Bowens Gemini 1500Pro studio flash
The Gemini 1500Pro is the most powerful Bowens monolight (i.e., studio flash with tube, generator, capacitors and all other electronics in the same physical unit). The Gemini Pro, according to the Bowens literature, is the series of Bowens monolights that emits the shortest light burst at full power. For the 1500Pro at 1500 Ws, a discharge of 50% of the power takes place within an interval of 0.9 ms (according to third-party data) or 0.7 ms (according to Bowens). Bowens monolights are not among the cheapest (quite the contrary), do not come in the colors of the rainbow, are not named after extraterrestrial insects and do not look like a ray gun from Star Wars. However, they are very dependable, honestly rated and, in daily use, feel more like a Rolls Royce than a Soviet-era Skoda. Incidentally, if you are like me, you will probably be interested to know that Bowensdirect.com often offers factory-refurbished units at substantial discounts. The only things that tell me my recently arrived 1500Pro is not a factory-new unit are a few small signs of use and my slightly thicker wallet (for now - money always seems to find its way out of there). In a way, a refurbished unit may be more reliable than a factory-new one, because it has been tested in real use by a previous owner and whatever could go wrong because of defective components already went wrong and was repaired. I am not particularly concerned about the possibly reduced life span of a used flash tube compared to a completely new one, since I was planning from the start to replace the standard coated tube and dome with uncoated ones (see Figure 5 below), which is indispensable for some of my applications. In any case, with the use I make of my Bowens units, they are probably going to outlast me. MountingThe Gemini 1500Pro is approximately twice as long and twice as heavy as the Gemini 500R, which I reviewed here. The 1500Pro is equipped with a correspondingly longer hanging bracket ending in a socket for a baby stud, which allows the unit to point straight down (when hanging down from a support) or up (when mounted atop a tripod). Personally, I find the thought of a 1500Pro mounted atop a lighting stand rather unnerving, and I much prefer an arrangement with the unit hanging down from a ceiling fixture or other type of rail. The long hanging bracket also means that this model will hang lower than a 500R from a ceiling mount, which may be a concern if the ceiling is already low. The hanging bracket also has a transversal hole that allows the bracket to be mounted at a 90° angle to the mounting stud. This adds some flexibility for mounting the unit, e.g., with the bracket horizontal, but is hardly practical because of the high weight of the unit and long lever arm of the bracket. I am not sure whether a brass or aluminium stud, for instance, can resist this force with a suitable safety margin. The 1500Pro is almost too heavy (5 kg) for amateur-class studio pantographs like the ones sold by Walimex on eBay, which are rated at 6 kg. They hold up a 1500Pro only after their spring tension is substantially increased by turning the friction disk (near the bottom of the pantograph) clockwise, and might wear out relatively quickly under this load. Safety cable attachmentAll Gemini units have a drilled hole near the end of the hanging bracket, meant to be used for threading a safety cable through the hole. However, the hole is somewhat under-dimensioned and only about 8 mm wide, which is only enough to allow the passage of a very thin and flexible safety cable. The weight of the 1500Pro does require a substantial safety cable: 2 mm is a minimum, rather than the 1 mm cables distributed with some hobby-class studio equipment (twice the thickness means four times the cross-sectional area and tensional strength, so this is a big difference). The loop at the end of a 2 mm safety cable is way too stiff to pass through an 8 mm hole, and the ferrule holding the loop closed is often thicker than 8 mm. I solved this problem by mounting a maillon, i.e., an elongated steel ring (roughly similar to a carabiner, but held closed by a threaded nut) through this hole. The safety cable easily passes through this maillon, or the maillon can be opened to admit a cable that ends with carabiners at both ends. As an alternative, the rear handle of the unit could be used for anchoring a safety cable, but this is often in the way of changing the inclination of the unit and requires a rather long safety cable, which partly defeats its purpose. A safety cable should be as short as practical, to prevent a falling piece of equipment from accelerating to a dangerous speed and to prevent it from wildly swinging around after being stopped by the cable. Longer cables than 0.75-1 m are rarely necessary. In some studios, I have seen safety cables so long that they might not prevent people from being hit on their heads by falling equipment. The safety cable should also be sufficiently strong to bear a weight multiple times the actual weight of the equipment (5 to 10 times higher is a good bet). In