Quick update. After some discussion with @joseph_1979 and @KimchiGUN I tweaked some settings and the results are as follows:
1) Increasing the post-detection recording from 8-seconds to 15-seconds and 30-seconds does not prevent fragmenting of the event the video. It only affects the first fragment of the event. The first fragment is 8, 15, or 30 seconds (based on setting) and the remaining event video is fragment at random lengths. I suspect this is a firmware oversight as the post-detection timer should trigger ONLY after motion is no longer detected. The way it is now, you get either 8, 15 or 30 seconds of static video before another function takes over.
2) The default PIR sensitivity is 80. Increasing to 90 reduced some fragmenting. Increasing to 100 further reduced fragmenting however I could not get fewer than 3 fragments even with 2 minutes of continuous motion. First fragment was 30 seconds, second fragment was 48 seconds and third fragment was 46 seconds. Having the sensitivity up that high introduces other issues but I will attempt to overcome those via settings before I comment.
Without a better understanding of how the camera determines when to split video, I can only assume that there is a firmware oversight or a bug.
Argus PT Ultra -- Motion Tracking Version
-
I have had the opportunity to test this model. Just some brief thoughts here and I am happy to answer any questions to the best of my ability.
- Motion tracking with a camera that has to wake from standby mode sometimes does not trigger. In many cases, the object that triggers the PIR is already gone.
- Still no 360-degree PIR.
- When active, motion tracking seems to work best when followed object is no closer than 3-4 feet. If too close to the camera, it looses tracking and returns to Monitor Point.
- Motion tracking is not smooth. I jumps/stops/jumps as the object moves.
- 3W solar panel is inadequate. Perhaps my unit is faulty. I get no increase in charge percentage from the solar panel even though the connected icon is present.
- Zero signal issues with the camera on the outside eve of my deck and the Home Hub Pro on the other side of the house. Full 5GHz bars. App and/or camera desperately need a Wi-Fi signal PERCENTAGE meter.
- Video output is not a contiguous video. For any given motion event, the camera breaks up the video into random length segments. For one of the tests, the motion event was 1:58 and the event was broken into 7 video segments ranging from 8 seconds to 27 seconds each (video is below). I have not been able to determine what criteria it uses for breaking up video.
1View 7 replies-
@chopstix Some observations:
- Indeed battery operated cams have only the PIR and the WIFI ccts which are active at all times so as to elongate the battery charge. So the PIR cct has to trigger the bootstrap to run the application to start detection the object type, check non detection zone, start recording if scheduled and activated, check whether a push/email notification has to be send, etc and this entails for time. So if the object passes fast over the FOV then it will be missed and may not even recorded. So the best cams are the ones powered through dc adaptors and POE which are constantly recording and can provide recording priori the event trigger. But again better than having no cameras. Personally, I do not think that tracking is something to add to battery cams but rather would have exploits other features.
- To have 360 degrees is achieved if you have a full 360 view. This could be two DUOs back to back or having a pentagon type cam with 5 lenses. Nice to have.
- Camera is able to follow one motion and so important actions may be missed. Hopefully that the cam always retrun to its monitor position and no need of frequent calibration.
- I have E1 Zoom, E1 OD Pro and TM. the best motion tracking is provided by the TM.
- Yes, I agree that the 3W is rather low power solar especially if it is triggered often. It needs at least 6W solar panel.
- Can you rotate the antenna downwards and notice if there is any improvement? I have asked for RSSI indication on the camera and some stats on packets such as loss packets, latency, etc. This helps the user identify issues on the WIFI network. The bars shown are nowhere accurate.
- What is the recording time set on your camera? I normally set them to max permissible.
-
@joseph_1979 From my limited use of this unit so far, agreed... Battery cams that support 24/7 recording (always on) are better candidates for motion tracking. I have found that if the object is coming towards the camera, there is enough time to power-up and track. However if the object is moving across the FOV, it has a high probability of missing. I'd like to give the Atlas a try at some point.
As for 360 PIR, they could simply add multiple sensors around the base. If one or more sensors detects motion, the camera rotates to that direction. It might not be that simple, but it works like that in my head.
I am going to have to test the solar panel. Granted it is winter here and the sun is at it's lowest azimuth to me (although I do have the panel angled to match), but even if there are zero detection events in a given day, the battery percentage does not go up. It has been a steady decrease at an average of 1-percent per day. The days that I test the camera, it drop 2-3 percent. I am wondering if perhaps the panel is defective.
I like the specs on the TrackMix but for this area I don't need wide-angle as I have a Duo2 overhead on the 2nd floor roofline. Honestly, I don't even really need motion tracking for this application (I just need to monitor the deck door) but I like the option of manually panning over to the pond and capturing events.
I am having zero Wi-Fi issues. My Home Hub Pro is at the front of the house on the second floor so assuming the antenna has a donut signal shape, this orientation is ideal. But to our points, having RSSI or a percentage would help users fine-tune placement although I think percentage would be more user-friendly. Or why not both? Case-in-point, the HHP is in the room directly above my Wi-Fi doorbell and that shows 2 bars signal strength. No signal issues but I am very curious on he actual signal strength.
