01304 827609 info@use-ip.co.uk Find us

DS-7608nxi-i2/s (c): No more rules are allowed. How many and what types of rules?

hgeorges

Member
Messages
11
Points
3
Hello;
This message seems to be close to my problem - it is worded differently, but I suspect its underlying issue is the same.
My setup consists in the following devices, from different generations, but all on their latest posted firmware:
DS-7608NXI-I2/S NVR
DS-2CD1327G0-L x1
DS-2CD2347G2-L(U) x1
DS-2CD1047G0-L x2
DS-2CD2442FWD-IW x1
In the basic events section, the NVR shows a persistent warning, that there are not enough resources to support motion detection. therefore for all those cameras, I ensured motion detection is disabled.
AI by Device: Enabled for all cameras.

I proceeded in the smart events section, and started setting up intrusion detection first, then line crossing detection.
Going through each camera, drawing contours (intrusion) and lines (line crossing) where I need them to be.
I've only been able to do 3 out of 5 cameras this way. the 3rd is unfinished. When I need to finish the last setting, I get "No more rules are allowed" message. And that's the end of it.
The documentation doesn't talk at all about it, and it is therefore something one would have to figure out through trial and error.
Any thoughts about this issue, how to set this up properly?

Unrelated to the above, another question:
I was trying to find out where to set the length of the video recording after an event occurred. The pre-event is in the right place you'd expect it to be:
Recording schedule - Advanced parameters - Pre-record duration
But there is no post-record there. You'll find Post-Record only in the Capture Schedule - Capture Parameters.
Any thoughts about this one?

Thank you!
 
Unrelated to the above, another question:
I was trying to find out where to set the length of the video recording after an event occurred. The pre-event is in the right place you'd expect it to be:
Recording schedule - Advanced parameters - Pre-record duration
It is in the same place. However it appears to have been renamed (in the web browser only) to "Record Delay"; it's the same parameter though and shows as "Post Record" if you view it on a HDMI connected monitor.
 
Hello;
Coming back to this thread. I'm surprised no one has come across the rules limitation...
I wonder if, given that one of the cameras is acusense, I can set the rules on the camera and not on the NVR - but is not clear if all the alerts and triggers will be passed thru to the NVR for accurate recordings. If that would work, then I could disable AI by device for the respective camera, and keep it only for the older generation cameras, and thus freeing up rules.
Again a question here to this forum, about other's experience of enabling MD and /or BE/SE on camera and how is that working with the NVR. What is the typical approach people take? Optimal combination between camera's features and NVRs features?

By the way, I'm rather pressed for some help, as I'm doing these experiments on a system installed "professionally", to monitor remotely and record only human movements at my parents' home. The physical install, wiring and baseline setup were ok, but clearly no experience with filtering noise and false alerts. So I'm trying to figure the correct setup and avoid messing up something which will require me go there in person. Any help is greatly appreciated!

Phil:
I've set Rre-record 10s and Record delay 1min. The storage schedule is set to record events. Right now the NVR is set "enable AI by device" for all cameras.
I'm baffled by the apparently random behaviour of this system.
DS-2CD2347G2 camera is set with both ID and LC detection, the recording linkage is set as it is also the notification to Surveillance center.
Yet what i see on Hik-Connect mobile app, it is only sending notifications and recordings on ID and nothing for LC detection.
The event recording lasts exactly 10s in the intrusion detection events.

On the other hand of the two DS-2CD1047G0 cameras, set both with ID and no LCD, one is recording 10s of event, and the other a full 1min. I'm not able to understand this variability, since the recording length is set for all in the storage section.
Because of the rules limitation I decided to keep only ID on all and no LCD, and rely on the length of the recording to capture all detail. But it doesn't seem to render reliable results.
Appreciate any advice from the experts!
 
One more thing, unrelated to the above.
I discovered this security bulletin SN No. HSRC-202311-03: Security Vulnerability in Some Hikvision Products implying that pretty much 2CD1xxx and 2CD2xxx older products have some vulnerabilities. So does the NXI NVR.
The article points users to download updated sw from https://www.hikvision.com/en/support/download/firmware/, but of course DS-7608NXI-I2/S and DS-2CD2347G2 were the only ones (being a bit more recent) to be found there with a firmware update.
Is there an official page where Hik-Vision announces its products life cycle policy? When they officially end their support?
For older products, seems that finding the last officially published firmware is an adventure (for me at least). Any advice on this matter?
 
In the basic events section, the NVR shows a persistent warning, that there are not enough resources to support motion detection. therefore for all those cameras, I ensured motion detection is disabled.
AI by Device: Enabled for all cameras.

