The camera needs to know when it is in the home position by a bit map image comparison and the stepper home location needs to be set at the privacy location if the images are not similar form the current to the last transit home. I noticed my cameras if bumped out of position ( stepper axis slip). The position and image won’t be the same indicating the stepper has lost home position. If this is the case the camera should be sent to privacy position and home reset and then sent to the home position where it checks the image similarity for the know home. A poor camera mounting and or a significant change in the home position image could present a problem. Homing using a image comparison validation.
And therein lies the problem. They don’t do an image comparison when returning to the home position. It is based solely on movement. That is, it only recalls the X and Y coordinates. If anything slipped in the process, those coordinates will be off. To do an image comparison would be processor intensive. And what happens when your home position has something in it that moves? The two images to compare would never be the same.
Appreciate the responses. I realized I may not have been clear. By “home” I meant the view I set in detection settings. In fact, the camera did return to what might be true home, pointing straight out from the bracket. This of course breaks my detection preset.
The first instance was caused by a power cycle from an outage. I just now cycled the camera, and it did correctly return to my preset view. This is a guess, but I suspect since it could not get back online right away (my internet gear takes a long time to reboot) it just went to the default home. Going to confirm with a couple more tests. Maybe I will finally do the battery backup I’ve considered for a while now for the Router and AP.
If I were designing a system, I would reserve a few horizontal and vertical lines at the edges to be used internally to the camera assembly to optically track position.
No optical recognition needed. It knows how far it’s supposed to be spaced from the stops. It only has to verify occasionally that it’s not confused by returning back to the stops and moving back to home.
Gear lash isn’t a legitimate argument either because the camera always goes in the same steps to the same positions when I manually move them. It never deviates.
This week’s update:
That’s not the problem. The gearing occasionally slips. So going back to the stops and then advancing so many stops would work, it doesn’t work that way. Maybe it should occasionally? I agree that would probably solve the problem. But, it moves, it skips and then it returns to “home” and because it slipped, “home” is off from where it should be.
Now, what many have done to do exactly what you mention, we’ve put in a rule to make it reset once/twice a day. That effectively does what you suggest. We’ve found it helps, but still doesn’t cure the issue because it doesn’t appear to know home is so many steps from the stops. Instead, you set home and then when it moves it counts how far it moves from there, not from the stops. If “home” was indeed X number of steps from the stops, logically, that would solve it as long as it finds the stops on occasion. I don’t think they want to do that tho because finding the stops means it essentially moves until it can’t move any further and resets to that position. That is, it’s hard on the movement mechanisms.
I guess we’ll need to wait and see if this new firmware improves things.
I have my motor set on number five pretty sure that’s what came from the factory and I never have any issues I only use motion tracking I don’t allow the camera to run back and forth all day long like some people are doing
It is pretty subtle on mine, I only noticed it after quite a while. It becomes more noticeable when you have a detection zone set (especially with specific things included/excluded). I also am set to 5 with motion tracking only, and it definitely creeps. Some days more than others. Mine is facing a somewhat busy street so it tracks a lot, which I’m sure adds to how quickly/how much it gets “off”. The one in my driveway definitely takes longer to get noticeably “off”, partially because of less tracking and partially because what it does track is much slower (at least I’m assuming those are both reasons).
That is exactly why I can’t leave the Motion Tracking on with mine. Eventually the Offset Drift gets so bad that it brings the previously blocked road and cars into the included DZ and it ends up being ADHD all day.
Do you also find that the edges of the DZ also appear to just be “suggestive” or that they are more of a progressive fade into the excluded zone?
Mine consistently chase cars moving within the excluded DZ when they near the edge boundaries between the included and excluded DZ. I end up adding extra excluded blocks to the DZ which then reduces the effectiveness of the DZ since I now have to block out extra FOV real estate that I wanted monitored.
I’m still getting adjusted to detection. It seems far more sensitive to stuff closer to the camera (a small bug flying by will trigger it when a package being tossed up on my steps won’t) and stuff in the middle of the FOV seems more “detectable” too. My pan I’m just excluding a tree in the middle that blows in the wind so can’t speak to that one specifically, but my OG I have a pretty specific zone set and I haven’t seen it trigger on anything outside that zone, it does seem the smaller the zone, the less sensitive. Still toying with trying to get a good happy medium between false positives and detecting stuff I want it to detect. I can tell it won’t be perfect ever but want to minimize the noise.
I am not experiencing this on any other cam model either. Just the PanV3.
Yeah honestly I haven’t played much with DZ on the two Pan v3s, waiting on the issue to possibly be resolved, so just have some crude ones set. Mainly so it will return “home” after moving manually but that hardly ever works either when track motion is on. It waits until it tracks some motion then goes back to home.
