Wyze camera vs tricycle stealer

Recently a person came to our garage at 5 in the morning and took away a tricycle

This spot was not covered by just one camera, there was a V3 camera pointing straight at the tricycle and a V2 camera that was pointing at the path he had to walk to get to the tricycle. Both cameras had a microsd card. Both were set to record continuosly and both were set to transmit events to the wyze cloud if only the first 8s.

However, the one camera overlooking the path stopped writing to the microsd card back in early june and there was no notification about it. It did not register the person (which was square in front of it, moving towards and away from it) and it also did not send a clip online. So this camera was useless for this event.

The other camera, was completely blurry because of the light conditions, it did record something but not enough to identify the person, it did also not upload a recording but at least that one was recording to the microSD drive.

In this instance I am feeling the impotence and anger that comes when something that belongs to you is taken away from you without right, and I would like to request Wyze to please make the detection/reporting on whether a camera has a microsd and if it is in working condition easier and better.

Only one thing is worse than having no camera to see who stole something, and that is having a camera pointed exactly at where the thing happened and it not working.

6 Likes

That situation does suck. It’s like you got burned twice.

2 Likes

I originally bought a V3 and a couple OGs and returned the V3 pretty quickly. It was too blurry.

I don’t have a V2 but on my OGs and Panv3 when I go into the live view of the camera it does give a notice if there is a MicroSD issue. Has only happened once or twice. I would be nice if the app would send a notification or even warn you right when you open the app rather than when you open the camera itself.

I will say that thieves have gotten “wyze” to cameras and often have a hood up and mask on etc. So even the clearest camera can’t tell you much. But these cloud cams use so much compression, none of them are going to give you crystal clear images, for that you need a higher end camera system.

As far as the V2 it may have just totally locked up (or partially). That’s why I have a rule (now automation) to reboot all my cameras once a week. May be something you want to try.

Upgrading to OGs or V4s (both priced very reasonably) would definitely give you better images (both day and night), and probably reliability since the V2 is so old and may be wearing out. The downside is if you’re using Cam Plus Lite, those cams will not work with it, they only support the paid subscription.

You can also use automations to say turn on notifications (and/or the alarm, etc) when motion is detected in the garage after a certain hour that there normally wouldn’t be. That gives you the chance to intercept or notify the police even if they are all covered up.

1 Like

I check the cameras often to make sure none are offline, which is a common occurrence somehow. I have to go reboot them frequently and, in the most annoying ones, I have placed smart contacts so i can powercycle them remotely. I have never seen a notification that the microsd was not working. I have seen a few where it says that the microsd has no format or can’t be used but none where it tells me that it is no longer recording or that the last recording was weeks ago.

the person was less than 2mts away from one camera and as close as 1mts from the one that did not detect him nor create an event when it was on.

I will be adding that as a task now. to verify that the cameras are recording every week. not just to verify that the devices are online.

Thanks for your suggestions :slight_smile:

2 Likes

You could set a nightly reload, for me that seemed excessive (especially since a couple of my cameras are Pan ones and it is excessive wear on the motors). But for stationery cameras, it shouldn’t be a big deal. I don’t know if Wyze unmounts the SD card before rebooting but I’ve never had an issue with the SD being corrupted or not recognized after a reboot, so I’m guessing they do.

If the cameras are having that much issue it makes me wonder if they have a wifi signal strength problem. If they can’t communicate with the Wyze servers they will show offline and eventually some even stop recording to SD or detecting events. Other thing to consider is what SD cards you’re using, if they’re qood quality endurance rated ones etc.

Some cameras—Cam OG, notoriously—will reportedly continue to write to microSD during an Internet connection loss but after 30 minutes of no connectivity will change the location on the card where it stores recordings, as I understand it. I believe @grapefruityoda has documented this behavior with some Cam v3s, as well.

Is it possible that the Cam v2 pointed at the path lost its Internet signal and started doing this? I’d be curious to learn if anything can be salvaged on a PC from the microSD card pulled from that camera.

Guess I can consider myself lucky that I haven’t had an internet outage in many years, not since fiber was installed probably 10 years ago. Our power is very reliable and everything is on good sized battery backups since I work from home. So other than occasional router reboot or maybe some overnight maintenance (can’t even recall that happening in a long time) my cams have never been offline in the year or so I’ve had them. So can’t comment on how the OGs behave after a longer outage.

Seems a bit odd behavior. I did notice on my distant pan v3 that one time lost wifi and didn’t reconnect for a while, all the events from that time did get uploaded once back online. So maybe it buffers a certain amount of time to upload when it comes back online, but then switches to a different spot that won’t get uploaded when too many build up? You’d think it would still “stitch” that into the timeline but perhaps that was just too much logic to try and build in.

I agree. For what it’s worth, this is a relevant detail from a Wyze employee:

This is why I wondered if pulling the card from the Cam v2 for analysis would yield anything useful.

These are other related posts noting that it’s not just a Cam OG issue (but may depend upon specific firmware versions):

Ah that makes sense, if the cam reboots and has no internal battery (doubt it does) then it will reset to whatever initial date the firmware has built into it, and yes totally different folder since it names the folders using the date.

I have these pretty old dash cams in my car, they have no wireless or GPS connectivity. The way it sets the time is you put the SD card in your computer, in their utility tell it what time it is (I add a minute to allow it to boot up and read the card) and it basically timestamps the config file with that time, which the camera then reads on bootup and uses as its own time. It does have a battery so it remembers that time, just need to do it if the time gets off for some reason (or the first time you install it). You’d figure maybe the Wyze cam could just read the date and time of the previous file, see that it rebooted itself, add a couple mins and be semi-accurate. But I guess there is potential for inaccuracy there, if it lost power during the reboot etc then it would be wrong.

