Why our Smart Video Doorbell will not be able to support HomeKit Secure Video
get.netatmo.com
get.netatmo.com
I think the reasoning may be that with Apple you pay extra to know that something will work well out the box without any issues (yeah yeah, I mean the theory). People don't know what FPS is or anything about internals. They connect the device and the video is not smooth. Looks like something is wrong. It is a good approach not to explain to your users that what they see is actually right, but instead find a solution which doesn't make them feel that way.
You don't need to integrate with Apple, there are plenty of solutions. If you do, then there's a reason you use Apple. I think it's mostly that they make choices for users so users don't have to (and it's actually valuable when done right).
Every instance is specific and has its constraints, but there are plenty of high resolution and fps IP cameras sitting in the sun and doing just fine. Most seem to have at least 2-3x power usage. 2W seems really great. Gopros work somehow. So maybe they thermal design could use some improvements.
Using doorbell as a quick check what's happening out there is also a fair use case which shouldn't require somebody standing next to the camera.
I don't know if this lower supported minimum would make much difference in heat though. Maybe Netatmo should be using a chipset that has a lower thermal draw.
Here's what a teardown of their doorbell looks like https://fccid.io/N3A-NDB01/Internal-Photos/Internal-Photo-pd...
And one of the benefits of HomeKit integration is that Apple gives you a certain amount of video storage. The “closed ecosystem” has nothing to do with your ability to use the app. You just can’t integrate with HomeKit. If you are concerned with automation, if their app supports it, you can still use the automation.
Their app is not as polished as the Nest app, but it is sufficient. From what I understand though, after Nest moved everything under Google, their app isn’t as good anymore either.
What I guess it would come down to is whether or not HomeKit's local processing to determine if a person, animal, or car is in the scene actually needs to see movement over short enough timescales that you actually need 30 FPS.
Seems like it would be much better to let people decide what quality they want from their doorbell rather than mandating it as part of the integration.
I don't want to store 1080p 30fps video all the time and have it consume the entire 50gb that I buy with iCloud.
Looks like it's unlimited only on the higher plan -
For $0.99 per month, iCloud+ will offer 50GB of iCloud storage and support for recording video from one HomeKit Secure Video camera. For $2.99 per month, iCloud+ will offer 200GB of storage and support for up to five HomeKit Secure Video cameras. For $9.99 per month, you get 2TB of iCloud storage and support for unlimited HomeKit Secure Video cameras.
With the popularity of NVRs, there's a niche market for those unsatisfied.
With PoE and a UPS too.
Your front door by definition is on an outer edge of your house. You can run conduit externally or whatever internally. It isn't beyond the wit of man to deploy wiring. If you need your "smart" doorbell to always work then you will get ethernet to it and ideally PoE too to power it. A UPS will keep it alive when the power is out.
I also recommend that it is able to trigger a non smart chime too so if all "smart" stuff is unavailable, you still get notified.
When my home (Doorbird) doorbell is pressed: Separate battery backed chime goes off, Sonos speakers say "Ding dong, there is somebody at the door", TVs get a notification, our mobile phone apps twitter.
It is still possible to miss a doorbell press but quite hard. I do all my IoT stuff like that if possible - important stuff still works without the hub/internet/Home Assistant.
Usually I would try and talk to people over it and they wouldn’t even hear that I was talking. If I turned the volume up so that they notice, the over modulation was so bad that they couldn’t understand me.
I ended up just going with a Ring even though it goes against my cloud stance. At least it works.
The issue is that the proper size wire to provide power to a doorbell is larger than a Cat 5 cable. So if you run Cat 5 to your doorbell, you can't actually use it as a simple doorbell. (Cat 5 is 24 gauge, but you need 16 gauge for a doorbell.)
If you are building a house and you want to leave options for the home owner what do you do? PoE Cat 5? Or standard doorbell wire?
The market seems to have decided that you do a standard 16V AC 1A power to the doorbell, and then WiFi for the data.
See: https://support.ring.com/hc/en-us/articles/360041807292-Sele...
As you said, they exists but are rare (and expensive). It seems to me there is a niche market that nobody is recognizing and taking advantage of.
Or you could run something like this: https://www.cablesforless.com/bundled-cable-cat5e-and-4c-16a...
I googled a bit and it seems you might be able to run low speed Ethernet over untwisted wires if they are not long.
They aren't exactly in widespread usage yet, though.
--
(ETA: And apparently there are even 100 and 1000 Mbps variants now too [0] -- with lower maximum distances, of course.)
[0]: https://en.m.wikipedia.org/wiki/Ethernet_over_twisted_pair#S...
I don't have enough electrical know-how to figure out if this works...
Edit: Based on a table I found; 24 gauge wire has a resistance of about 25 ohms per foot, where as 16 gauge has a resistance of 4 ohms per foot. So; assuming each wire is a parallel resistor, 2 pairs(4 wires) would be about 6.25 ohms -- so it could heat up a bit more than proper wires. so .... maybe as long as someone doesn't hold the button down for too long?
I'm not sure if they are polling those contacts fast enough for 'normal' doorbell use, but it's there ...
Cheap plastic shit is cheap only if you can make lots of them, because the variable cost is low - but as soon as you're talking about small quantities of niche products, the fixed costs dominate as soon as you try to do anything custom physically.
Not available yet but I believe they're aiming for this year. I have other Reolink cameras + NVR and it's pretty solid and doesn't rely on cloud at all
Your on Hacker News after all.
On the other hand I dont feel one bit comfortable streaming any video of my home to "someone elses computer" without end to end encryption.