Print View |
Note 26/05/08: Someone has suggested that the start-up delay, and the delays between photos that I mention below, may be due to using a slow SD card, and not the camera itself. I haven't checked this yet.
I was pleased to receive my first trail camera (shipped from the USA) a few weeks ago (in March 2008). After considering various makes and models I settled on the Moultrie GameSpy I-40 infrared still/video camera.
We have a small hole in our back fence, and I know the cats use this to get in and out - so what better place to deploy this camera during its test run, than on this spot?
The GameSpy allows you to name your camera, so welcome to Charlie's world. You'll see his name on the info strip printed at the bottom of each image - you can turn this feature off. Apart from the camera's name, the info strip displays the date and time of the image, the temperature and the moon phase.
Camera options and preferences
Still photos versus video
The basic choices here are whether to shoot still photos or motion video. If you choose still photos you can opt for 1, 2 or 3 shots at a time, with the caveat that there is a 13 second delay between shots. (I have since seen a camera model with a delay of less than 1 second between shots). The second choice is image quality - low, medium or high.
If you choose to shoot video, the camera still creates a thumbnail image - and this appears to be a separate photo, either before or after the video (I'm not sure which), but it is not a frame capture of the video itself. With video you have two quality choices: low or high. You can also choose the video length which is (from memory) 5 seconds, 15 seconds or 30 seconds. It seems that regardless of your choice, night-time (infrared) video is always of 5 seconds duration.
Regardless of whether you're choosing still photos or video, you also have to choose how much delay to allow before the camera re-arms. Options include (from memory) about 1 minute, 5 minutes, 15 minutes or 60 minutes. In my case the deployment was not baited, but on an obvious animal trail - so the 1 minute reset was appropriate. I would imagine that if you're using a bait (to keep the quarry in field of view) then you'd want to delay for at least 5 minutes before re-arming the camera.
Multiple photos per trigger
For the same reason, I found that a single still shot was sufficient for the purpose of identifying the cat which passed the camera. With a 13 second delay between shots I very rarely captured a cat in shots number 2 or 3. Again, if I was targeting a baited site, then 2 or 3 shots - even with the delay - would be appropriate.
Infrared versus visible light
The camera operates in both daylight and at nighttime. During the day photos and videos are captured using visible light. At night the camera emits infra-red light which produces an image rendered in black and white. The camera has a light-detector to determine which mode to use.
The infrared flash emits a soft red glow. If the camera is taking still photos, this flash is brief, but if the camera is shooting video then the light is continuous. This has the small potential for attracting human attention should humans be looking towards the camera at that time. My understanding is that the vast majority of Australian mammalian fauna will not visibly respond to infrared flash.
The benefit of infrared flash over white-light flash from an ethical standpoint is made clear when you consider that prey species are dazzled for up to 20 minutes by white-light flash. White-light flash is unlikely to cause any permanent vision damage to nocturnal wildlife (after all, nocturnal species do quite well in daylight) - but rather it increases the risk of predation as the animal is less able to detect predators and less able to negotiate escape routes. However, the infrared flash adds to the cost of the camera also.
When the camera is triggered, an audible sound is emitted. During the first trial, cats moved past the camera while I was still watching the setup and they clearly heard the camera as it was triggered. One of the early videos shows one cat investigating the camera (which was at ground height). Being city cats they very quickly got used to the noise and ignored it.
It is important to note though that the camera operates silently until triggered, making it less likely to deter cautious species. Please note, however, that I am unable to test for sounds outside human hearing range and there is always the possibility of some sound being emitted during the operation of detecting changes in the background infrared signal. (If anyone has insight, please do write).
Backtracking a little, the first thing that needs doing before deployment is installation of the batteries. Here I was a little disconcerted because numerous times as I opened and closed the battery cover, the o-ring seal worked its way loose. This was only detectable when the cover itself wouldn't move correctly into place; in other words, only when you were actually in the process of fatiguing the rubber ring. The USB connection and SD card slot can both also be found under the battery cover - thus, opening and closing this compartment will occur as frequently as you check the camera for images. With care, however, damage to the seal can be avoided. I would rank this seal as an area for improvement for this model.
Laser aim and test mode
Apart from the battery compartment, the entire front face of the unit lifts up in order to access numerous controls and buttons. A slide switch allows for 3 positions: on, off and "aim". When set to "aim" a laser beam is emitted in order to target the camera, however it must be said that in sliding the switch back over to the "on" position, and in closing the cover, the camera may well move from its intended target. One useful mode is in testing the camera - this allows you to move through the target area while you watch an LED (light) on the unit. If you're within trigger-range the LED lights up. Again, when you close the camera you may adjust the targetting.