On an NXI model NVR, when you are using the NVR to provide AcuSense filtering, you can only have Motion Detection OR Smart Event globally. You cannot have the NVR do AcuSense filtering for Motion Detection on one camera and Line Crossing on another. If the NVR engine is doing Motion 2.0 and you set up line crossing on a camera, that line crossing has no AcuSense filtering.

I wonder if, given that one of the cameras is acusense, I can set the rules on the camera and not on the NVR - but is not clear if all the alerts and triggers will be passed thru to the NVR for accurate recordings. If that would work, then I could disable AI by device for the respective camera, and keep it only for the older generation cameras, and thus freeing up rules.
Yes you can (and should) do that. The camera allows you to have Motion Detection and Line Crossing, Intrusion etc all enabled simultaneously with AcuSense filtering, whereas the NVR engine can only do one or the other globally. It's important that before you select AI by camera, that you clear and disable any events for that camera. You then recreate your rule after changing the switch (otherwise the previous rule is still there in the background)

Yet what i see on Hik-Connect mobile app, it is only sending notifications and recordings on ID and nothing for LC detection.
For notifications 'Notify Surveillance Center' needs to be selected. The setup and that selection is done via the NVR menu NOT the camera (the camera cannot send an event message to Hik-Connect as it has no path to the internet)
 
On an NXI model NVR, when you are using the NVR to provide AcuSense filtering, you can only have Motion Detection OR Smart Event globally. You cannot have the NVR do AcuSense filtering for Motion Detection on one camera and Line Crossing on another. If the NVR engine is doing Motion 2.0 and you set up line crossing on a camera, that line crossing has no AcuSense filtering
Ok. That I understood. So I didn't enable /choose motion detection on the NVR. In VCA section I've chosen Perimeter protection.
Nor on the Acusense camera - I've chosen instead to work with Smart Events (SE) only.
I'm not even sure if (having the option to set up MD on the camera) I should enable it, since I want to use SE and avoid recording and notifications on false alerts. What is the typical setup for outdoors monitoring? Is MD of any use when I can use SE?

Yes you can (and should) do that. The camera allows you to have Motion Detection and Line Crossing, Intrusion etc all enabled simultaneously with AcuSense filtering, whereas the NVR engine can only do one or the other globally. It's important that before you select AI by camera, that you clear and disable any events for that camera. You then recreate your rule after changing the switch (otherwise the previous rule is still there in the background)
This one I thought I understood, but perhaps not. Turning off AI by device is in essence turning it on by camera (toggle switch). I did that.
Went to ID and LCD and disabled them too (on the NVR).
And then I went to the camera configuration SE menu to configure ID and LCD definitions on it. But I'm not sure if this is correct, I think this was possibly not what you meant.
The other possibility would be to turn on AI for camera, and do the SE configuration still on the NVR, presuming that those settings will pass through to the camera.
What is the correct answer?

For notifications 'Notify Surveillance Center' needs to be selected. The setup and that selection is done via the NVR menu NOT the camera (the camera cannot send an event message to Hik-Connect as it has no path to the internet)
I understand this one too, that's why I think I must understand earlier point correctly.

Thank you
 
I'm rethinking the use of ID and LCD, and instead use region entrance /exit detection (RED).
My use case is monitoring and recording the surroundings of a house. Small front and backyard. The house sides are sandwiched between the neighbors properties and fences run around the sides, front and back, starting where the building ends. That's why I drew a lot of lines... assuming all possible directions of access/intrusion. And that's where the NVR complained about too many rules, and stopped taking them at 3rd camera.
It may be that RED might be more useful. Comments? Any issues I should be aware of?
 
There's some confusion as to which way around you set the 'Enable AI by Device' switch. From memory, with the switch on 'Device' is referring to the camera not the NVR. You can confirm which way around it is by logging into the NVR from a web browser rather than using the local on screen menu; the browser refers to 'AI by camera'/'AI by NVR' I believe as opposed to 'device' so you know you have it correctly set. It's obviously massively important that you get it the right way around from the get go. As I mentioned earlier, if you set it events of any sort using the NVR, then switch to the camera, unless deleted and turned off the previous setup is still there on the NVR.

The other possibility would be to turn on AI for camera, and do the SE configuration still on the NVR, presuming that those settings will pass through to the camera.
Yes they do. Always do the settings on the NVR.

