Dahua P5AE-PV

alessi89

n3wb
Feb 4, 2025
8
6
Italy - Rome
Hello guys, I'm facing with the following issue:

I've setup a wireless dahua P5AE-PV with SDcard (to store snapshot) and connected to NVR-5208-ei.
I've setup some IVS rules (disabling MD/SMD)
I've correctly setup the scheduling time to 24/7 (both for snapshot and recording)

All should be ok, in fact sometimes it works - BUT:

1- Some times in DMSS app arrives events with no snapshot (lock icon instead - see pic). When I tap on it, to see the recording, video is not fluent, lags, and missing frames.
2- For the same events in the meantime, in the playback (smartPSS or any other dahua sw) I see "black holes".. see attached video. This is a serious problem because I can say "ok no icon attached.. no problem, just see the recording" but the recording itself is corrupted.

Has anyone had the same problem?
 

Attachments

  • 20250302_111158.mp4
    20250302_111158.mp4
    4.4 MB
  • WhatsApp Image 2025-03-02 at 11.14.37.jpeg
    WhatsApp Image 2025-03-02 at 11.14.37.jpeg
    83.4 KB · Views: 0
First, it appears to be a "do everything" wireless piece of garbage, but beyond that:

It seems the dead space is actually a separate IVS clip (IVS and full time are separate video clips spliced together on Dahua equipment)

There seems to be some problem with the 2nd IVS activation clip being locked (so its not overwritten/lost). I'm not sure how thats done on the camera, its a feature after the fact on the NVR

It would be interesting to see the same video capture of the event at approx 11:38 where the clip BEFORE the shown IVS activation is empty and locked as well
 
  • Like
Reactions: JDreaming
This is the video at 11:38 (approx) - As you can see, person appear in the IVS zone and does not trigger the event.
Look at playbeck when the cursor falls into the orange part.. is not fluent and does not represent significant event.
 

Attachments

Camera is too high to reliably trigger.

The AI was trained based on more of a straight on view and not a top of head view.
 
  • Like
Reactions: JDreaming
Camera is too high to reliably trigger.

The AI was trained based on more of a straight on view and not a top of head view.
I know - this is the reason why I set multiple IVS rules in the left area of the image. Even if it's not perfect I do not understand the strange recording behavior.
 
  • Like
Reactions: JDreaming
Yeah agree with @wittaj that triggering will be hit and miss at that height/angle, but the gaps in recording dont make sense.
The recording gap is almost exactly 50 seconds on both clips....

What does you recording schedule look like?
 
  • Like
Reactions: JDreaming
all schedule set to 24/7
All IVS rules are set to save both snapshot and record
 

Attachments

  • 1741008060954.png
    1741008060954.png
    22.7 KB · Views: 0
  • 1741008083105.png
    1741008083105.png
    22.6 KB · Views: 0
  • 1741008241268.png
    1741008241268.png
    34.5 KB · Views: 0
  • 1741008263640.png
    1741008263640.png
    27 KB · Views: 0
  • Like
Reactions: JDreaming
Are you recording to an NVR as well as events on the camera SD card? Sorry if I missed that
 
  • Like
Reactions: JDreaming
with this configuration I suppose to record all on NVR. And only event's snapshot on SDcard. I'm little bit confused about this but with this configuration, if just one IVS rule is triggered, all the system works as expected (correct playback, correct notification, correct snapshot)
If more than one IVS is triggered simultaneously, camera will fall into a kind of "crazy state" (missing frames, missing trigger...)
 
  • Like
Reactions: bigredfish
Ok that sounds like it just doesn’t have the juice to keep up with multiple triggers in a short period of time.

So if you delete or disable the other rules and just have one it works ok?

Or try disabling/removing the sd card. They can introduce weird behavior sometimes
 
So if you delete or disable the other rules and just have one it works ok?
I've just setup it yesterday and seems to work, only "appear IVS" on the left, and "tripwire IVS"on the right (near the car)
I should perform another test with 2 person violating simultaneously the two IVS rules and check what happen.
 
I guess you could.

I would try disabling the SD card first
 
Hello guys, I'm back after some tests.

I removed the SD card from the camera and although at first everything seemed to work, sometimes the camera drops frames.

