The following review has bee updated three times since I originally posted it. Each update is identified below. Also, I uploaded a nighttime video that you can view.
I originally purchased a Bushnell 8MP Trophy Cam HD Max, but the trigger sensitivity was extremely poor, and the pictures and video were disappointing. I returned it. I have now used this camera for several days, and have explored all the options and settings. I like this camera much better, and at 2/3 the price of the Bushnell it does just about everything that the more expensive camera claims to do (although the Bushnell does provide more setting combinations).
However, while this camera does the basic job of triggering when motion is detected and then capturing either stills or videos (but not both at the same time), there are several things it could do a LOT better, and there are several important issues that are not mentioned in the reviews at various trail camera sites.
The first thing I tested was how far away the M-990i would detect me as I walked across the field of view. The Bushnell was unable to detect me any further than twenty feet away, which is why I sent it back. By contrast, using the exact same test at the same outdoor location (same time of day and temperature), I was able to get fairly repeatable detection up to seventy feet away during the day. This is quite good.
[update]I placed the camera on a California grassy hillside, in video mode, and retrieved the card after a few days. I ended up with about three hundred twenty-second daytime videos of grass waving back and forth and nothing else. This is the downside of having a sensitive trigger. I much prefer having a trigger that is too sensitive, so I don't mind this, but it does mean that you must take the time to do some tests before you leave the camera unattended for weeks or months at a time in order to make sure that the placement doesn't result in false triggers.[end update]
Still picture quality
The still pictures are a little noisy, but the amount of detail they can resolve is better than the Bushnell.
Like the Bushnell -- and I suspect this is true of most other trail cameras -- the "HD" video quality is not even close to being HD. Yes, it is 1280x720 (720p) at 24 fps (sort of -- more on this below), but the ability to resolve detail is not much better than SD video. I suspect that the sensor is probably not capable of 1280x720, and the video is being "up-res'd" from some lower resolution.
However, one problem that is not talked about in any review, but is quite evident in all the video I've taken is that the camera doesn't actually take 24 frames per second, as claimed. Instead, it takes about 10-15 frames per second, and duplicates frames to get to a 24 fps frame rate. This is easy to see if you put the video into a video editing program and go forward one frame at a time. I have written to Moultrie three times, but after several days have still not received any response -- not even a "we got your email" response. I'll update this review if they ever respond. [update] Two weeks later, and they never did respond.[end update]
You can easily see this for yourself if you look at the videos posted online. I don't know how I missed this before I purchased.
[update]I just watched some video the camera took at night, and the frame rate is even lower than the daytime video, probably about 5 frames per second. I suspect that the camera is actually taking a series of flash pictures and then converting them to video. To conserve power (because the flash uses a tremendous amount of battery power), the camera reduces the number of frames per second. I understand this, but it is even further away from the "24 frames per second" spec than the daytime video.[end update]
Time lapse is very important for my applications.
This camera has a time lapse feature so it will take one still picture every 5, 10, 15, 30, 60, 120, or 300 seconds. However, there are all sorts of issues with this feature that are not discussed in any review or in the manual. First of all, you cannot take high-res photos in this mode. All you get is 1280x720 which is slightly less than 1 megapixel. This is a major disappointment. In addition, the only way the camera will store the time lapse photos is inside an "MLT" file, which is Moultrie's version of a "motion JPEG" (MJPEG) video file. Fortunately, I found out that you can simply rename the file extension from MLT to MPG, and most video editing software will load it, assuming you have a motion-JPEG codec loaded (Panasonic and MainConcept have free MJPEG playback codecs).
I tried using Moultrie's "Plot Stalker" software which lets you play back these video files and automatically search to points in the timelapse where motion is detected (i.e., where something is in the frame that wasn't in the previous frame). Their software is a crude hack that looks and feels like it was done by a high school student. I am a professional videographer and a retired software programmer so I am probably a lot more critical than most, but I am very, very glad that I can use my video editing programs to view timelapse videos instead of this rather poor program.
The next thing about time lapse is that it only runs during the day. It starts fifteen minutes before sunrise and stops fifteen minutes after sunset. The only option you have is to turn it off 1, 2, 3, or 4 hours before sunset and after sunrise. You cannot have different settings for sunrise and sunset; the number of hours you set is used for both durations. You also have the option to leave it on during the entire day, and that is what I chose to do.
