical camera. This procedure meets the most important condition of reproducibility, removing all (all at least most) of the bias and randomness factors I've listed above.

  According to NEXcell's, an RCR-V3 (1200 mAh) stores enough charge for 1700 of such cycles, while a pair of NiMH AA's (2200 mAh) — 2200 cycles. This seems closer to my results than most other claims I've seen, although in my book the AA's may have a somewhat greater advantage.

• The Dan's Data technically-oriented and knowledgeable Web newsletter deals with the RCR-V3's in two issues: No. 89 and No. 90. The author is also skeptical about the conclusions of the Steve's Digicams "review", although much of the performance data he quotes after his respondents was obtained under uncontrolled conditions. The discussion of underlying mechanisms is quite educational.
• Thomas Distributing does not post any general article on RCR-V3's, but their Moby Power RCR-V3 Page claims that these batteries "Will actually out perform [sic!] a true 2300 mAh NiMH battery due to normal voltage" (the same creative spelling is used at least three times in two different pages, I'm not inventing it!).

  While I haven't tried Moby-branded RCR-V3's, I doubt whether their performance is twice as good as of those I've got from NEXcell (they look slightly different), and I wouldn't trust such claims not supported with any additional information. If any of the Readers has these batteries and is willing to do a controlled-conditions comparisons on them, I will be very interested in seeing the results.

  As a matter of facts, similar claims are made for the NEXcell batteries, contradicting the manufacturer's own data.

• The Moby Power Web site consists as of this writing (June 20, 2004) of just one page (in addition to the front cover), containing a RCR-V3 Cost and Performance Table, accompanied by a single paragraph of text. (For me this is an indication that the company may be a couple of months old.) Guess what? This is the same table we've already seen at Steve's Digicams and BatteryX, without any information about the data source or even the camera for which the data was supposedly acquired.
• Delkin Devices, a company which distributes the same batteries and the same charger as NEXcell (the charger is identical, just with a different label) provides a PDF document on their (?) batteries. It contains the same table as that used by three other sources above (obviously, inconsistent with NEXcell data on the same batteries). Go figure.
• Batteryspace.com have their own branded variety. The most detailed performance information is that the battery "offers more shots than any AA Ni-MH batteries". Life is simple.
• Although this is not directly related to the RCR-V3, you may want to know that Dave Etchells of The Imaging Resource performed a gigantic task of comparing the performance of about 40 various models of NiMH batteries.

  His test methods seem solid and are well-documented, and if you are looking for hard (and mutually comparable) numbers on NiMH batteries, look no further. If we beg Dave long enough, he may run some RCR-V3's through his benchmarks, and this could clear a lot of confusion.

• Steve's Digicams RCR-V3 "review" was mentioned above. 'Nuff said about that one.

Not much, and most of it quite suspect, I would say.

Generally, do not waste your time to watch online discussion threads, for the reasons I've described above.

In technically-sounding articles and postings, try to figure out the level of writer's technical literacy. One easy symptom is using "power" instead of "energy" or "charge" (these two are also not equivalent; multiply charge by voltage to get energy). If the writer does not provide description of procedures used ("too much for your pretty little head"), do not pay much attention to the conclusions.

The bottom line and recommendations

At this moment, the RCR-V3, in spite of inflated claims and urban legends, does not deliver the performance (in terms of energy) of a decent set of NiMH AA's.

The lower self-discharge rate of the RCR-V3 is probably true (out of general considerations), but I was not able to verify that, or to find anyone who did it.

The prices of RCR-V3 are still very high. Compare $52 for two of these against $13 for a set of four highly rated, 2200 mAh NiMH PowerEx batteries by Maha.

The available chargers accept only a single RCR-V3, a major inconvenience for most digital camera users, who use two such batteries at a time.

I would recommend waiting until the RCR-V3 evolves into a higher capacity. The technology will probably improve, it just needs some more time. For now, treat yourself to a set of premium NiMH AA's from a respectable maker.

Still, if you want your batteries to stay charged for weeks at a time while your camera is not being used, and if you are willing to accept less energy per charge and the much higher price, go ahead and try the RCR-V3 out. Just consider yourself warned.

Follow-ups and more

Since the original posting of this article, I'm receiving quite a few emails from other photographers who use RCR-V3s with a number of cameras, not necessarily the C-5050Z. In most cases they just report their own disappointing experience with these batteries, but some found the RCR-V3 a viable solution.

In particular, Bob Dzuricky (a master pool cue maker and an experienced photographer, see his Web site for some impressive stitched panoramic images, and not only) wrote:

I enjoyed your article comparing NiMH vs. RCR-V3 batteries. Very informative & entertaining! However, I'd like to relate my experience with batteries and my Sigma SD10 camera.

I was only getting between 30 and 80 pictures with NiMH batteries. The switch to RCR-V3s allows me to regularly take 350-500 pictures. (...) It seems that Sigma has a low voltage sensor that is set too high and prevents one from using the NiMHs to full advantage. The RCR-V3s apparently put out just a bit higher voltage that keeps the sensor happy.

The shelf life of the RCR-V3 is remarkable, too. [...] If I waited a week without recharging the NiMHs, I would be lucky to get 10 pictures. With the RCR-V3s, I don't think twice about freshening up a charge and they have never failed me in that respect. In this instance, the RCR-V3s have a definite advantage. (...)

My only complaint is that I've run into some weird problems occasionally with the RCR-V3s. I believe the internal chip gets flaky sometimes. I install a pair of freshly charged batteries into the camera, but it will not turn on. A voltage check before and after installation shows that one of the batteries will instantly go from 3 to 0 volts. Putting the dead battery back into the charger for one second is enough to make it show 3 volts again. Re-inserting the batteries back into the camera may or may not allow the camera to function yet. Sometimes I will have to repeat this drill 10 times before the camera works. From this point on, I'm good to go for the next 400 pictures.

In other words, looks like the Sigma users will find the RCR-V3 a lifesaver. Good to know.

Another positive experience with the RCR-V3s was reported by Ron Godenow, who is using them with the compact Olympus C-560:

Freshly powered 2100 mAh NiMH's give me about 100 shots [...] using LCD and flash about half the time [...], an RCR-V3 gives me upwards to 250-300. I've run the test several times.

Well, it looks again that the C-560 may be a bit fussy about the voltage, and the RCR-V3, in spite of lower capacity, keeps the voltage better, thus providing more usage. Interesting.

The only way to be sure is to check the performance with your particular camera model. If you are lucky, the RCR-V3's may, in spite of my results, prove to be a good investment.

As of this update (three months after acquiring my RCR-V3s), my experience did not change my initial opinions. I keep these batteries for my C-5050Z and E-20 being temporarily not used (or only occasionally used), where their low self-discharge is, indeed, a blessing. However, when I take one of these cameras for a serious workout, I replace the RCR-V3s with freshly charged NiMHs.