Announcement

Collapse
No announcement yet.

Need your opinion on what you would like to do with Snapshots and why.

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Need your opinion on what you would like to do with Snapshots and why.

    I could use your insights on what you want the Ring snapshots for.

    Already a few versions back of this Ring PI, I added the ability to retrieve and store snapshots based on your selection criteria :

    1. In the standard or user defined directory.
    2. Subdirectory based on device ID.
    3. Options: don't store, only when motion, when motion or when camera generates one (could be as fast as each 30 seconds), when motion or on a user defined period (ex. every hour).

    Recently I began coding up "managing" these snapshots and for me a few things stood out (or confirmed my reluctance) quickly:
    1. If you store all snapshots made by the camera, it very quickly becomes unmanageable. For example, I had this function on for a while and gathered 35,000 snapshots and when I was testing some webpages, I had been working on (almost identical to the video clip webpage), not only can you not add 35K elements to a webpage, it takes too long and fails.
    2. If you "chunk" these 35k snapshots say in chunks of a hundred or perhaps one thousand, looking through them is tedious and still not clear what you'd be looking for.

    So here's the question, why do you need this?

    Answering from my own view:
    1. to capture events, but why? The cameras can generate videos when there is motion, isn't this much easier? Now I believe this comes at a monthly cost, is it this we are trying to avoid (have to admit I have no idea if you can operate the cameras without a subscription or what exactly you can do without a subscription). I'm ok this being a valid reason, just trying to get feedback.
    2. Store all these snapshots on a big disk for some off-line processing or for some other purpose. Again, if this is for event capturing, isn't the video clip a few orders of magnitude more efficient?
    3. You like some time-elapse videos, I can see that, but having the same time-elapse everyday, wouldn't that be boring rather quickly. Not to mention, you would have to turn these into a time-elapse video.
    4. If you like a series of snapshots of just the motion event and want to keep those, that can be done with the features included today and that I can sorta see. Having that said, the series of videos would do exactly the same and probably better because taking snapshots when there is motion, seems to be hit and miss with Ring. From my own experience and others that have gone before us, it seems that (more so with) battery operated cameras, when they are recording, they pretty much ignore any other commands, such as taking snapshots.
    5. Could I add support to turn these snapshots into , say a daily video? Well that adds another level of complexity, as I would have to depend on some 3rd pty app to do that, and good luck now to do this on all flavors of Windows and Linux, not to mention licensing rights when i was to bundle this with the PI. Having the end user being responsible to install said app, well I can see how that is going to go.

    Now here is the teaser (assuming you are still hell bent on getting these snapshots ), I found a command to create and retrieve time-elapse videos from the snapshots the camera took by itself. But with everything I experienced so far with Ring, it comes with its limitations and quirkiness as none of these features has been created for API hackers to use, but made to support their apps.

    So here is what I've seen:
    1. So far I've only been able to make these snapshot videos of non-motion events, so what exactly is the purpose (except from using your app on your phone and scroll through time to see time-elapse pictures).
    2. Ring seems to "chop" these into videos of approximately 10 seconds. for example: I instruct the Ring server to generate this video for say, yesterday, the server will respond with an undefined amount of links to small clips, so not one video based on an input time range (say one for each hour or day or something on request) but based on the number of snapshots the Ring server stored and glued together into a ~10 second clip. This presents again for me, quite a challenge how to use this or present this to the end-user being you.

    So what are your thoughts?











    #2
    Currently, I use the snapshot in your plugin to see the current status of a location. However, it would be great to have some version on on demand live video.

    I do not see an option to show the last motion event.

    Comment


      #3
      Originally posted by chad611 View Post
      Currently, I use the snapshot in your plugin to see the current status of a location. However, it would be great to have some version on on demand live video.

      I do not see an option to show the last motion event.
      Thanks for the feedback. Are you on the latest version?

      1. The live video , I will look at it, but that falls under the category: "hard". It falls in the same category I described above where I will need to make use of 3 pty apps (SIP clients) to make this work .... on different Window OS versions and Linux kernels. It is on my list but don't keep your fingers crossed too long.

      2. "I do not see an option to show the last motion event". Care to comment, not sure I understand. If it is video, go to the "history" tab and you can see all events that are caught or available on the Ring server. The latest version, allows you to download the clips, look at them and even takes snapshots when it happens (or sometime around the event, depending on the camera ). If you just want to keep snapshots when motion happens, you set them on the settings page (general tab) and select for example the snapshot action to be "only when motion events", it will store all the motion snapshots, without overwriting them in a dedicated/per device subdirectory. In the version available on line, there is no support to look at these except to go into the subdirectory and use whatever tools you want and do whatever you want with them, hence the purpose of this thread to understand what I should try to support that makes sense to a larger audience. I'm about to post a new version, that can show the snapshots, delete them etc. but if you have thousands of snapshots, this is a totally useless webpage (in my opinion).

      3. If it is about home automation, there is a snapshot feature that will be updated when motion happens (technically when the snapshot becomes available on the camera and could lag quite a bit in seconds). You can use that to trigger an event and do whatever you want to do.

      Thoughts?

      Comment


        #4
        What's the accepted (or supported) method for viewing the snapsnot in the HS4 web interface? I can take the snapshot URL and manually paste it in my browser, but should (can?) this be a link to make that easier?

        Click image for larger version

Name:	image.png
Views:	95
Size:	78.4 KB
ID:	1618772
        Still playing around trying to decide if I want to mess with snapshots and motion recordings in HS4 or just leave them in the Ring app.

        Comment


          #5
          Originally posted by windracer View Post
          I can take the snapshot URL and manually paste it in my browser, but should (can?) this be a link to make that easier?
          Not sure I understand, you would prefer this to be a "clickable" hyperlink? If so, don't think changing the feature status string to a full URL , I don't think it would make it clickable because that would be something HS would need to do. It would indeed be nice if you could click on say, the image there on the left and it would open up in full resolution but that would also be a HS feature, not sure what trick I could pull to make it happen

          Comment


            #6
            Ok, thanks. I just was trying to figure out how people were viewing their snapshots through the HS4 web UI.

            Comment


              #7
              What determines the folder used for snapshots? I have my option set to "motion only" and this is what my snapshots folder looks like:

              Click image for larger version

Name:	image.png
Views:	83
Size:	25.5 KB
ID:	1619263
              The motion folder is empty. The per-device folders have some images, and then as you can see the high-level folder also contains some device-specific images. I'm guessing the one without "motion" in the name was generated by me physically requesting a new snapshot.

              Also, I'm not doing any video downloading (anymore, I did play around with it), but it looks like there are still snapshots being dropped into the videos folder?

              Click image for larger version

Name:	image.png
Views:	67
Size:	39.0 KB
ID:	1619264

              Just trying to figure out what causes images to put in each location.

              Also, for management purposes, it's up to us to purge these, right? If I do so automatically (i.e., with a script), that won't mess up any info in the PI will it? Or do I always need to maintain these from the History tab of the PI?

              Comment


                #8
                Originally posted by windracer View Post
                What determines the folder used for snapshots? I have my option set to "motion only" and this is what my snapshots folder looks like:

                Click image for larger version

Name:	image.png
Views:	83
Size:	25.5 KB
ID:	1619263
                The motion folder is empty. The per-device folders have some images, and then as you can see the high-level folder also contains some device-specific images. I'm guessing the one without "motion" in the name was generated by me physically requesting a new snapshot.

                Also, I'm not doing any video downloading (anymore, I did play around with it), but it looks like there are still snapshots being dropped into the videos folder?

                Click image for larger version

Name:	image.png
Views:	67
Size:	39.0 KB
ID:	1619264

                Just trying to figure out what causes images to put in each location.

                Also, for management purposes, it's up to us to purge these, right? If I do so automatically (i.e., with a script), that won't mess up any info in the PI will it? Or do I always need to maintain these from the History tab of the PI?
                1. not sure about the 'motion" subdirectory, either it was created because you set some path, or it was in created by one of the previous versions (by mistake), but it shouldn't be there, at least not for the PI.
                2. In de snapshot directory, you will find the latest snapshot that was taken by user command. Each time there is a new snapshot for a device, the previous snapshot is automatically deleted, so no need to clean up of manage anything. This directory has the on-demand snapshot of any of your cameras.
                3. In the subdirectories under snapshots, you find a subdirectories named after the ringId. I know some people won't like it, but it is unique, whereas using the GIVEN camera name, this may change or worse, it may have characters in the name that are not allowed in pathnames. In this directory, you find snapshots which were made either by motion or by your setting to take periodic snapshots. These snapshots are NOT automatically deleted, you can do this with tools or using the history page under settings to delete.

                Makes sense?

                Comment


                  #9
                  Perfectly, thanks for the breakdown!

                  Comment

                  Working...
                  X