My recording schedule is set for always. The post-detection recording is set for 8-seconds and oddly, the first video segment of any event is always 8-seconds. It's almost if the camera is not recognizing that the event is longer than 8-seconds so it stops recording and then immediately starts again. But that does not explain the continuous video fragmentation (at random lengths) for a single event. I am going to increase the post-detection recording time to see if the first video segment time matches.
-
Quick update. After some discussion with @joseph_1979 and @KimchiGUN I tweaked some settings and the results are as follows:
1) Increasing the post-detection recording from 8-seconds to 15-seconds and 30-seconds does not prevent fragmenting of the event the video. It only affects the first fragment of the event. The first fragment is 8, 15, or 30 seconds (based on setting) and the remaining event video is fragment at random lengths. I suspect this is a firmware oversight as the post-detection timer should trigger ONLY after motion is no longer detected. The way it is now, you get either 8, 15 or 30 seconds of static video before another function takes over.
2) The default PIR sensitivity is 80. Increasing to 90 reduced some fragmenting. Increasing to 100 further reduced fragmenting however I could not get fewer than 3 fragments even with 2 minutes of continuous motion. First fragment was 30 seconds, second fragment was 48 seconds and third fragment was 46 seconds. Having the sensitivity up that high introduces other issues but I will attempt to overcome those via settings before I comment.
Without a better understanding of how the camera determines when to split video, I can only assume that there is a firmware oversight or a bug.0View 8 replies-
@chopstix
Since your post, I've been watching my camera's logic. I get very few fragmented recordings, but I feel like it happens when my wife and dog barely step out of range of the PIR sensor and steps back in. Increasing the "Time to Stop Tracking" to 1 Mins for both options seems to drop fragmented recordings a little more on my end.
This is my 1st battery/solar cam in my system. I have all POE cams. It has been a doozy pin pointing these settings for my scenario. -
@chopstix
Forgot to mention this... Watching your video. I think your windchime distracts the camera. I had to disable "Other" Motion in the schedule detection type under recording. To reduce that cam confusion. I have this stupid bush that would distract my cam while it waves in the wind. I wish I could just cut it down, but my wife loves it. it would sit on the bush while my wife would keep walking the dog. -
@kimchigun What to expect with 'any motion' enabled! Any change in pixels will trigger a detection. On all my 20s cams I only enable AI detection.
FYI there are some issues on playback on the new v4.52 Android client. -
@kimchigun Happy wife, happy life something something hahaha. I did have to disable "Other" as well as "Vehicle" since turning up the PIR sensitivity.
-
@chopstix
LOL you are 500% right!
Its 0 degrees out right now... Bright, sunny, and not a cloud in the sky... and not charging. I wonder if the outside temp affects the solar panel? I wiped off the panel with a clean terry cloth rag, it looked "dirty", and no go. Unplugged the USB cord and reseated it, nothing. From 1pm till now (3:38pm) not charging.
It was working the week before until it got super cold. I'm assuming the cold weather is affecting it. -
@kimchigun it’s 8° today however last week it was above freezing, 34 to 38° daily on average with sun and I wasn’t getting charging. Could simply be some quality consistency issues with whoever supplies the batteries. I know of a lot of batteries won’t charge below 40F. Maybe that’s what I am experiencing.
-
@chopstix
Regardless of the chemistry (uh the days where we had to learn all those formulas !!) of the batteries implemented on Reolink bat cams, cold weather will impact their ability to function. Since batteries run on chemical reactions, and as temperatures drop, these reactions are slowed down and increasing the resistance (the electrolyte becomes more viscous and less conductive at low temperatures), limiting the battery’s operating capacity. Even though lithium-ion batteries perform the best in cold temperatures, they still have a safe operating temperature range that should be followed. Additionally, they have a built-in Battery Management System that prevents them from unsafe operating conditions and won’t allow the battery to operate outside the safe temperature range or accept or charge when the internal temperature of the battery drops below 25°F.
Reolink states that their batteries should be charged at temperatures between 0°C and 45°C (32-113°F), and do not use the camera if the temperature drops below -10°C(14°F). But this is the engineering threshold which is lower than the design one. It will operate at more extreme temperatures as customers have witnessed. This, of course, reduces life span and performance.
Here in Europe it's sunny with temperature reaching 18 degrees Celsius. -
@joseph_1979
Ah gotcha! Thanks for the explanation! It being 6F today and sunny, no charge... Now I know why.
So far, this camera has performed to my needs. it has been operating at 0F and -10F wind chills. So far so good. -
@kimchigun Experience is the best teacher and Reolink should learn from these teachers. ha ha ha......
-
@joseph_1979 I hope they are paying attention regularly. As a family, we do a lot of RV'ing. Not a day does by with the big rig that I wonder if engineers ever spend time camping/touring in a motorhome that they have designed. I am active in several branded motorhome groups an while there are factory reps that monitor them, it seems like the information goes nowhere.
-