I am experiencing a lag between when the camera see motion and when it is broadcast on my Echo Show (and all other Echo’s). The lag is appx 15-20 seconds. I have a Zmodo camera which is spot on with motion announcements. As soon as it detects motion I get a notification.
I am on fiber optic WIFI so speed is not an issue. What’s the deal?
I have my Wyze Cams announce PD on Alexa and also thru push notifications. They are not immediate but Alexa is regularly quicker than the push notifications. Usually between 7 and 15 seconds for Alexa and the push follows within 5s.
But, is the Echo announcing all motion events for the Wyze Cam or just the Person Detection events?
Is the Echo announcing the same types of events for the Zmodo as it is for the Wyze Cams?
The notification lag is the same on Both of my Echo’s, hockey puck and show (they go off at exactly the same time0, AND on my phone. So everything experiences the lag. Lately it has grown to around 30 secs. They are only announcing people.
The Zmodo is in the garage. It is announcing well ahead (instantly) of the echo’s.
The only reason I asked about the type of notifications from the ZModo is because some of the lag on the Wyze Cams is a direct result of the AI Person Detection (PD) routing and process.
I’m not sure how the ZModo handles it’s detection, if it has AI PD, and if that coding is cam based or server based. Since Alexa only announces PD on a Wyze Cam, I will assume that she only announces PD for the ZModo as well.
Wyze PD AI is Server based. This means that all motion activated video has to be uploaded to the Wyze Servers and then ‘intereogated’ or scanned on the server for the specific AI PD object. Once positively identified, that video is tagged and the Push Notification is initiated. It is at the time when this tag is applied that Alexa picks up the PD and announces it. In a perfect world, with blazing fast WiFi and ISP internet, this can happen in a matter of seconds. But, with slow WiFi and ISP speeds, that time can be longer for the video to be uploaded and the positive tag initiated.
Add to that variable the fact that Motion activated videos can be quite lengthy. The object identified may not enter the field of view until later in the video or doesn’t present a good image for the AI server and therefore it delays the AI PD tag until later in the motion event because it takes longer for the server to ID the object.
If the ZModo cam has its AI software resident on the cam, it will be much faster at identifying the object and prompting Alexa to announce it because the video doesn’t need to travel half way around the globe thru the interwebs. The drawback to Cam resident AI software is that it can only be updated by a firmware download and it is limited to the size of the IC memory installed when manufactured.