In VCA section I've chosen Perimeter protection.
Nor on the Acusense camera - I've chosen instead to work with Smart Events (SE) only.
I'm not even sure if (having the option to set up MD on the camera) I should enable it, since I want to use SE and avoid recording and notifications on false alerts. What is the typical setup for outdoors monitoring? Is MD of any use when I can use SE?
Nothing wrong with motion detection and in many cases it's perfectly adequate. When you're covering a large area, there's a large difference in perspective between a human that's close to the camera (occupying 40% or more of screen height perhaps) and at a distance (occupying a little as 5% of screen height perhaps) Maybe that's not ideal, as if the camera is basing it's decision of what's 'human' based on a simple shape and movement of those grouped pixels, there may be little difference between a human thats distant to the camera and a bird that's stood close to it, with regard to pixel density, shape and movement (I don't profess to have any knowledge of the workings of the AcuSense algorithm, so I'm just surmising). I never leave motion sensitivity at the default of 80. All of my own cameras are set at 20 and detect reliably with few false alerts.

Smart events (Line Crossing, Intrusion detection) have more adjustments possible. Once the basics have been set via the NVR (area, direction, sensitivity), finer adjustments can be added in the camera. Minimum and maximum target size specifically can be adjusted in relation to a human target in the position of the detection area, so it's easier to prevent alerts from birds, cats, foxes) Target validity allows further adjustment.

I don't think there's any 'typical' setup as every view is different. Line Crossing works best when the target crosses a line drawn vertically. That may not be ideal depending on the mounting location of the camera. Similarly intrusion detection won't work best full screen. For both line crossing and intrusion, the target ideally needs to be in view of the camera before crossing the line or entering the detection zones in order to process it's size in relation to the minimum and maximums set. I tend to set up Motion Detection without 'Notify Surveillance Center', combined with Line Crossing or Intrusion with 'Notify Surveillance Center'. A larger area outside of the property is covered with Motion Detection, the video is marked so that selecting 'Human' in the app finds anyone in the vicinity. Anyone crossing the line or entering the intrusion zone triggers a notification.

It may be that RED might be more useful. Comments? Any issues I should be aware of?
I wouldn't use Region entrance/exit. For reasons only known to Hikvision, Region Entrance will not trigger a Hik-Connect notification even when the 'Notify Surveillance Center' option is checked; it just doesn't work. However Intrusion Detection with a threshold set to zero seconds IS Region Entrance; there's no difference at all (confirmed by Hikvision)
 
There's some confusion as to which way around you set the 'Enable AI by Device' switch.
Yes they do. Always do the settings on the NVR.
this is clear now.
Nothing wrong with motion detection and in many cases it's perfectly adequate. When you're covering a large area, there's a large difference in perspective between a human that's close to the camera (occupying 40% or more of screen height perhaps) and at a distance (occupying a little as 5% of screen height perhaps) Maybe that's not ideal, as if the camera is basing it's decision of what's 'human' based on a simple shape and movement of those grouped pixels, there may be little difference between a human thats distant to the camera and a bird that's stood close to it, with regard to pixel density, shape and movement (I don't profess to have any knowledge of the workings of the AcuSense algorithm, so I'm just surmising). I never leave motion sensitivity at the default of 80. All of my own cameras are set at 20 and detect reliably with few false alerts.
Thanks, I'll revisit VCA setting. I had it set on Perimeter.
But then this issue gets in the way:
On an NXI model NVR, when you are using the NVR to provide AcuSense filtering, you can only have Motion Detection OR Smart Event globally. You cannot have the NVR do AcuSense filtering for Motion Detection on one camera and Line Crossing on another. If the NVR engine is doing Motion 2.0 and you set up line crossing on a camera, that line crossing has no AcuSense filtering
MD enabled at NVR level, takes away the ability to set SE.

Smart events (Line Crossing, Intrusion detection) have more adjustments possible. Once the basics have been set via the NVR (area, direction, sensitivity), finer adjustments can be added in the camera. Minimum and maximum target size specifically can be adjusted in relation to a human target in the position of the detection area, so it's easier to prevent alerts from birds, cats, foxes) Target validity allows further adjustment.
I kind of got the idea and set them. I'm surely lacking in the finer adjustments. Tedious process.
Unpredictable events and recordings - strange. I verified twice all the settings.

Line Crossing works best when the target crosses a line drawn vertically. That may not be ideal depending on the mounting location of the camera. Similarly intrusion detection won't work best full screen. For both line crossing and intrusion, the target ideally needs to be in view of the camera before crossing the line or entering the detection zones in order to process it's size in relation to the minimum and maximums set. I tend to set up Motion Detection without 'Notify Surveillance Center', combined with Line Crossing or Intrusion with 'Notify Surveillance Center'. A larger area outside of the property is covered with Motion Detection, the video is marked so that selecting 'Human' in the app finds anyone in the vicinity. Anyone crossing the line or entering the intrusion zone triggers a notification.
Re LCD best action.
Since I'm trying to detect intrusion and /or motion from many sides, my choice was to draw lines all over. Since 4 out of 5 cameras are not Acusense, I've chosen Perimeter in VCA setting. Switching to MD takes away SE from the menu.

