GHETTO TRIGGER TEST

Introduction

Strobist readers and others have discovered the value of off-camera strobes for good light. It has been argued that traditional PC sync cords are unreliable. Wired alternatives are available, but still leave the photographer tethered to their lights. Pocket Wizards and professional alternatives allow wireless strobe triggering, but are non-trivially expensive to the point that for many photographers, buying them is a choice to forgo purchasing another item of equipment. An alternative is inexpensive radio triggers, produced in China and usually purchased through online auction sites. These devices, typically known as eBay radio triggers or Ghetto Triggers, cost about one-tenth of the price of Pocket Wizards. They do not have the range of Pocket Wizards, and anecdotal evidence from users indicates there are at least some problems with the reliability of these devices.

Where users have reported difficulties with these devices, they have included random misfiring of the flash, failure to fire when a trigger signal is sent, and slower than desired maximum shutter sync times. Difficulties with these devices are particularly reported with Canon flashes. A number of users have proposed solutions to these issues, which have included moving the wireless receiver to less close proximity to the flash (often more than 30cm/12"), modifications to the trigger units themselves, as well as simply changing to another one of the channels available on the units.

This research was undertaken to empirically examine this latter solution. This study asked the question, will the use of some channels result in more reliable performance of a Ghetto Trigger system than the use of other channels? If so, can a particular channel be recommended as likely to be more reliable for other similar setups? These questions were examined in what was regarded as a challenging setup for these devices; triggering a Canon 580EX Speedlite, with the Ghetto Trigger receiver attached directly to the flash unit itself. On the basis of other users' reports, and initial casual use, it was hypothesised that some channels would be significantly more reliable than others.

Method

Equipment

The Ghetto Trigger unit selected was the widely available 16 channel unit, identifiable by the unit numbers: RF-616 on the transmitter, and RD616 on the receiver. The receiver was modified by removing the 6.3mm plug (1/4") on the built-in cable, replacing this with a permanently wired 3.5mm plug (1/8"). This was connected to a Canon 580EX Speedlite using an inexpensive hotshoe adaptor, modified to include a 3.5mm socket to which the receiver unit was connected. The receiver unit was held on the left hand side of the 580EX (when viewed from the rear) using a loop of elastic. The unit was held with the unit's channel controls against the 580EX, and thus the built-in PC sync plug away from the 580EX. This arrangement was used because casual observation suggested that the unit tended to randomly fire more frequently when the PC Sync connector was held directly next to the 580EX, although this was not systematically examined in this study. The flash was powered by four fully charged Uniross Hybrio 2100mAh rechargeable batteries, while the receiver was powered by two relatively fresh Alkaline AAA batteries that had been previously used to trigger the unit approximately 100 times. The transmitter was powered by the battery it shipped with, which again had been used approximately 100 times.

Procedure

The 580EX was set on manual at its minimum 1/128 power, to ensure it could fire as frequently and rapidly as it was triggered; in this mode it can essentially fire continuously. The flash-receiver combination was placed in an open plan living room environment which was likely to have contained a typical amount of ambient RF noise. It was approximately 5m from a desktop computer, stereo system, cordless telephone, answerphone machine, printer, which were all switched on, and a similar distance from a dishwasher, fridge-freezer, oven, microwave, and other kitchen appliances, which were in their typical inactive state.

For each of the 16 channels in the receiver, two measurements were made. Firstly, the unit was triggered ten times using the Test button on the transmitter, at one second intervals. The number of times the flash fired with no perceptible delay after the test button was pushed was recorded. (Random misfires also occurring in this interval were ignored.) Secondly, the unit was left powered and attached to the flash but not triggered for a two minute period, during which time the number of flash misfires were recorded. During the procedures, the transmitter was approximately 5m from the receiver, held in a cupped hand, and no particular attempt was made to avoid touching the hotshoe connector of the transmitter.