1741691200821.png

By the way, in all these "recording holes" you also lose the preset related to the PTZ. The shot you see in this photo is from 5:40 - and six minutes later it has moved

1741691255057.png
Note how the IVS at the top is no longer in the same position.

note:
  • the camera is about 1 meter from the network wifi antenna
  • another different camera is connected to the same network that does not have these recording holes (so no network issue).
  • the PTZ position of the first image is the one I called Preset1 so even if the camera restarts, this should be the target position (when I'm in the wrong position, If I set the Preset1 (from PSS) it goes in the correct position).

I have the impression that this camera has some problem.
 
  • Like
Reactions: bigredfish
The dropping of frames in your first post and this latest one all seem to point to why i dont recommend wireless cameras for critical coverage.
 
Wifi and cameras do not go together.

There are always ways if you don't want to run an ethernet cable.

You need power anyway, so go with a powerline adapter to run the date over your electric lines or use a nano-station.

Maybe you are fine now one day with wifi cams, but one day something will happen. A new device, neighbors microwave, etc.

Cameras connected to Wifi routers (whether wifi or not) are problematic for surveillance cameras because they are always streaming and passing data. And the data demands go up with motion and then you lose signal. A lost packet and it has to resend. It can bring the whole network down if trying to send cameras through a wifi router. At the very least it can slow down your entire system.

Unlike Netflix and other streaming services that buffer a movie, these cameras do not buffer up part of the video, so drop outs are frequent, especially once you start adding distance. You would be amazed how much streaming services buffer - don't believe me, start watching something and unplug your router and watch how much longer you can watch NetFlix before it freezes - mine goes 45 seconds. Now do the same with a camera connected to a router and it is fairly instantaneous (within the latency of the stream itself)...

The same issue applies even with the hard-wired cameras trying to send all this non-buffer video stream through a router. Most consumer grade wifi routers are not designed to pass the constant video stream data of cameras, and since they do not buffer, you get these issues. The consumer routers are just not designed for this kind of traffic, even a GB speed router.

So the more cameras you add, the bigger the potential for issues.

Many people unfortunately think wifi cameras are the answer and they are not. People will say what about Ring and Nest - well that is another whole host of issues that we will not discuss here LOL, but they are not streaming 24/7, only when you pull up the app. And then we see all the people come here after that system failed them because their wifi couldn't keep up when the perp came by. For streaming 24/7 to something like an NVR or Blue Iris, forget about it if you want reliability.


This was a great test that SouthernYankee tried and posted about it here:

I did a WIFI test a while back with multiple 2MP cameras each camera was set to VBR, 15 FPS, 15 Iframe, 3072kbs, h.264. Using a WIFI analyzer I selected the least busy channel (1,6,11) on the 2.4 GHZ band and set up a separate access point. With 3 cameras in direct line of sight of the AP about 25 feet away I was able to maintain a reasonable stable network with only intermittent signal drops from the cameras. Added a 4th camera and the network became totally unstable. Also add a lot of motion to the 3 cameras caused some more network instability. More data more instability.
The cameras are nearly continuously transmitting. So any lost packet causes a retry, which cause more traffic, which causes more lost packets.
WIFI does not have a flow control, or a token to transmit. So your devices transmit any time they want, more devices more collisions.
As a side note, it is very easy to jam a WIFI network. WIFI is fine for watching the bird feed but not for home surveillance and security.
The problem is like standing in a room, with multiple people talking to you at the same time about different subjects. You need to answer each person or they repeat the question.

Test do not guess.

For a 802.11G 2.4 GHZ WIFI network the Theoretical Speed is 54Mbps (6.7MBs) real word speed is nearer to 10-29Mbps (1.25-3.6 MBs) for a single channel


And TonyR recommends this (which is the preferred way IF you want to do wifi)

The only way I'd have wireless cams is the way I have them now: a dedicated 802.11n, 2.4GHz Access Point for 3 cams, nothing else uses that AP. Its assigned channel is at the max separation from another 2.4GHz channel in the house. There is no other house near me for about 300 yards and we're separated by dense foliage and trees.

Those 3 cams are indoor, non-critical pet cams (Amcrest IP2M-841's) streaming to Blue Iris and are adequately reliable for their jobs. They take their turns losing signal/reconnecting usually about every 12 hours or so for about 20 seconds which I would not tolerate for an outdoor surveillance cam pointed at my house and/or property.

But for me, this works in my situation: dedicated AP, non-critical application and periodic, short-term video loss.... if any one of those 3 conditions can't be achieved or tolerated, then I also do not recommend using wireless cams. :cool:
 
  • Like
Reactions: bigredfish