Other cameras have a much more configurable setup.
During time lapse, if you have the interval between pictures set to one minute or more, it will take additional photos if motion is detected. However, these additional photos are not saved as JPEGs, but are instead added as additional frames within the MLT timelapse video file. Thus, once you use timelapse, none of your images (except at night, when timelapse turns off) will be high definition.
To put it another way: the Moultrie M-990i timelapse is video, not a series of still images, so all the video specifications and limitations apply.
This mode is the same as timelapse, but instead of taking no pictures at all during the night, the camera will take real still pictures (i.e., high res, saved as individual JPEG files) during the night. I have now used this extensively, and it worked well.
The camera can be configured to take 1, 2, 3, or 4 pictures each time the motion sensor is triggered. The duration between these images can be changed. The fast speed takes several pictures in about a second, so it is almost like video. If you are using timelapse, and motion is sensed, you will get multiple pictures at the interval you set, so unlike the still image resolution settings which are NOT honored during timelapse, the interval settings ARE used.
The "black flash" works amazingly well. I cannot see any light whatsoever (not even a faint glow), and yet it provides excellent infrared illumination for B&W nighttime photos.
The SD card simply slides in and slides out: there is no spring latch or anything else. This makes it somewhat difficult to pull out the card, and it also means that the card might vibrate loose (when being transported) or work its way loose from thermal changes. However, I have not had either of these things happen, so perhaps this is not a problem. But getting the card out is not easy without a small tool or tweezers.
The setup screen shows live video which makes it really easy to set up the camera. I cannot imagine having a camera that doesn't let you see exactly what you are getting before you leave the camera for weeks or months. I really like this feature.
You have to use all eight batteries because this is a 12 volt system. Many other cameras only need four, but then let you load eight or twelve batteries to give you an automatic backup after the first set of four run down. In my application, I am not planning to leave the camera for really long periods of time, but for those who need long battery life, they should make sure that this camera will last a long time.
The manual recommends using alkaline or lithium batteries and does not mention NiMh rechargeables. Since I am taking lots of pictures via the timelapse mode and since the camera is mounted near my home, I opted to try using rechargeables in order to reduce operating costs. It worked perfectly with my Sanyo "Eneloop" batteries, and I was able to get about 1,200 daytime pictures and 200 nighttime pictures before the battery ran down. Since nighttime pics use the flash, that drains the battery much, much faster.
The barometric pressure is not very accurate (I own several real barometers, and the Moultrie is not even close to the correct pressure). However, it does go up and down, so the relative changes may help provide information about pending weather changes. However, I didn't buy this to be a weather station, so this is no big deal.
[update]Several people asked whether this product works with a class-10 32 GB memory card. I have been using a Patriot LX Series 32GB (Model PSF32GSDHC10) without any problems. However, I have seen reports that some people have not been able to get 32 GB cards to work, and that some report that Moultrie recommends using slower than class-10 cards. I have been using rechargeable batteries (NiMH) even though Moultrie does not recommend this practice. However, these batteries can deliver huge amounts of current compared to Alkaline or Lithium one-use batteries. I have a theory (untested) that people who have reported problems are using alkaline batteries and, when taking night photos, the large current required to charge the flash causes the alkaline battery voltage to drop significantly, thus interrupting the write operation, thus losing the picture. For anyone having this problem, try using NiHM batteries and see if that helps. [end update]
The detection range is the reason that I am keeping this camera. It is quite good during the day, and outstanding at night.
I am very disappointed in the way that timelapse works. I am also disappointed in the video quality, especially the very low frame rate and the way Moultrie is apparently "cheating" on their specs when they claim 24 fps.
Finally, being an engineer, here is what I would like to see in a really good trail cam, but which I have so far been unable to find, even in the very expensive Reconyx.
1. Stills and video with quality equal to a GoPro Hero. This camera is within the price range of trail cameras, so this is not out of the question.
2. Much better motion detection. Even if the detection circuitry within the camera cannot be improved (IR sensors have limitations), I'd like to see an external trigger input. One simple contact closure interface would let me use all manner of external triggers, possibly combining them together. Imagine having several wireless sensors, located at multiple locations around a plot, hundreds of feet away, all sending trigger signals to the camera.
I want quality pictures, and I don't want to miss anything. Compared to my ideal camera, the Moultrie M-990i is good, but it could do the job a lot better.