addition, the correct way to fasten a safety cable is to pass it through a ceiling ring or around the top of a ceiling rail, then through a ring or handle of the object to safeguard, and finally to attach the two cable ends together with the built-in carabiner, forming a closed loop. This doubles the intrinsic strength of the cable. An incorrect way to fasten a safety cable is to fasten it to the equipment at one end and to the ceiling support at the opposite end (making the cable a single line rather than a double one). Vinyl-coated safety cables can be used to protect equipment from wear and scratches, but the vinyl may fray against sharp edges, especially if the equipment is frequently moved or adjusted. PowerThe Gemini 1500Pro is specified as giving a power of 1,500 Ws. The power ratings specified by Bowens are honest, unlike those of many manufacturers of cheaper studio flash units, who use meaningless measurement units like "effective Ws" and the like (see also here). It is not possible to directly convert Ws to GN (Guide Number), because the latter depends on the reflector and light modifiers mounted on the flash. Bowens specifies a GN of 150/m at ISO 100, which seems reasonable, and roughly corresponds to what I get with this unit and a small reflector. Some sources give a GN of 160 for the earlier Bowens Esprit 1500, which had an identical power rating in Ws (e.g., http://olegkikin.com/guidenumber/). In comparison, a battery-operated, portable unit like the Nikon SB-900 is rated at a GN of 34/m at ISO 100, but the actual GN in this case depends on the zoom setting of the flash head. In any case, a GN of 150-160 means approximately 5 stops of light more than a GN of 34, or being able to shoot at f/11 instead of f/2. 1,500 Ws is the same amount of energy as 1,500 Joule. This is almost ten times the muzzle energy of a .22 caliber round, or three times that of a 9 mm round. The reasons why a 1,500 Ws electronic flash does not instantly kill a person is that the energy emitted by a flash tube is not coherent - the light waves are not in phase with each other - and its energy is typically spread across a wide surface. As a comparison, a shower of sand grains (not coherent) is obviously more survivable than a falling concrete slab (coherent). This is also why laser radiation (coherent) is dangerous even at a much lower power than an electronic flash. In practice, a "pop" at full power from a naked 1500Pro in a medium-sized room is painfully bright even if you were looking away from the flash. For some of my special applications, I shoot with the flash unit at close distance from the subject (half a meter or less), and developed a habit of closing my eyes while firing the flash, which is a real necessity. At these close distances and near the full power setting, there is also a risk of causing superficial skin burns, damaging small heat-sensitive subjects and even igniting flammable objects like thin dark paper. When working at very close range, I would recommend to use barn doors or a snoot to restrict illumination to the area where it is really needed. Incidentally, the 1500Pro does produce a relatively loud popping sound when firing at high power, even with the protective pyrex dome covering the tube. Some readers may question why I really need 1,500 true Ws of flash power (which is roughly equivalent to 3,000-4,000 effective Ws as specified by many manufacturers of flash equipment). My main reason is that I use studio strobes, among other things, as sources of multispectral radiation (ranging from near UV to near IR) for the scientific photography of special subjects under these wavelengths. The solid-state sensors of current digital cameras are not very sensitive to near UV, even after the camera has been modified by removing the internal UV and IR filter. These special subjects are small, and require close-up photography and macrophotography techniques, and consequently lens apertures around f/8 to f/11 for optimal DOF. With a 500 Ws strobe with uncoated tube (the Bowens Gemini 500R), even with the flash tube placed 30 cm away from the subject I often need to flash at full power. This leaves me no margin in case I should want to illuminate a larger surface, to work with unusually UV-absorbing subjects or to use special filters that transmit only narrow bands of wavelengths and/or low levels of radiation. The 1500Pro, on paper at least, gives me a margin of 1.5 stops, which is not much but better than nothing, at a price that is still relatively affordable (quite a bit less than my last camera body, an Olympus E-M5). Two independently adjustable radiation sources are also better than a single one for certain uses. However, I had a big surprise when I started testing the 1500Pro in UV photography (see near the bottom of this page). As a definite plus, both studio strobes, after replacing the tubes and dome, remain available also for general-purpose studio photography, and in fact I use them a lot more often than battery operated strobes like my Nikon and Metz ones. Indeed, for these uses I rarely find I need more than 250 Ws, but the 500R