Attachment strap and angle-adjustment
The most criticised feature that I could find in online forums was the angle-adjustment arm and the attachment strap. The angle-adjustment arm is simply a metal rod protruding from the back of the unit, the length of which can be adjusted by turning a dial. This pushes the top of the camera away from the anchor point (such as a tree) thus angling the camera downwards. The range of movement here is minimal and I was able to work around this tedious control by simply wedging a stone in behind the top of the camera. Secondly the attachment strap is very basic but in my opinion is quite suitable for the task. (I think the criticism is a little harsh - after all this is probably the cheapest infrared unit available, and this is a non-essential feature).
Smart camera placement
Even these basic tests performed in the backyard revealed some techniques that should be followed (probably even regardless of camera model) in order to obtain the most useful photographs. On the first night the camera was positioned so that it directly faced the rising sun. There was dew on the ground in the morning, and this included being on the lens cover, thus the early morning photos showed almost nothing but white. Certainly they were useless and no animal could be identified.
The dew may be overcome by raising the camera off the ground. On subsequent nights I attached it variously to a gum tree trunk and to fence palings but in these tests I wasn't as much concerned with daytime photos as night-time. If possible, the camera should be placed in a south-facing direction. This ensures 2 things: firstly the photo is not washed out by direct sunlight, and secondly the sunlight will illuminate the subject and reflect back towards the camera, leading to better quality images.
Secondly, the camera seems to have a significant delay between being triggered and taking the first photo. I'm sure this could easily be measured, but I have not done so. Rather, during most nights the camera was placed such that the main thoroughfare of the wildlife (cats) ran from left to right across the frame. In every shot where the animal clearly continued walking along its path, it appears in the second half of the frame (and sometimes partly off-frame). The only time the animal appeared in the first half or centre of the frame is when it paused for some reason (and this can be seen by its posture).
I had read this advice on a deer-hunting website - to position the camera so it looks along a trail, not across it. Again, this point is relevant only for monitoring trails. Other suitable targets include mud wallows (for deer) or baits, both of which tend to attract animals to a central location.
The start-up delay, when combined with the 13 second delay between shots mean that for my purpose of photographing a small target species such as the Eastern quoll, the single-shot, 1 minute-to-re-arm setting is appropriate for still photographs.
Please note that I am really only reviewing the infrared night-time shots and video.
Deployment 1 - Low res video at ground level
It seems most trail camera reviews are written by people who are hunting deer. I could not find any information on how well these cameras perform for small target species. In my case I intend to deploy this camera in search of Eastern quolls on the mainland of Australia - a carnivorous marsupial about the size of a house cat. This first deployment simply had me set the camera on the ground, cover it with some branches and make sure it's pointed at the kitty-cat highway in the yard.
This trial used low-resolution video. I was surprised to find that despite choosing the 15 second video length option, the camera shot all night-time (infrared) video at 5 seconds only. This makes the camera much less useful than I would have hoped for filming nocturnal species.
- Video 1 (549k) - This video shows the quality of the low-res setting well enough. You can see there is enough pixellation to obscure detail on the animal (and in the background) but overall the quality is good enough to identify the species.
- Video 2 (552k) - In this video the cat came up to sniff the camera. Because of the proximity to the flash, much of the detail on the animal is washed out. This can be overcome for terrestrial species by mounting the camera above the ground but is more difficult to overcome if attempting to detect largely arboreal species.
- Video 3 (552k) - This third video illustrates the effect of the start-up delay; the cat has most probably entered the frame from the left (based on my knowledge of this cat's usage of this route) and is shown only briefly before disappearing behind the cubby house (which in the wild might as well be a shrub). The major portion of the video shows no animal.
Deployment 2 - High res still photos at ground level
For the second night I opted to investigate the best quality photographs, taking 1 shot at a time.
- Photo 1 (click to enlarge) - demonstrates the under-exposure in the foreground, caused by aiming the camera up towards the sky. The timestamp indicates 6.23pm, about 1 hour before sunset and facing north-east. This underexposure could probably be avoided in the field if the camera can be placed under a tree canopy. However it may be a real issue for more open terrain and is certainly a criticism made of earlier Moultrie models on other review websites. More (daylight) testing is required in order to determine the frequency of this result.
- Photo 2 (click to enlarge) - demonstrates the effect of the start-up delay between triggering and firing. Again, this should be avoidable by smarter camera placement: higher up and not facing across the trail but rather, along it, or using a bait or natural attraction such as a mud wallow.
- Photo 3 (click to enlarge) - in this photo the cat is fairly close to the camera. It is full stride, walking across the frame and motion blurring is very evident. This photograph leads me to wonder about the merits of trail cameras in identifying cryptic species; if this were a thylacine, would the evidence be believed. That is a matter for another discussion, however even with the animal only 2 or 3 feet from the camera the image is at once less than acceptable by some measure, yet also reasonable enough to identify the species. It leads me to wonder how readily a mammal the size of a cat may be identified at a distance.
