Search results

  1. J

    Home Assistant VERSUS Hubitat

    I've been using a now ancient, C5 generation Hubitat since January 2020 (just barely pre-Covid). The choice at that time seemed easier: HA was (and perhaps still is) the hard core DIY/programmer's choice. The latest Hubitat C8 Pro seems like an even more capable system, tempting even for my...
  2. J

    Chuck Norris Gone....

    Cute, but his actual birth date seems to be March 10, 1940. Still one of the greats.
  3. J

    Problem configuring a simple zone crossing

    At least for testing purposes, try turning off the AI analysis/confirmation just to establish a known baseline that the basic trigger works as expected. I would also lower the contrast significantly in the trigger (I generally run about 17 -18 in the daytime, lowest setting of 12 at night), but...
  4. J

    Has anyone tried Homey Pro?

    No, but thanks for bringing it to my attention. Very complete platform, but it should be for $399 starting price for the Pro version ($199 for the Pro "mini" that strips out Z-wave and a few other items, so a nonstarter for me). I think I'll stick with my now ancient C5 gen Hubitat.
  5. J

    BI v6.0.2.7 built-in AI not confirming small moving objects

    But even more importantly, change the timing between each one further apart, as I indicated above.
  6. J

    6.0.3.1 - Broke my webserver

    Has anyone tried today's 6.0.3.2 to see if any (most?) of the problems/concerns have been addressed?
  7. J

    BI v6.0.2.7 built-in AI not confirming small moving objects

    I would say that your settings are tuned to fast moving objects (vehicles), not people or animals. Try changing them to 3 pre-trigger and 10-15 post trigger images spaced apart at 750ms or 1 second each.
  8. J

    6.0.3.1 - Broke my webserver

    Please elaborate: how many things will likely be broken by this? Pushover notifications as well? Can I presume you have sent off a support request to Ken?
  9. J

    How Many Cameras Do You Actively Manage?

    The problem with this poll is that it will automatically and inevitably skew towards the higher counts: those with modest numbers will tend not participate at all (fear of being mocked as a n00b, nothing worth bragging about, etc...). It won't really represent a random sample of this forum...
  10. J

    Internal AI vs CPAI performance

    The "AI Inspection" tab (same set of tabs that include the running log, camera stats, CPU and RAM usage, etc...) has a set of stats in the lower left corner with (overall average) analysis time and total number of analyses since BI was last restarted (or a manual Reset has been done in this tab...
  11. J

    Periodic Scan

    OK, I guess it depends on whether the above scenario happened frequently, or was a one-off false alert. I have very occasionally had these sort of fluke detections, but can dismiss them as long as they aren't a constant daily problem. Otherwise it is normally covered already by the AI zone...
  12. J

    Periodic Scan

    Happy to hear that a fix has been released, but it still begs the question: what is this for, and what exactly does it change? How can we make use of this new feature?
  13. J

    Periodic Scan

    I tried this version for a few hours yesterday and wound up reverting, as I was getting Pushover alerts for zones that I wouldn't normally. Honestly, it is getting to the point where there are way too many places where we have "zone controls" and it is becoming difficult to follow all the...
  14. J

    BI 6.0.2.7 - very odd alert behavior

    I fully agree: a reboot shouldn't have been necessary (for all the reasons you mentioned), and yet...somehow...I had different results entirely after a reboot.
  15. J

    BI 6.0.2.7 - very odd alert behavior

    Oddly, I reinstalled 6.0.2.7 today, but this time I did an immediate reboot following the installation and all seems normal this time around. One-off fluke on Friday I suppose.
  16. J

    BI 6.0.2.7 - very odd alert behavior

    Yes, I had enabled periodic scanning (and checked the box to "ignore static objects detected during periodic scanning" ) a day or two prior, in an effort to cut down on other false detections of truly static vehicles shown as moving. I seem to have fixed that with this combination, but...
  17. J

    BI 6.0.2.7 - very odd alert behavior

    I tried updating BI to 6.0.2.7 (from 6.0.2.6) yesterday, but wound up reverting in about 5-10 minutes as it was acting very odd with the alerts. The version notes indicate that it added "AI object counting", and associated Macros, but little in further explanation or suggestions for how to use...
  18. J

    Blue Iris Version 6.0.1.24 Now Supports Custom Models

    I'd suggest sending this info (JPEG And .DAT file) to support (Ken) at BI. He needs samples to show that whatever he changed about 2 minor releases back has flipped the situation around for the worse.
  19. J

    Blue Iris on Linux

    I agree that WinBoat looks like an interesting project that could simplify running (some) Windows apps under Linux (probably better than Wine), but yeah, running BI in there just sounds very likely to be slow. Too many layers of "box inside a box" virtualization for something that needs near...
  20. J

    Blue Iris Updates - Official Thread

    You do know that this thread is really meant to be just the automated script/bot giving us the update info notes, right? Manually posting "another beta is up" isn't needed.
  21. J

    Blue Iris Version 6.0.1.24 Now Supports Custom Models

    Seems like 6.0.2.3 may have only been released for a few hours, 6.0.2.4 was the only available "new" release yesterday afternoon for me. In any case, I am now seeing essentially the opposite problem: static (parked) vehicles being identified as moving; although it might be partially due to...
  22. J

    BI6 AI detects object but still reports "nothing found"

    Keep in mind that even your canceled alerts will all have a minimum duration of the 15-18 seconds you specified, so it will use slightly more storage per day than a shorter analysis time, but I haven't found this to be a problem as long as you keep false triggers down to some reasonable number...
  23. J

    BI6 AI detects object but still reports "nothing found"

    Yes, that should work (I am assuming that boat isn't one of the named objects, but you might want to double check, or use something like giraffe). Your settings now basically mimic my own, with 3 seconds of pre-trigger, and 15 seconds of post-trigger analysis.
  24. J

    BI6 AI detects object but still reports "nothing found"

    The break time is not directly related to AI analysis time, it is a separate consideration. In your screenshots, you effectively have 2 seconds of pre-trigger analysis (6 x 333ms) and AT MOST 6 seconds of post-trigger analysis (18 x 333ms), but subject to shorter (less time post-trigger) if AI...
  25. J

    Blue Iris Version 6.0.1.24 Now Supports Custom Models

    Please send your details (DAT files and JPEGSs), as well as your description of the problem, along to support (Ken). The more examples he has, the more likely he can sort it out.
  26. J

    Blue Iris Version 6.0.1.24 Now Supports Custom Models

    The responses from Ken: "Thanks for sending that over. The update was not squarely aimed at solving this ... this is something else. Since you are not running periodic scanning and/or not selecting to ignore static objects, othering should be marked as static ... but we see EVERYTHING is...
  27. J

    Blue Iris Version 6.0.1.24 Now Supports Custom Models

    I did send an email to Ken late last night showing another canceled alert with the DAT file and multiple JPEGs of the object motion from the relevant intervals (T+0ms, T+1000ms, T+2000ms, etc...). Hopefully this is useful info for him and fingers crossed that I hear back.
  28. J

    Blue Iris Version 6.0.1.24 Now Supports Custom Models

    Same here...someone needs to send yet another email to Ken telling him it still needs work :(
  29. J

    Built-in AI not detecting (DEBUG ATTACHED)

    Interesting, I installed 6.0.2.2 about an hour ago and have already seen instances where I'm still getting false "occupied" cancelled clips where there is clear and obvious vehicle movement in relevant zones. Ugh...we are still not there yet, even if I thought I had tuned the timings to avoid...