The plastic lamp housing appears durable and cleverly designed to enhance the dancing flame effect at close range by strategically masking the gaps between individual LEDs and reinforcing the illusion of continuous moving flames, a kind of spatial analog to the "magic" of using rapidly displayed still frames to create the illusion of motion pictures.
The dancing effect itself is complex and random enough that it feels quite natural — not mechanical or repetitive — and the shape and size of the flames are very satisfactory.
I purchased a pair of these with the idea of attempting to adapt one of them for use in a concrete (faux stone) Japanese garden lantern. We'd lived in Japan for several months and noticed that these ornamental lanterns are sometimes illuminated with candles or lamps, especially in temple gardens, but we didn't want the risk or inconvenience of an open flame here in southern California.
The adaption (see accompanying video) worked better than I'd hoped. Basically, I separated the LED lamp unit from the housing that contains the solar panel/sensor and the on/off switch and reconnected them with two six-foot double cords (4 wires total). This so that the solar panel assembly could be positioned at a distance from the Japanese garden lantern to preserve aesthetic qualities.
The original housing containing solar panel and switch is inconspicuously staked several feet away amidst low shrubbery, and connected to the LED lamp unit via extension wires coming up through a central hole in the bottom of the lamp. To be on the safe side, I strung the wires through a short piece of electrical conduit (see video) to thwart enthusiastic gardeners severing it with shovels or trowels. The conduit will eventually disappear under mulch and plants.
I'd tried this re-wiring trick in years past with other solar lights without success — I'd never been able to get them to charge properly — but splicing the extra lengths of wire into this device was easy and hasn't seemed to have impaired the charging, switching, or operating of the lights a bit. I used crimp-on connectors for both the butt splices and the "disconnect" pairs (placed so that the LED lamp can be easily installed and removed for maintenance or replacement without taking the whole setup apart).
Detail: Since I had exactly 4 wires to deal with, I used two different styles of disconnects (bullet and blade) and alternated the male/female ends so that each of the 4 wires can only connect with its proper mate. Avoids decomposable labeling or trial-and-error fussing. Since this is outdoors, I added drops of hot glue to seal the connectors to prevent moisture getting in and corroding the wires. A few drops of hot glue in the reassembled solar panel/switch housing relieves the strain from the wires. Disassembling and modifying the unit no doubt voids the warranty, but it's lots of fun!
As you can see in the video, the effect is pretty convincing. I kept one of the lamps as-is and placed it in another corner of our garden to compare performance with the modified set-up, and so far both are performing flawlessly and beautifully. Both are placed in partial open shade, but the generous size of the solar panels appears to be enough to charge the battery for at least 3-4 hours full strength. We'll see how they do in the winter with fewer hours of sunlight. If need be, I'll reposition the solar panels — another plus of separating them from the LED lamp. I'll try to remember to report back here in a few months.
My usual critics (read "family") loved these lights so much they requested a couple more to place elsewhere around the garden.
To date I've purchased three different "brands" of these lights — one 20% cheaper than the other — and it appears they are really just different vendors — the lamps themselves — right down to the instruction sheets — are absolutely identical, so I'd suggest shopping around for the best price per unit at any given time. A good way to find all the options seems to be to search Amazon for specific features such as "96-LED" or "Dancing Lights" or... well, you know how to experiment with search terms.
Update: I've had the original lights I purchased for over a year, and they're all as good as new, as nearly as I can tell. During the reduced daylight hours of our southern California "winter" they would last only about 5-6 hours after sundown, which was perfectly acceptable for us, and bear in mind these were not positioned ideally to catch all available sun, quite the contrary. When the long days of summer again rolled around all of the lights seemed to be performing at full strength, with no maintenance whatsoever. More props to the designer(s).