I wouldn't use Region entrance/exit. For reasons only known to Hikvision, Region Entrance will not trigger a Hik-Connect notification even when the 'Notify Surveillance Center' option is checked; it just doesn't work. However Intrusion Detection with a threshold set to zero seconds IS Region Entrance; there's no difference at all (confirmed by Hikvision)
Got it. Thanks, I won't fiddle with it anymore. Thank you, JB.
I've re-read everything and tried a few things along the way.
I've switched now to MD VCA setting and I'll leave it a few days, to see how it performs.
Having mostly older /no acusense cameras complicates the matter. it's either MD or SE. SE is more granular, but the limitations makes it tedious to set up. I could certainly use an hour of consulting :) Let me know if this is an option.

Thanks again
 
Here I come again. I messed things up, simplifying the setup, I get no notifications and no recordings on all but Acusense camera. Even that is not what it needs to be... I'd really benefit from a 1-1 assistance, as this takes too much time I don't have. Any possibility to get an expert on a 1h call with me?

Current state:

Changed VCA from Perimeter to Motion.

Changed the four non-Acusense cameras to Basic Events.
Enabled all options: Enable Motion Detection, Enable AI Configuration and Enable Dynamic Analysis for Motion.
Checked Detection target: Human, Sensibility 40
Arming schedule 24x7
Link method: Notification to surveillance center and recording enabled.
In the storage section. I chose initially to record events - nothing was recorded. Then selected motion - nothing was recorded either.
In advanced settings: pre-record 30s, recording delay (which was supposed to mean post-trigger recording duration) 1min.

For Acusense camera, I've set both Basic and Smart events ( intrusion detection.)
It works somewhat, but not as I need it to. It is not
In Basic events:
Enable Motion Detection: Enabled, Enable AI Configuration: Disabled, Enable Dynamic Analysis for Motion: Enabled
Checked Detection target: Human, Sensibility 40
Arming schedule 24x7
Link method: Notification to surveillance center and recording enabled.
In Smart Events:
Enable AI by Device: Disabled (it is done by the camera)
Enable Intrusion Detection: Enabled
Detection target: human
Threshold: 0
Sensitivity: 50
Arming schedule 24x7
Link method: Notification to surveillance center and recording enabled.
Storage record: Events 24x7
Advanced settings: same as above.

This channel records activity, but only late (the person is already in the middle of the scene), missing the beginning of the action. Presumably the threshold is not sufficient, and I have to change sensitivity to 0 too, if I want it to record from the moment if touching the set perimeter. I haven't tried that.
The duration of the recording is 10s - the 1min set is ignored.

Notifications and events late in hikconnect mobile app - probably hour(s). lots of the outdoors activity missing. So totally random.
...
Pls help. Thank you
 
I think unfortunately you've reached a point where you've tried so many things and altered so many settings that it's nigh on impossible to provide instructions on how to put it right here.

I've never set an NVR up to record on event/motion only. It's better to record 24/7 and use events for notifications and/or easier searching of recorded footage as I mentioned in post #8.

I'm not actually convinced you're interpreting the 'Device' in 'Enable AI by Device' correctly. I don't have an NXI-I model to check but from memory on the NXI-K models 'Device' refers to the camera. So with the 'Enable AI by Device' switched on/enabled, the camera (not the NVR) is doing the processing. You seem to indicate the opposite with the below statement:

Enable AI by Device: Disabled (it is done by the camera)

Unfortunately Hikvision confuse the matter, as the terminology/wording differs according to whether you are viewing the configuration menus on the locally connected monitor, or via the web browser menu (I'm not sure whether this has improved on more recent firmware but previously the local menu uses 'Device' and the browser menu uses 'Camera'/'NVR' which is clearer). Once events have been set up with that switch the wrong way around, it can be awkward to correct as toggling the switch to the other position for a channel does not delete the previous event setup on the camera/NVR. The only way to be sure is to clear and disable all events for all cameras with the 'Enable AI by Device' in one position, then repeat with it in the opposite position. When complete there's a menu in the NVR where you can check the NVR event engine. That should show that it is not in use. Only then are you back to square one where you can begin to set it up again.
 