Results

Data for both random flashes and missed triggers are presented in Figure 1. In this figure, higher bars represent larger amounts of unwanted behaviour. Therefore, an ideal channel would have no bar for either issue being measured. As can be seen in the figure, a number of channels have this property, indicating both no random misfires in the two minute period, and that all ten triggers resulted in a successful firing of the flash.

What can also be observed in the data, upon closer inspection, is that there is a noticeable periodicity in the data. Every fourth channel provides ideal performance, with no misfires and all triggers firing. Likewise, there is a clear peak of random misfires every four channels, while every second channel displays a high degree of missed trigger signals. However, the most likely explanation for systematic variation in reliability between channels would be if there were particular frequency ranges in which the flash was giving off (higher levels of) RF interference, or in which the Ghetto Trigger units were particularly sensitive to RF interference. While a periodic interference pattern is not impossible, it seems unlikely.

A more parsimonious explanation presents itself. In plotting the original data, channels were arranged in the figure from 1 through 16 as per the original instructions (otherwise in Chinese): 0000 = Channel 1, 0001 = Channel 2, and so on, when reading binary values left to right, i.e. in the order: 1234. However, if the switches were read in reverse order (right to left) and the resulting binary values were used to define channel order, a clearer pattern is observed. Figure 2 presents again the same data as Figure 1, but with channels re-ordered with these new values. Note that to (hopefully) reduce confusion, channels in Figure 2 are still however labelled with binary values read from left to right, to enable comparison with the data in Figure 1.

In Figure 2 the pattern becomes obvious. There is a range of four channels in which random flashes are common, occurring more than 100-200 times in two minutes for each of these channels. There is a varying degree of missed triggers. "Above" these channels (although we do not know in which direction actual frequencies are moving), there is a range of four channels that in this test displayed perfect performance. Then beyond this there are eight channels which have a relatively low (though not insignificant) level of random firing, alongside high degrees of unreliability.

Discussion

The results of this study indicate that there are clearly some channels of the receiver studied that display considerably higher levels of problems, while there are some channels where reliable performance can be expected. This is particularly notable given that throughout all trials the receiver was attached directly to the 580EX. This position has been reported to be problematic in the past, leading to the suggestion that users keep the receiver 30cm away from Canon Speedlite units. This study raises the possibility that provision is unnecessary providing the user selects the correct channel (and, perhaps, ensures the PC sync socket is facing away from the Speedlite unit).

There are a number of limitations to the current study. The performance of only one 16 channel unit was tested, using just one 580EX Speedlite, in one location. It is possible that other individual units might lead to different findings, depending on the degree of variability in the RF emissions and sensitivity of the units. It would be ideal for these findings to be replicated in other settings with other units. However, the stark differences observed in various channel ranges suggests the strong possibility that these findings would be replicated in other situations. The study also examined only one Ghetto Trigger type—a commonly used 16 channel model. It is impossible to predict how other units might fare in a similar test. However, the results clearly demonstrate that the channel selected can result in dramatically altered reliability. A 16 channel model provides far more choices in this regard, and on that basis the few extra dollars required for these units would seem to be well spent. Finally, the study examined unit reliability over a short period of time (only 2 minutes) and with a small number of trigger signals (10). Regular use requires reliability over much longer periods and a more substantive test would be warranted. Having identified reliable channels within the unit will be a useful starting point for any more extended reliability test in the future.

On the basis of the present results, it appears users could select any of the four channels that end in "10" (reading switches from left to right): 0010, 1010, 0110, or 1110; all with equally good reliability. To prevent any residual left-to-right confusion, channel 0110 is recommended to the reader as the easiest to get correct, since it is symmetrical.

The reader is thus left to mull on one final data point from this research... Since the best answer seems to be 0110, it is suggested that the universe must be comprised of six dimensions, rather than the four we usually identify. After all, counting in Base 6, the number 0110 translates to none other than our old friend, 42. :)

— © 2007