lets me reduce its power by up to 5 stops, and the 1500Pro by up to 7 stops (note that both units, in addition to their "Stops" dial, provide an extra stop of regulation via their "Tenths" dial). With the 1500Pro placed outside a 1 m light tent, I can usually shoot product photography at base ISO and a power dial setting of only 2 to 4. Cooling fanThe 1500Pro is equipped with an internal cooling fan that blows air from the inside of the metal casing through the pyrex dome that protects the tube and modeling light. The fan turns on whenever the modeling light is in use (even at very low power, but the fan speed is variable, so it does not blow a lot of air and make much noise at low power). The fan does not activate when the flash is operated with the modeling light switched entirely off. It is possible that the fan may switch on even without modeling light to prevent overheating when the flash is fired numerous times in quick sequence, but so far this has not happened to me. Differences from other models and series
This model offers all the settings available in the Gemini Pro and R series (see also here). In fact, there are no great differences in settings and functions between these two series, except that the Pro series is equipped with a multi-voltage power supply, so you can use it almost anywhere in the world (provided you change the modeling light bulb), and provides higher powers. Where the Pro and R series mainly differ, instead, is the flash tube and its electronic circuits. In the Pro series, the flash tube has three electrodes (in addition to the external trigger wire; see Figure 6) instead of the traditional two. I am not sure how this works, but the Pro tube perhaps does not work like two traditional tubes joined end-to-end, with one shared electrode in the middle. This is just a guess on my part, since I decided for now not to dig into the electronics of this unit. However it works, this is a somewhat different tube than ordinary xenon arc tubes, and the electronics must also be partly different. The tube of the Pro series has a shorter radius of curvature than the R series, which reduces the length of the path between the end electrodes (see Figure 6 below).
The patent describes a tube with a pair of anodes at its ends and a single cathode at its center, and is described as providing a broad illumination surface (which is a consequence of having a relatively high length) combined with a short flash duration (which results from a relatively short path between anode and cathode). Thus, as described in the patent, this tube works like a parallel-coupled pair of separate tubes. However, this is probably not the case of the Bowens Pro tubes, because, with extended use at high power, one (and only one) of the end electrodes develops a whitish deposit on the interior of the glass tube near the electrode (possibly caused by metal sputtering). If the polarity of the two end electrodes were the same, this deposit should be expected to appear near both end electrodes. It is therefore quite possible that the Bowens three-electrode tube works like a series-coupled pair of separate tubes. This is likely accompanied by a pair of internal independent power supplies and capacitor banks, or a power supply with a center-tap carrying a voltage intermediate between high-voltage and ground. Recycle time at full power is 2.3 s, which is longer than models of lower power but not so long for 1,500 Ws. The internal power supply is very well stabilized, and unlikely to be affected by reasonable mains power fluctuations. Like all Pro and R series units, the 1500Pro can be equipped with an internal radio trigger module and powered by a battery-operated Bowens external generator. An external antenna for the radio trigger module can be attached after removing the button visible at the right in Figure 2. It is also entirely possible to attach a cheaper external radio trigger receiver to the sync jack on the rear panel (like I did). Doing away with the sync cable does make wiring simpler in a studio setting. All R and Pro units come ready to be controlled remotely with a Bowens IR remote control (just turn the Stops knob to the REM setting). Two models of remote control handsets are available. The smaller and cheaper one, according to some users, may be insufficiently powerful to reliably operate in some large studio settings. The window of the remote control receiver is located near the LEDs on the left side of the unit, and is separate from the trigger photocell (which is at the top, near the rear of the unit). All models of the Gemini R and Pro series have a built-in microcontroller that carries out most functions and sits between the controls and the hardware (unlike in cheaper studio strobes, it seems none of the Gemini manual controls is directly coupled to the hardware). Many settings are accessible only through the "user set-up mode" (as Bowens calls it), which is entered by switching on the mains power while keeping the flash test button pressed. Keep the button pressed for a few seconds, and the user set-up mode is entered. At this point, the