My OGs seem to have the opposite problem, it seems you have to set the detection zone a bit bigger than you want or it isn’t sensitive enough… But I’m still toying with them to see which is better, very small DZ with 100 sensitivity, larger DZ with 50 or 60, etc.
My WYZE Cam Pan V3 has a mind of it’s own.
It cannot hold a position. It wonders all over the place. When it stops , it usually is zoomed in to a view it has chosen on it’s own. I have reinstalled app… Changed setting. Extremely frustrating considering I bought this camera to help with a problem that needed eyes on.
I don’t know what I can do to keep this camera focused on one area.
The camera doesn’t zoom on its own. Maybe time to change your password, or maybe there is some glitch with the app where it is sending commands to the camera?
If you have motion tracking and pan/scan off it should not move on its own at all. If you have them on check your settings, maybe you have waypoints that are messed up or sensitivity is too high.
One of mine will follow bugs that are attracted to the infrared lights at night, but turning the sensitivity down helped.
You have it shared with someone else? They probably moved it manually and left it there.
My V3 just recently got all crazy, I believe after the last update. No detectives n zones or anything turned on, I have 4 locations set, one to the sidewalk to the house, one to the driveway, one to the road, and one ti cover the neighbor’s yard for them. As of Wednesday it will not keep the locations. Everytime I go to view the camera it’s pointing back at the window, it’s almost like it turned all the waypoints 180⁰. I’ve reset the points multiple times now and it keeps doing the same thing. I’m going to reset the whole thing and re-add it today and see if that helps, if not just contact support and get some advice there. They’ve worked great since I got them months ago until just this week.
(update) 12/1/2023 found out that the plug in wall cord instead of the USB cord that comes in the box has been the problem all along. I concluded that the charging cable that is included in the box is deffective or does not provide an adequate source of power for the unit, I have a new one up and proceeded to setup a waterproof powerstrip and extended it out towards the camera, bought the plug in in wyze"s shop with the Lshape adapter and so far no wifi connection problems. My previous assesment of brick walls blocking the signal was all wrong. I have tried 25 foot usb runs, 25 feet usb with power adapter runs and nothing had enough compatibility.
Old post: Hi everyone, I have been testing this feature repeatedly weekly in every whichway. There is a bug; no matter how low of a setting I set this, it chases the leaves and bushes. The camera sensor can not tell what a person or a brush is because it tells me that a person is a bush, and sometimes it is a person or an animal. I set this up 20 feet away, and the Wi-Fi on this unit is horrifying, especially if you live in military housing townhouses with concrete walls and steel doors; I set up a repeater outside on a 2.5 megahertz, but it’s not responding to it, then I added a 20 foot USB cord extender adapter with a plug in the adapter wall outlet to boost and that worked until it didn’t. So far, this is my second unit that has gone bad. These cameras are not worth putting outside! I think they are great indoors or on your patio, away from the rain or snow. They are the worst cameras for outside weathering the elements. That is my conclusion so far after one motor locked up on me and the other the Wi-Fi doesn’t reach my router. My router is the latest router that money can buy. My modem is the latest modem that money can buy. Neither the modem nor Router are from Cox cable. I prefer to tweak my own system so I am in control.
I having only but troubles, i have 5 different ones, v3 pro, pan can v3 etc.
I reboot them every night. Turning off the mesh wifi it works a bit better, even installed a wifi repeater outside and one meter away they have only 2 bars range, all settings are off for motion tracking.
The camera sensor and the firmware on the camera is incapable of identifying any object. PanV3 cams have no onboard AI. They only see motion thru a change in the light\color\shading pixelation in the FOV. Your sensitivity setting only determines the relative degree of cumulative pixelation changes required to trigger the upload to the server where the Server AI Engine then interrogates each still image frame for an object that resembles one of your selected objects.
Because the AI Engine does not measure or recognize motion, it is possible that something within your FOV, usually not the actual object ‘moving’ or the one marked by the green Motion Tagging box if you have that enabled, resembles the shape or outline of the object being tagged.
The AI gets updated regularly. When it does, it sometimes starts tagging objects and shadows that were never previously tagged. This starts to slow down after users submit training videos for the next update.
Yes. The PanV3 is very succeptible to connectivity loss due to a poor signal. Moreso than any other cam I have tested. But, it only works on 2.4GHz. Not sure if that was a typo or the wrong freequency. Also, the cam has an affinity for the root router to which it is initially attached. If you install a new extender\AP, it may take a reboot of the network, the cam, or a reinstall for it to connect to the extender\AP.
Not sure what that is. Power or EoP WiFi?
I have all four of mine outdoors in the elements since installing them when they released. The outdoor rating isn’t the issue.
The Pan V3 has 3 weaknesses right now: Antaenna is far too weak, it is highly succeptible to low power glitches, and the motor gearing lash \ home calibration is faulty.