I have a question about recording to the micro sd card.
I just got my starter pack a week ago. I’ve been playing around with them and doing continuous recording. Tonight, I set one of them to just record on events to see the difference. Imagine my surprise when the 1 min event length is time based NOT event based.
A motion event started at 1135pm 56seconds. The 12 second video shows the 12 seconds from this moment for the next 12 seconds as expected. Playback from history on the card has the first 4 seconds only. Instead of saving a 1 min of recording around the motion it saved the recording from 1135pm to 1136pm thereby only saving 4 seconds of the actual event.
Is this a bug? or as intended? Something wrong with my setup? I can’t believe this is the intended result.
I see notes saying it should record those 1 min clips until motion stops. So what happened? perhaps it only saw motion for a second in the last few seconds of the minute and thought it was over? There really should be a span of time that is always saved from before the detected motion to after it thinks it stopped. Personally I’d say at least 15 seconds before motion detection and 30 afterwards. Preferably as a user setting to determine times.
Even if this has no internal buffer (does it?), why would it not save both 1 min clips to be sure it got all the motion when motion was so close to end of the minute. Super easy fix to ensure events are recorded.
Unless of course… its already doing that and mine is not working correctly…
There is a topic where you can vote for recording a few seconds prior to motion. They use a pixel differential algorithm to detect the actual motion. So for example, my cameras catch wasps and bees flying around in my garden. But, they are so fast I only see them leave the screen or right after they actually leave and see nothing.
If my V2 cam experienced the scenario you described, it would record to the SD card two 1-min clips: one from 11:35:00 to 11:36:00, and then from 11:36:00 to 11:37:00, thus capturing the entire event. My camera has captured numerous short events that span a minute boundary, and it saved 1-min clips on both ‘sides’ of the boundary.
What might be happening in your case is that the motion activity within the detection zone was sufficient to trigger the Wyze ‘save to SD card’ algorithm for the 11:35-:36 buffer of video, but not for the 11:36-:37 buffer of video. The detection algorithm is not really detecting ‘motion’. It simply triggers if a sufficient number of pixels (in the detection zone) have changed by a sufficient amount. For event recording to the SD card by a V2 camera, it captures the video stream in a 1-minute long RAM buffer, analyzes the buffer contents, and decides if it meets the ‘motion’ criteria. If so, it writes the 1-minute buffer to the SD card. If not, it flushes the buffer. Rinse and repeat. The SD card will thus contain a series of 1-minute clips each of which contains some ‘motion’.
With this approach, the process does implement a form of pre-record. If ‘motion’ is detected at :45, you’ll get 45 seconds of pre-event video on the SD card. But if motion starts at :05, you’ll only get 5 seconds (since the 1-minute buffers are synchronized to the top of the clock.
And note that what constitutes ‘motion’ for recording to the SD card is different for recording to the Wyze cloud. There are numerous reports on this forum of instances where the cam detected and recorded something to the cloud but not to the SD card. Or vice versa. Evidence suggests that the Sensitivity parameter only effects record-to-cloud detection and not record-to-card.
Longer version is that the camera when set to record events only to the uSD card, retains every minute - HOWEVER, those one minute records are written to the uSD card ONLY if there is motion detected within that minute. In your scenario, apparently there was motion in the 2335 minute so it was written to the uSD card, but there was no motion in the 2336 minute, so the camera did not write that minute to the uSD card.
Thanks everyone. There was a bit of motion at the start of the second min but there was no green box so I guess thats what happened. It saw the beginning bit of motion at the end of the one min but not the end bit at the start of the next min. Hence it never saved it.
Better use of the buffer would easily fix that but until they make that change I guess i’ll set it back to always record so I don’t miss anything important. Nothing important happened in this case, just was wondering why the card only had part of the action while the alert had it all.
Quickly moving from event to event was nice when it was doing event recording only. But with always recording those buttons just skip to the very start of the recording, not to each event as desired. I think I read they may be working on that so that will improve things.
I can’t wait till they do RTSP. 3 days is fine most of the time for the micro SD cards I have in them. It will be nice to have better control and to pull the data to a network drive for longer storage auto purging as needed.
IFTTT started working properly today too. I think I had something not quite right the last couple days, I’ve been tinkering. Its now enables the alerts when I leave but disables them when I get home. So thats a bonus.
I’m still playing. But these are already far better then the crappy Zmodo stuff I have.
There is an issue where the 1st couple seconds at the top of the minute can get dropped from the video file. That issue is being investigated by Wyze.
In theory, even a second or two of motion at the top of the minute should result in the full minute being retained. But if that second or two got dropped due to this issue, that might explain why it wasn’t saved.