digital display flashes a sequence of four 2-digit numbers. The first number is the firmware version. The following three numbers tell the total number of flashes done by this unit since it left the factory (but the count restarts after the 1,000,000th flash). This may be a way to tell how much a second-hand unit has been used (but I don't know how easy it would be to fraudulently alter or reset the count, and this tells nothing about the used power levels - 100,000 pops at full power certainly wear down the tube and capacitors more than 100,000 pops at low power). Once the user set-up mode is entered, the Tenths dial can be used to select a setting (as described in the manual) and display its current value in the digital display. The modeling lamp function switch can then be used to change the value. Once finished, don't forget to press the flash test button to save the new values. One of the available settings is a reset that returns the unit to factory-default settings. You also need to power down the unit, then up again to exit the user set-up mode and return to its normal operating mode. While not extremely difficult or lengthy, changing these settings is not comfortably carried out in the middle of a session. Most of these settings (e.g., number of idle minutes after which the modeling light turns to a lower power, sound on/off, photocell on/off, number of preflashes to ignore when using TTL "digital" flash units as remote triggers), are not normally changed after the initial setup. If you frequently change some of these settings, it may be a good idea to attach a copy of the relevant manual pages on or near the unit. DumpingThis model offers autodumping, i.e., after reducing the power settings it is not necessary to manually fire the flash in order to get rid of the excess power already present in the capacitors. Bowens monolights do this for you, and offer two ways of doing this: resistive dumping (i.e., discharging the excess capacitor charge through a bank of power resistors) and flash dumping (firing the flash to empty the capacitors of the excess charge and then recharging them to the desired level). It is possible to select in the firmware settings whether to always use resistive dumping or to let the unit perform a flash dump whenever this is faster than a resistive dump. With the 1500Pro, resistive dumping from maximum to minimum power takes several seconds, but a full charge dump is unlikely to be needed very often. I have my units set to resistive dumping, which is the default. Models of lower power tend to have a faster resistive dumping, because the amount of energy to dump is much lower. The LEDs around the manual trigger button blink quickly while recycling and more slowly while dumping. Bowens monolights automatically dump the capacitor energy when the unit is switched off, in order to avoid the risk of a (probably fatal) electric shock when replacing the flash tube or modeling light. Cheap studio flash units typically do not do this, and short of using a multimeter you never know whether their capacitors are still charged and for how long after switching them off. After switching off a unit it is usually impossible to manually fire it to make it safe, and even when possible, a residual charge of around 100 V is typically left in the capacitors after manually firing the unit. Modeling lightThe modeling light can be set with the rear panel controls to full power (i.e., always at maximum intensity), proportional power (approximately proportional to the flash power setting), custom (the intensity can be manually adjusted and is independent of the flash power settings) and off. The last used setting is remembered the next time the unit is switched on. Three LEDs on the rear panel indicate the current setting. Three more LEDs on the left side of the unit make this setting visible even when the unit is hanging from a high ceiling with the rear panel out of sight. 
The 1500Pro uses a 500 W halogen modeling bulb in E11 mount (which is a threaded socket much smaller than ordinary light bulbs). It seems difficult to find an adapter for using another type of halogen bulb in this socket (e.g., if you need to use a much less powerful modeling light). A 300 W replacement is available from Bowens for 250 V mains (250 W and 100 W replacements are available for 117 V mains). A further thing to note is that the Gemini R and Pro series sense whether the modeling light has blown out and refuse to work in this case. You may find it impossible, for instance, to replace the modeling light with a small halogen bulb or LED light, because these light sources use a low current and may fool the flash unit into believing that the modeling light has burned out. It is of course possible to manually turn the power of the original modeling light way down to reduce its intensity to almost nothing, but light becomes quite orange-tinged this way and most of the lamp emission spectrum moves into the near-infrared, so much of the heat is still there even when the light decreases. This may or may not be a concern, depending on the planned use of the flash. 