NTP intercept on the Asus is a good solution in that case. It will reply with its own time even if it can’t reach an NTP server (and the assumption is its own time will be accurate, as long as it did not lose power. If it did, you’ll be no better off as they also rely on NTP to set the current time after reload, they have no battery).

I have this ridiculously accurate carrier grade switch with GPS antenna inputs, supposedly a “binned” quartz crystal with like .0000001% drift and certified to Stratum-3E, which goes to Stratum 1 if you hook up the GPS, and even has the old serial BITS timing output. Uses PTP (a step up from NTP). It only gets fired up when I need to test something for work in my lab. Don’t think I need that kind of precision all day every day, but maybe I’ll set it up and for a monthly fee everyone can synch to me :rofl:

The nice thing about working for a large networking company is the pallets of “1 generation back” stuff that they send to the destruction and recycling company. The more I grab off the pallet, the less they have to pay to destroy. This stuff was millions new, now sitting in my house and obviously valued far lower than that. Enterprise networking gear depreciates faster than anything else I’ve ever seen once it goes end of sale (and especially end of support).

The rack can only be turned on for testing, otherwise my electric bill and ears will both go crazy. But nice knowing I have a backup heat source if my furnace dies. Meanwhile my Asus router I bought on clearance for $35 handles all my internet, even though the Cisco/Palo Alto/etc gear is bullying and mocking it all day.

End of tangent/random babbling.

1 Like

So the dash cam utility is basically just doing a touch to a config file? Weird.

This is awesome and also funny because it’s true.

Yup, there are all the other configurations which do update the date and time on the config file to current system time when you change them, but then at the top there is a running clock, you can add or remove minutes, then click the “clock setting” button (made in Korea, not the best translation) and it basically deletes and re-creates the config file with whatever date/time you specified. When I first got it I toyed around with it to try and figure out what it was doing/how it worked, each click just made the file disappear and reappear with the new created date/time. It is actually interesting in that you can set it to an hour in the past (or future) and it will change the creation date to reflect that. So it is definitely doing a bit of hackery with the file properties.

Somewhat ingenious I guess, sometimes the best solution is the simplest.

1 Like

In a previous role I used these huge 13 slot (x48 port per slot) switches with POE on all ports for the IP phones. They used dual 6000 watt power supplies. The rooms they were in on each floor had A/C rated for whatever was standard at the time the building was built, nowhere near what two of those switches were putting out heat wise. It was good enough to keep them within their rated operating temps, but working in that room was not comfortable since the fans were at full jet engine speed and it hovered around 80 degrees. I had an opportunity to take 4 of them but at almost 3 feet high and like 250lbs fully loaded, even I had to turn that down. Not like I have 250V 40A receptacles in my house anyway. Though having a switch capable of 2 terabits of throughput would have been a nice mantel piece (significant mantel upgrade required).

When POE+ and ++ came along, they went up to 9000W power supplies available for them, never had to use those though, can only imagine what they weighed (getting the 6000s in there was hard enough).

1 Like

Good idea. I will have to add that as a task. I don’t have the issue as far as I know, but then …

Something I don’t miss, the never ending baby sitting of Wyze cams…LoL

It’s a shock when you see how much is missed by a Wyze cam when you have another brand in the exact same viewing area. Guess What? I don’t have any issue with my Reolinks…

1 Like

Is this the Reolinks forum? :rofl:

I can’t speak to the older cams, but my OGs and PanV3s if anything pick up too much stuff and I had to get a bit more strict with detection zone and sensitivity.

I know within a few seconds of a package being delivered, and by the time someone is at the top of my front steps and about to ring the doorbell, I already know they’re there (and multiple cameras, even ones far away, picked them up and recorded an event).

We’ve had this discussion here before. From what I understand, the cameras have clocks, just not battery-backed. It does an ntp call on boot so the clock is correct. It’s what I think shows on the liveview.

What seems to be the bug is that when the internet connection fails, the wrong time is used to save the clips. Possibly the code is incorrectly sending an ntp call and failing, rather than using its internal clock.

Well, if the power fails the cameras usually start up before the network is up and connected to the internet. Internet connection does not fail on its own because the internet connection has failovers. I do see the cameras broacasting their own wifi signal for a bit until an internet connection is detected. The microsd is normally seen in the drive but it stops writting, sometimes reformatting the microsd is required for it to be detected again by that camera. I have tested the micrsd for defect or read/write errors and I have yet to find one that has become faulty.

I will keep that in mind, but i would be surprised if just by not getting a connection the writing to the microsd fails. That would be a bit too much.

The cameras definitely have clocks, otherwise the time on the live view would only update when they synch NTP which is infrequently. The discussion here is that apparently some cams reboot after 30 mins of no internet, at which point they lose the time and revert back to whatever their hardcoded default is. Causing them to write to a different named folder (since the name is based on the date).

My understanding is certain models of cams (possibly all) need to authenticate before they start doing anything. However if they’re already authenticated they should write to SD during an internet outage, at least for a certain amount of time. As has been mentioned, check and see if it starts writing to a different folder, apparently that behavior has been seen when the cams have no internet for a while.

SD cards can be difficult to diagnose as they will just mark a cell as bad eventually, even during a format or test write etc. So it can cause an issue once, but then once reformatted, work for a while, until another cell goes bad (typically once the card starts going bad, it will just keep happening). Also make sure they aren’t fakes, reporting incorrect size, as that will definitely confuse the camera and just about anything else that tries to write beyond the true size of the card.