Deployment 3 - High res video from tree trunk
On the third night the camera was placed about 6 foot off the ground, on the trunk of a gum-tree, facing across the yard. From this position there were no triggers by cats although I can't say whether this is due to an absence of cats or the camera not detecting their movement. The camera certainly was triggered by my taking a bowl of scraps to the compost bin - but again the start-up delay is evident. Earlier videos on this night showed less static while a floodlight illuminated the scene from off-frame.
- Video 4 (1.93mb) - A high level of static in the dark of night; trigger-delay is evident; the camera was not triggered by cats on this night.
Deployment 4 - High res still photos from fence
Back to testing high-res photos, this time the camera was placed at the top of the fence palings - approximately 5 to 6 feet off the ground. During this deployment the camera was configured to take 3 shots each time it was triggered. In the majority of cases, shots 2 and 3 didn't show the animal, or only showed a tiny portion. This was due to the animal moving off-frame during the lengthy 13 second delay between shots.
- Photo 4 (click to enlarge) - this photo shows the level of detail available for an animal at mid-distance from the camera. My legs are in the frame, to given an indication of distance, size and perspective. At a guess the cat is about 10 to 12 feet from the camera. The resolution is sufficient for an ID but again the animal is stationery.
- Photo 5 (click to enlarge) - Although the animal is closer to the camera, it is less distinct because of motion blur. When you consider that the size of a cat is comparable to an Eastern quoll, Tasmanian devil or juvenile thylacine, you can get a good idea of the difficulty of obtaining a concrete ID from motion-activated infra-red photography.
Even with the motion blur, however, this is sufficient to identify the species. By far the more common issue (not shown in the images presented) was the animal being cropped, partially out of frame. Thankfully this is a scenario that can be largely overcome (as described above) but again, this remains to be tested in the field
Deployment 5 - High res video from fence
In this last deployment, high resolution video was again used but this time from the fence paling. On this night cats did trigger the camera, but again the depth of the field of view was much less than on deployment 3.
- Video 5 (1.93mb) - this video was typical of most triggered on this night: the cat is midframe when the video begins, quickly walks off-frame and the majority of the 5 second video does not show the animal. One clip (not shown here) showed the animal for an even shorter period of time and begins in fact with only part of the animal visible in the frame.
- Video 6 (1.93mb) - Here is perhaps the ideal video clip: the animal appears will lit, is clearly mid-frame, looks towards the camera and remains in the centre of the frame for the entire duration.
- Video 6 (1.93mb) - I included this clip for two reasons. Firstly, it shows the stripes on this cat. One might expect that a thylacine's stripes would have a more solid appearance than the stripes on this cat, so hopefully this camera is adequate for showing this important feature. Secondly, the cat is partially obscured by plant material. It is likely that in the wild this would be observed more frequently than has been seen during this backyard testing. With the limitation of having only 5 second infrared videos it is useful to see the effect of foreground vegetation.
When you take into account the price of the unit (being the cheapest infra-red flash camera I could locate) and features included, this unit does present as excellent value for money. From other forums I have learned that battery life can be expected to reach 5 months which by all reports is impressive. The camera takes SD memory cards, so 4GB of storage is the theoretical maximum.
There is one minor build quality issue to speak of (being the loose seal on the battery compartment) which can be worked-around if used with care. Others' complaints about the attachment strap seemed ill-founded - the strap is adequate.
The 13 second delay between shots is a drawback for the kind of photographs taken during testing - that is, of moving animals - especially when other units are able to deliver 3 consecutive shots 1 second apart. Nevertheless, the option for consecutive shots is there and the delay may in fact work in favour for other species and environments.
Image quality is adequate to identify species in most shots, but as with all modern digital photography, would probably not be sufficient to constitute proof of existence. However, images of cryptic species captured with this camera should be sufficient to prompt further investigation in the area of the sighting.
The trigger-delay works against this camera also and results in the need for careful camera placement. Over and under-exposure during daylight hours also result in the need for careful camera placement and taken together these requirements may prove restrictive in the field where positioning choices are limited. Again, this testing was conducted with cat-sized marsupials in mind.
At numerous points I have noted that baiting would likely increase the chance of good captures. I must point out that baiting for the purpose of photography of native mammals is most likely illegal in all jurisdictions if a research permit approved by an authorised ethics committee is not obtained. On the subject of ethics, the infrared flash stands this camera in good stead but adds to the cost.
Overall the vast majority of still and video captures were adequate for identifying the species except where the camera was placed on the ground and the animal walked past it before it triggered.
My own experience with trail cameras amounts only to this unit and readers should bear this in mind. Other forums suggest it is worth investing an additional $100 to $200 in a higher-range model (from any of a number of manufacturers) but it is perhaps unfair to compare this entry-level camera with anything costing nearly twice as much.
I am certainly happy to have the camera and happy to have tested it. I look forward to its first deployment in search of Eastern quolls in New South Wales and stay tuned - there is certainly more to come this year as regards remote camera placements!