My unit came with a 500 W 230 V modeling light bulb mounted in the unit and a spare in its box. I don't know if this is standard, or they supplied a spare because the one in the unit has already been used for some time. My factory-new 500R came with only one bulb. A quick online search shows that 240 V E11 halogen lamps are available in powers of 75, 100 and 150 W, albeit I have not tried them and I don't know whether they will work in this flash model. In general, low-power E11 halogen bulbs seem to be more common for 117-130 V than 250 V. 
Coated or uncoated tube and dome?The tube and modeling light are protected by a large and thick pyrex dome with ventilation holes. Before removing or replacing the dome and tube, you should power off the unit, disconnect its power cord and, if necessary, let the unit cool down. The modeling lamp, tube and dome can get extremely hot during use. It is a good idea to use clean cotton gloves to handle the dome, tube and modeling bulb, in order not to leave fingerprints that absorb IR radiation and cause the glass to expand unevenly, possibly causing microfractures. Latex gloves are less suitable, because they may leave starch powder or oily substances on the glass. These operations require a good access to the front of the unit, which means any reflector or modifier must be removed and the unit placed at a comfortable height for observation and access. The protective dome is made of pyrex and is removed by pressing in two brass-colored buttons that sit in holes of the dome (wider than the ventilation holes) near its base. This operation requires no tools, and in fact is much more safely performed by hand. Simultaneously pressing the two buttons may be difficult, but pressing one and then slowly bending the dome outwards eventually frees it from the second button. The unit can be operated even without the protective dome. However, the dome is useful to limit the severity of burns in case of accidental contact with a hot unit, and also to contain glass splinters and reduce the likelihood of a softbox catching fire if the tube should accidentally explode (which is a rare but not impossible event, and more likely in high-power units). 
Replacement flash tubes for this monolight are large, remarkably thick-walled and quite expensive. They are not compatible with the tubes used in other Bowens Gemini series. They can be replaced by the user, but the 1500Pro uses a difficult to reach, tamper-proof Torx screw (a.k.a. Torx pin-head screw) to fasten the trigger wire to its electrical contact on the unit. This type of screw requires a matching driver, bit or wrench (ordinary Torx tools won't work). A single tube type is used in all models of the Pro series, which includes four models with maximum power ranging between 500 and 1,500 Ws. The Pro units are delivered with a standard tube and dome, both coated with a gold vacuum-sputtered layer that filters away most UV radiation. Uncoated replacement domes and tubes are also available from Bowens. Ordinary digital cameras, unlike photographic film, are essentially insensitive to UV radiation (the sensor itself is somewhat sensitive, but a UV-cut filter is always present in front of the sensor, and most lenses strongly absorb UV in any case). Thus, a UV-cutting coating of flash tubes is not really necessary to avoid the so-called UV fogging sometimes seen, for instance, in landscape photography with film cameras. I guess this coating on flash tube and dome is there mostly as a habit (since photographers are well used to the performance of UV-coated flash tubes), to provide a daylight-balanced illumination (uncoated flash tubes generate a slightly "colder" light) and possibly also to reassure users who may be concerned about exposure to UV radiation. However, many fashion photographers now routinely use uncoated tubes because they prefer their more neutral emission spectrum. Color balance with digital cameras, of course, is easily adjusted in post-production and is no longer a critical issue. Incidentally, uncoated tubes are also somewhat cheaper than coated ones. Uncoated pyrex flash tubes do emit larger amounts of UV-A than coated tubes, at least between 340 and 400 nm (I have no first-hand data for shorter wavelengths), but pyrex as a material, even when not coated, typically absorbs virtually all the dangerous UV-B and UV-C. The gold UV-absorbing coating present on coated tubes and domes cannot be made so strong that it absorbs all UV-A. A thin coating of this type is visibly yellow-brown in color and cuts also a modest amount of blue light, giving a "warmer" illumination. A thicker and stronger coating would transmit only yellow and red. As a result, coated flash tubes, in general, only reduce the emission of UV-A radiation by a couple of stops, rather than eliminating it. How much UV radiation is actually emitted by a studio flash unit with uncoated tube, and is it any real danger? My practical experience is limited to the longer wavelength range (340-400 nm) of the UV-A. In this wavelength band, my experience in UV photography tells me that a 500 Ws flash at a distance