For this wishlist, I’d like the ability to choose to show event thumbnails from the middle of the video, rather than the start
at the start of videos, most often the person is just barely in frame, basically showing an empty porch with nothing on it. Then I have to watch the entire video just to see why the event was triggered.
If you instead had thumbnails from the middle of the video, I could quickly know if the event was triggered by a mail-person, a cat, an unknown person etc., and not have to watch the entire video
It would be great to be able to select middle of video instead of start of video thumnail. Or at least had the option to select “start, middle, or end” of video as the default thumbnail display option
I think the middle of the event could also be problematic. For example, say someone drops off a package or whatever and that only takes 5 seconds, but then other motion keeps the event going for the full 5 minutes, like cars driving by, or branches waving around or whatever else. If the thumbnail is taken at 2min30sec into the event when there are no longer AI detections, that will be even worse than having it happen right at the first time it recognizes the AI detection. At least the AI object is still in the thumbnail in that case. It would also delay notifications by a LOT instead of getting notifications with the thumbnail to us as fast as possible.
What they might be able to do without causing huge notification delays, is something like: Send the notification immediately as soon as the detection is made, and send it with the first available thumbnail…then notifications are still quick, but then have it update the original thumbnail to one that is a few frames later. I know of at least one company that offers notification updates similar to that…they will send an immediate text-based notification as fast as possible, and then later update the notification with a thumbnail. So I guess it is possible to do that. The problem is that it doesn’t work very well. When I have that option turned on, I often never get a thumbnail at all.
I don’t think Wyze will change the way it is done right now though. Their choices are to either delay notifications (something most people would be very upset about, and very much oppose), or else send an update, which means they are more than doubling the amount of work their server has to do for every event…hundreds of millions of events. That adds a bigger cost than most people would think. I am guessing they would decide the cost-benefit doesn’t add up for them.
It would be nice, but I would be shocked if they agreed to do this. Maybe they can just delay it a few frames more without affecting things very much. If they just grabbed a few frames later than the first frame, it wouldn’t even take an extra second, but could provide a significantly better thumbnail image. That might be possible to convince someone on the Wyze Cam Plus team to try out during the next Cam Plus Team AMA.
Event videos have a pre-roll buffer of X seconds, but the thumbnails on the Event tab show the trigger that started the event capture. If you have Motion Tagging turned on, you’ll see a green rectangle somewhere in the thumbnail. Depending on type of cam, thumbnail contrast and motion tagging box size-to-trigger object size, the green box may be so small you may not see it. The point is, rather than guessing at where in an event you wish to create a thumbnail, if the gist is to determine trigger by thumbnail, it already exists. Examples (enlarge/zoom in to see trigger/motion tagging boxes: