Newbie with general questions...
Hello,
I'm new here and in the trial phase of SecuritySpy. I have a lot of questions for which I haven't found answers yet, or which are more like suggestions and might even be silly because I'm using SecuritySpy in a relatively specific environment. Perhaps the developers or some users here can chime in - sorry in advance for the many questions and thank you for your patience!
- At the moment, the app is still running on a rather old MacMini (headless) under HighSierra with four Axis cams, partly in 1080p, partly in 720p (so that the computer doesn't crash...). If I decide to go with SecuritySpy, I would purchase a new MacMini in the medium term. Do I have to buy the license again, or does the license move with my "account" as long as the computer is online?
- Regarding my application environment: I would like to access the recordings managed by SecuritySpy in my home automation system, which runs isolated on several iPads. Here, practically only a very simple browser is opened, in which I access the web interface of SecuritySpy and then see an overview of the recordings via "Captures". This works very well so far, despite the old MacMini, but the "touch control" is a bit fiddly, and as far as I can see, there is no API to, for example, control the video playback in the browser more precisely. Desirable would be direct http requests, e.g., for fast forward, fast rewind, or also next recording, previous recording. Have I overlooked something here, or does this not exist (which I would understand given my somewhat special application)?
- Is there perhaps an alternative way to enlarge the controls below the currently running video, i.e., "Play", "Back", "Forward", and the progress bar? Something like a layout for iPads?
- This question has probably been asked before, but perhaps something has changed in the meantime: Are there perhaps direct "links" to the recording types or recordings of the individual cameras, or perhaps also a link to the latest recording of a camera? So, for example, something like:
- 192.168.xxx.yyy:8000/cam3/all
- 192.168.xxx.yyy:8000/cam3/vehicle
- 192.168.xxx.yyy:8000/cam3/lastmovie
5.And this is a little embarrassing... I don't really understand the AI. This is definitely not due to the software, but rather either my "scientific" understanding or my poor language comprehension:
a) If I deselect all AI elements (human, car, animal), then all movements will be recorded?
b) What is the difference then if I select all AI elements? Will only humans, cars, and animals be recorded, and other movements (for example, trees and leaves) be ignored?
c) Or is the AI function only about the fact that in the recording overview, the frame that triggered the event is always displayed immediately and that I can sort the recordings according to the AI triggers?
d) If I then completely deactivate the AI elements (checkbox shows "-"), will none of the deactivated elements be recorded at all? If I observe a busy street with this setting, for example, and the checkbox for cars is set to "-", will only recordings of animals and pedestrians be generated?
e) What does "Apply motion mask to AI detection and capture thumbnails" mean? If I select this option, will every movement, regardless of whether it's a tree, leaves, human, car, or animal, be ignored and no trigger activated as long as it is behind the mask? Or the other way around: If I deselect this option, will the software record animals, humans and/or vehicles although they are moving "behind" the motion mask?
f) If the "Motion Detection Sensitivity" slider is set to a high value, for example 90, recordings will be triggered more frequently, and the probability of false positives will increase, right?
How do the values for "AI object detection thresholds" relate to this? If particularly low values are set here, will fewer recordings be generated in general, or is the AI then just more generous in recognizing and assigning the detected objects (human, car, animal)?
Thanks for reading...! 👍️
Comments
I'll do my best to answer your questions in turn:
a. Yes, if you deselect all AI elements, any movement will cause a trigger.
b. If you select all AI elements, then any of those objects will cause a trigger. Other movement that isn't one of these objects (e.g. a tree swaying) won't cause a trigger. It's best to select only the elements that you need here, because the more you select, the more opportunity there is for a false-positive trigger.
c. When applied to Motion Capture, this feature limits what actually gets recorded. So a swaying tree will not be recorded at all.
d. If you hold the alt (option) key on the keyboard and click one of the AI elements, it change to a minus "-" indicator. In this case, only situations where this object is NOT present will be recorded. So for example if you put a tick in the Animal box, and a minus in the Human box, then only animals that aren't accompanied by humans will be recorded (e.g. a stray racoon will be recorded, but a person walking a dog won't be). This is a fairly advanced and slightly confusing feature, it's best to not use it unless specifically required in your situation. The majority of users will simply want to enable the AI element(s) that they need, leaving the others switched off.
e. This option is being phased out because it leads to confusing results. From the next version of SecuritySpy (out soon), everything behind the mask will be totally ignored, which is what users expect.
f. Your statement here is correct.
g. The sensitivity relates to motion detection, whereas the AI thresholds relate to AI object detection. Higher AI thresholds lead to a lower chance of the associated AI object causing a trigger. Leave these on the default values unless you have a particular issue - for example, if you find that you are getting a lot of Human false-positives for some reason, you can reduce this threshold.
Hello Ben!
Thank you very much for your answers. I think I see things much more clearly now and might just venture into getting a license...
I understand that my use case is quite specific, and therefore, from your perspective, it probably doesn't make sense to revise the controls in the web interface. However, I believe you underestimate how well the SecuritySpy web interface is already designed, especially for recordings. It would just be nice if the controls were a little better adapted for use on an iPad. With the SecuritySpy app, I naturally can't control my amplifier, and with the app for my television, I can't control the cameras. Besides, I like to have control over the appearance.
I wouldn't want to trade this
for that
or this
for that
And I think even the "normal" user, without my specific requirements, could benefit from some direct links and slightly larger controls: You could then easily create favorites in Safari on the iPad, for example, as a direct link to the recordings of a specific camera, and then quickly skip through them there.
But I understand, of course, that as a business, you need to offer an app, even though your web interface is actually much better and more appealing, and it's excellent work.
Thanks again for your help, I'm off now to buy a license !
Thanks for your additional feedback, and compliments on the web interface! I understand your point of view. Even though the web interface isn't designed specifically for iPad, you're right that it does still work well on this platform, and perhaps some tweaks to get it working even better are warranted, like larger movie playback controls. We'll also look at adding some shortcut links in a future update.
Hello Ben, thanks for you response!
When I write a lot or am in a hurry, I sometimes have my texts translated by Google's Gemini - perhaps sometimes the language comes across a bit too harsh. So please don't think that I am not satisfied with SecuritySpy when I make suggestions or when I ask questions that you have to answer in the negative.
I think you vastly underestimate your own software! The basic function alone, to quickly and reliably stream all recordings to my MacBook and the iPads distributed throughout the house, is fantastic. Congratulations! And I have to admit that asking for http-requests to control playback of the recordings within an iPad is something nearly nobody needs and of course I have not the faintest idea of how complex this might be...so sorry.
But looking at the aspect ratio of an iPad screen it would be great if there was an option to place something like fast forward / fast reverse and next recording / previous recording below the screen showing a recording (just like the buttons on a CD player). If you need someone to do beta testing - I am here😃
A quick question:
I tried this : 192.456.789.123:8000/triggermd?cameraNum=3 to force a motion triggered recording. This is great when the cam is armed but unfortunately this doesn't work when the cam is not armed. Am I doing something wrong?
To explain my "use-case": I want to trigger a recording from one of the action rules of my Axis cams, but I don't want to arm the cam within SecuritySpy.
Thank you!
Nocko!
P.S.: I am using the beta 6.10b7 (registered)
Don't worry, your post didn't come across as overly harsh, I appreciate your clear feedback.
Recording, either Motion Capture or Continuous Capture, only happens when the respective mode is armed based on a schedule.
If you want to enable/disable recording from a script, independent from regular Motion Capture, then the easiest way would be to set the schedule for Continuous Capture first to armed, and then a short time later to unarmed, using the following URLs:
http://192.456.789.123:8000/setSchedule?cameraNum=3&mode=C&schedule=1
http://192.456.789.123:8000/setSchedule?cameraNum=3&mode=C&schedule=0
For this to work, the Continuous Capture settings should be configured to create movies, with a new one created each time the camera is armed.
More info on the HTTP API can be found via the Web Server Specification page.
Does this do what you want?
Oh - I know the API quite well. Please just translate again:
But I would never have thought of simply turning the recording function on and off again. Thanks!
Now I just need to figure out how to pack that into a single http request. But that's my problem... Thank you very much!
Nocko!