I think unfortunately you've reached a point where you've tried so many things and altered so many settings that it's nigh on impossible to provide instructions on how to put it right here.
That's quite a true statement :)
It's better to record 24/7 and use events for notifications and/or easier searching of recorded footage as I mentioned in post #8.
I actually had it working on events recording but I tried to resolve the fact that I was running out of rules. Your approach is ideal, but I'll have to beef up the storage and I'm remote now. I'd have to see how many full days I can fit in the 10TB.

I'm not actually convinced you're interpreting the 'Device' in 'Enable AI by Device' correctly.
Actually through trial and error I determined that is what it is. Device is the NVR (at least in this version of the firmware :) - I understand Hikvision keeps everyone busy trying to decipher their changes from one version to the next)
I don't have an NXI-I model to check but from memory on the NXI-K models 'Device' refers to the camera. So with the 'Enable AI by Device' switched on/enabled, the camera (not the NVR) is doing the processing. You seem to indicate the opposite with the below statement:
I just did a full search on the forum and found this fellow's statement which appear to confirm my finding
Unfortunately Hikvision confuse the matter, as the terminology/wording differs according to whether you are viewing the configuration menus on the locally connected monitor, or via the web browser menu (I'm not sure whether this has improved on more recent firmware but previously the local menu uses 'Device' and the browser menu uses 'Camera'/'NVR' which is clearer). Once events have been set up with that switch the wrong way around, it can be awkward to correct as toggling the switch to the other position for a channel does not delete the previous event setup on the camera/NVR. The only way to be sure is to clear and disable all events for all cameras with the 'Enable AI by Device' in one position, then repeat with it in the opposite position.
I sort of did this going back and forth between one of my non-acusense cameras and the only acusense camera I have and compared the NVRs messages each time.. "Enable AI by Device" on implies the NVR does the job.
Only then are you back to square one where you can begin to set it up again.
Would you consider teaming up with me for a consultation? Thank you, JB.
 
Alright, looks like my thread is not getting any further ideas... I'd really appreciate a bit of a direction.

Current state:
I have set continuous recording on all 5 channels (I posted above my configuration)
I have set the System > vca config > Engine config on motion detection, instead of perimeter protection
In the Event setup
- all non-acusense cameras (4) are set on Basic Event handling with all options enabled (Enable Motion Detection, Enable AI Configuration, Enable Dynamic Analysis for Motion). Sensitivity 80. Human detection. Arming schedule 24x7. Linkage method: Notify and record.
- acusense camera
in Basic Events: Enable Motion Detection: On. Enable AI Configuration: Off. Enable Dynamic Analysis for Motion: On
in Smart Events: Enable AI by Device: Off.
Scene Change Detection: On, Sensitivity: 50
Intrusion detection: On, Sensitivity 50, Human detection
In storage all recording continuous.

Currently only the acusense camera channel is behaving as expected after this setup.
The recording is continuous, and the intrusion detection is notifying and marked on the recording playback timeline.
The non-acusense camera channels are recording continously. No notifications what so ever, no marking of events on the timeline, just long recordings to watch endlesly.

Where is my error, please help me, I really need some suggestions. Thank you!!
 
Hello;
This message seems to be close to my problem - it is worded differently, but I suspect its underlying issue is the same.
My setup consists in the following devices, from different generations, but all on their latest posted firmware:
DS-7608NXI-I2/S NVR
DS-2CD1327G0-L x1
DS-2CD2347G2-L(U) x1
DS-2CD1047G0-L x2
DS-2CD2442FWD-IW x1
In the basic events section, the NVR shows a persistent warning, that there are not enough resources to support motion detection. therefore for all those cameras, I ensured motion detection is disabled.
AI by Device: Enabled for all cameras.

I proceeded in the smart events section, and started setting up intrusion detection first, then line crossing detection.
Going through each camera, drawing contours (intrusion) and lines (line crossing) where I need them to be.
I've only been able to do 3 out of 5 cameras this way. the 3rd is unfinished. When I need to finish the last setting, I get "No more rules are allowed" message. And that's the end of it.
The documentation doesn't talk at all about it, and it is therefore something one would have to figure out through trial and error.
Any thoughts about this issue, how to set this up properly?

Unrelated to the above, another question:
I was trying to find out where to set the length of the video recording after an event occurred. The pre-event is in the right place you'd expect it to be:
Recording schedule - Advanced parameters - Pre-record duration
But there is no post-record there. You'll find Post-Record only in the Capture Schedule - Capture Parameters.
Any thoughts about this one?

Thank you!
Q.1: If your cameras support Accusense, then you should disable "AI by Device" on the NVR and let the Camera handle it via its onboard Accusense.
Q.2: Under Storage, you should be setting "Continuous Recording" so it's recording 24x7.
 
Back
Top