Interesting situation here. Wondering if I can get some input from all of the wyze & knowledgeable folks + @Mavens here in the forums?
Last week, I was visiting the parents and had in the back of my mind to install a spare WCO v1 inside their medium-sized bird feeder. The idea of course being to get customer close-ups… After (3) attempts, I finally got the positioning & aspect where I wanted it… and all dining birds well within the PIR effective area. Since a single cam only covers one side of the feeder, I decided to get a second cam (WCO v2 from the local Home Depot) in order to cover both sides. Positioned the new one similarly to the first but on the opposite side.
The weird thing is that I don’t ever get any motion-based capture events. Trying to determine if the plexiglass behind which both cams are mounted somehow blocks infrared to the point that no motion registers. Depending on the time of day, there is definitely reflective feedback from each opposite side of the feeder, but most of the time the field of view is clear. Weirder still, I seem to get smart alert captures periodically when birds arrive or depart, but those are few & far between. The only time I ever seem to get motion-based captures are when the units actually power on for each daily cycle. I’ve created no rules attached to sensors or anything else that would explain smart alerts to my knowledge…
I have dialed the proximity setting (detection distance) down to (2) since the birds are very close up obviously. Also set the motion sensitivity to (73). I have tried various settings including proximity at (1) and motion at (99) or (100), but results are the same. I’ve also experimented with changing the cool down periods alternating between 1min & 3min. There are definitely several timespans per day when birds are prevalent, but also long swaths when no birds appear.
Wondering if anybody knows the secret ingredient to this recipe?
Might need an @Seapup intervention, but here’s a manual recording of a cardinal bully…
Even though thermal scanners in movies see thru walls, that was my first inclination as well. I have noticed in other posts that folks actually do the opposite of what I did with sensitivity settings and REDUCED it way down instead of increasing it. However, the specific post actually involved a cam outside of the bird feeder itself.
I’ve also noticed plenty of motion boxes when I observe the live feed, however. Since these operate off of PIR, I gather that those are essentially cosmetic for detection purposes.
Actually, I’ve noticed that if I walk in the field of view I do get a person detection clip logged of myself. So I’m wondering if the angle needs to be straight on perpendicular for PIR to work instead of adjacent on the diagonal?
I’m no longer on site, but don’t really want to modify the bird feeder operation itself since the cameras are more of a “nice to have” and not mandatory. An alternative idea I had is to use a circle saw bit to place a few holes in the plexiglass above typical seed levels, within the PIR zone, and at bird level. I’m thinking that that will actually end up capturing what I’m after, but still in the deliberation stage.
About the only other interim plan that makes sense to me would be to observe patterns of peak activity and simply set the cams to manually record during those periods.
Clumsy & battery draining. Wish there was an option to turn on motion based recording for WCO.
First I would try increasing the detection distance, have you tried it at 100 to see if you ever get events for the birds. When you get a chance I think it would help to cut holes in the plexiglass as well.
The angle is more Paramount if you’re attempting to use the AI features. For instance in the past it was noted that person detection could be less accurate if the camera was angled or if the person had approached and an awkward angle. This is something that has very slowly been improved upon with AI updates submitted by the community. But it is still being worked on and refined more.
More than likely your biggest detriment is indeed the plexiglass. I had tried to use the outdoor cameras for similar things and the only time I’ve gotten consistent recordings is when there’s no barrier at all. I actually ended up changing a location so that a feeder was hanging off of a tree and I could Mount the camera to the tree. After doing that having no barrier in between I got plenty of recordings even on tiny birds. I can even get recordings on hummingbirds as long as there’s no barrier. And fortunately for comparison I have a thermal imaging camera that uses passive infrared, although it is far far more sensitive than our cameras, it’s amazing to see how much you don’t pick up by even a thin barrier. Direct contact or direct line of sight is the only way that PIR is truly effective.
As an alternative next time you’re back there, you can take us very small thin board and screw it to the bottom so that it extends five to six inches out on either side and just place the camera on that. That will give you a good view of how the birds would be able to sit and they would probably even be closer to the camera as well giving you some cool angles
Hmmm, so is this setting kinda like a golf score where a higher setting is actually for shorter distance?
How would sensitivity factor into this?
It was an expensive Christmas gift last year, so I actually don’t really want to modify the stock panels. If I go to this route, I’ll probably purchase new plexiglass, cut it to fit, and season to taste with circle saw holes.
We actually noticed last year when we replaced the previous bird feeder with this one that no birds trusted it for roughly 2 weeks after installation. There is concern that any appearance modifications might cause a similar hiatus. Preferable to keep them internally mounted and see if I can improve event capture in this case.
We are planning on some sort of bluebird house down the road and I will factor in an external mount for this at inception.
I think the sensitivity setting is what counts as motion (think size, and how much it moves), while distance is well distance. Not really sure how it can tell distance, maybe just by temperature? If that’s the case the birds are small, so a greater distance could be needed. Either way, if your not getting events, first try raising everything to the max so you can see if it’s cam settings filtering it out or something else.
Yeah, the sensitivity part I think I get. Birds are relatively small, but they are up close & personal based on the mount points in this case. So I figure that they occupy a relatively large chunk of the frame. That said, I have raised sensitivity to max with no effect. Right now it’s in the 70s.
Roger that. I was reading the distance setting more as like a focus with binoculars or something along those lines. Lower values meaning that it focuses closer in and higher farther out. Therefore I went very low on the scale thinking that equates to close-in subjects. Since the low-end actually says close and high-end says far, I feel like I’m on the right track with distance.
I now have another issue in that the WCO v2 seems to bleed battery charge very rapidly even though it’s not capturing any events. Wondering if this Home Depot purchase was part of the original production run that seems to have been problematic. No solar panel or other charging is in play. Both of these bird feeder cams are just straight up battery powered.
12min worth of live stream (WCO v2) is really the only item of significance. That’s within 48hrs. Doesn’t seem like that should have chewed up 11% of battery…
Comparatively, the companion WCO v1 shows 8min of live stream and is still at 100% charge over the same time period. Both cameras we’re fully charged simultaneously.
SIDE NOTE: I have had the issue below several times here in the forums where I choose to quote the entire post and the system removes that from my message. I then go back and add it manually a second time and it sticks. Is that a glitch or by design?
UPDATE: The WCO v2 I deployed here turned out to have a defective battery or other charging trouble. Literally mounted next to a WCO v1 still @ 98% battery that hadn’t been charged in 2 weeks, this WCO v2 is now around 60% after a full charge & hard power cycle earlier this week. It literally lost a point while chatting online to the support rep…
Thankfully, Wyze support is shipping me a replacement under ticket #2471909. Support was fantastic after I furnished my documentation, etc. However, this underscores that there was a large batch of troubled WCO v2 in the wild. Purchased this WCO v2 (starter kit) locally @ Home Depot which might be indicative of early production run. Packaging was green & tan, unlike the white boxes from online orders. Also had to get warranty replacement for the other WCO v2 deployed back at this site back in June and it’s been perfect. Hopefully the squirrely cams from early on have nearly worked their way out of inventory.
Lastly, I’ve obtained some aftermarket plexiglass and will prep it thusly in coming weeks to see if that satisfies conditions for these cams to detect PIR events. Stay tuned!
After (3) attempts, I finally got the positioning & aspect where I wanted it… and all dining birds well within the PIR effective area. Since a single cam only covers one side of the feeder, I decided to get a second cam (WCO v2 from the local Home Depot) in order to cover both sides. Positioned the new one similarly to the first but on the opposite side.