of 25 cm from the subject (i.e., much closer than any ordinary use of a studio flash unit at this power) is equivalent to an exposure time of very roughly 1-2 seconds in full sunlight. Thus, in ordinary use, a human model at a distance of 2 m from a 500 Ws flash unit with uncoated tube should receive 6 stops less of this UV radiation, or the equivalent of roughly 40 ms of sun exposure, and during a long session of 1,000 shots the equivalent of less than a minute in full sunlight (at the wavelengths we are discussing). On the other hand, certain types of photo-damage to the eyes (especially, permanent damage to the crystalline that causes the onset of nuclear cataracts) are known to be due to molecules in the eye tissues that absorb two photons in quick succession. If the time interval between the two photons exceeds a specific length, the molecule can recover after the first photon and no permanent damage occurs. Therefore, a high-intensity flash may cause more irreversible damage than a low-intensity exposure of higher length that involves the same total amount of energy. For this reason, xenon flash may be more dangerous than its equivalent in solar light exposure, and it may be a good idea to avoid any unnecessary exposure to the direct flash emission and its strong reflections. This (in addition to practical reasons, like avoiding being temporarily blinded) is why I developed a habit of closing my eyes before each flash burst at close distance. The tissues of the eyelids completely absorb UV and substantially reduce and diffuse visible light and IR. I also use a snoot to concentrate the light emission onto a small subject area when feasible, especially when using focus stacking, which requires tens or hundreds of pictures to be taken in automated sequence, over the span of several minutes. When I need to stay in the same room, I also shield most of the equipment (photomacroscope, subject stage and front of the flash unit) with a dust cloth before leaving it to operate unattended. Performance in UV photographyThis section contains only preliminary results, albeit quite surprising ones (at least for me). In order to use the 1500Pro in near UV photography, I replaced its dome and tube with uncoated ones (original Bowens parts available for separate purchase). Since the 1500Pro is rated at 1.5 stops more light than the 500R, I was expecting a comparable performance in the near UV. This means, in practice, that for an exposure that requires the 500R to be fired at full power (maximum settings on both Stops and Tenth dials), or 500 Ws, the 1500Pro should be set at 5 on the Stops dial and 0.5 on the Tenths dial, equivalent to 1.5 stops less than full power, or approximately 500 Ws. Note that the Stops dials on the two Bowens models have a different number of click-stops. These settings do produce an equivalent exposure in the visible range and near IR (give or take half a stop or less, which may be caused by slight differences in flash-to-subject distance, slightly different reflectors and slight approximations in the actual settings and their calculations), showing that my reasoning contains no major flaws. However, UV imaging with a test subject shows that the 1500Pro emits massively higher amounts of near UV at power settings equivalent to 500 Ws. So far I tested only one subject, and since the autumn has already begun here it will be difficult to test with more suitable subjects like flowers that display UV reflectance at a variety of UV wavelengths. Nonetheless, these preliminary tests show that the 1500Pro set at 500 Ws produces a UV exposure equivalent to roughly 1.5 stops more radiation than the 500R set at 500 Ws. 
Results are similar at other power settings. In the above figure, the left column shows images taken with the 500R at 250 Ws and the right column with the 1500Pro at 95 Ws. As expected, the image in the visible range with the 1500Pro (bottom right) is darker, because of the lower power setting. The corresponding UV image, instead (top right), is lighter in spite of the lower power setting. 
Further testing (Figure 8) shows that other subjects also respond well in the UV, which is critical for my needs, and in the visible and IR bands, which was expected, since virtually all good flash units perform well in these bands. It seems I found an ideal light source for UV photography. It is possible that the very short light burst of the 1500Pro and its unusual four-electrode tube (including the trigger wire) and electronics cause the substantially enhanced UV emission, compared to the more traditionally built 500R. Fused silica flash tubes - a health hazard Fused-silica Xenon flash tubes are very expensive and can emit massive amounts of UV radiation even at 200 nm and below, which constitutes a serious health hazard. This type of tube is used for industrial applications, like pumping high-power lasers. As far as I know, fused-silica tubes are not used in photographic flash units, but could be found in surplus industrial equipment. |
A flash tube with three internal electrodes (plus an external trigger electrode) is described in