Review - Dahua SD49225XA-HNR 2MP 25x Starlight + IR PTZ AI Camera with Deep IVS & SMD Plus

2 years later and how does this camera work out? Still worth getting?
 
Mine died, lost power and when the power came back on all my cams came back except this one. Its been to cold and to much snow on the ground for me to put a ladder and look at it.
 
I had high hopes for this but didn't work for me. Auto tracking was actually decent but night image wasn't that good for 2MP. I think I will have to swallow the cost and go with 4MP with bigger sensor.
 
I had high hopes for this but didn't work for me. Auto tracking was actually decent but night image wasn't that good for 2MP. I think I will have to swallow the cost and go with 4MP with bigger sensor.

What about it was the issue? Were you forcing it in color or B/W with infrared?

From the other thread, I am thinking you have only had it for a few days? Give it some time to dial in the settings. These are not plug-n-play and takes some time to dial it in for the areas you are wanting to cover.

Any videos you can share for illustration purposes?
 
Last edited:
What about it was the issue? Were you forcing it in color or B/W with infrared?

From the other thread, I am thinking you have only had it for a few days? Give it some time to dial in the settings. These are not plug-n-play and takes some time to dial it in for the areas you are wanting to cover.

Any videos you can share for illustration purposes?
Yeah I didn't have long but it was enough to determine that I want a better night image for my coverage. I tried color and B&W neither was satisfactory for me. It's not a bad camera just didn't work for my night imaging needs.
 
Folks, a couple of questions:

1) I am having issues with the camera triggering when a moth or bug flies by at night and it's lit up by the IR. This results in a lot of spurious alerts. This is even with SMD turned on and the object filter set, and sensitivity set to low in SMD.

If I turn off motion detection and only use IVS, the camera seems to miss a lot where cars moving at a distance or other events are not captured.

Any tips on reducing these false

2) As part of debugging 1), I notice that I can't seem to turn off the object filter in the IVS menu. Is that normal? I was thinking if I turned off objection detection but motion detection off, that would improve my false positive situation, but that doesn't seem to be an option.

Thx
mike
 
Finally I have some good weather and took my cam down and I think its dead, water dead. I thought I thought of everything but I missed one spot and water got in. Water ran down the conduit, down the cable and into the plug. It was a great cam, need to get another one.
 

Attachments

  • 20220212_144821_LI.jpg
    20220212_144821_LI.jpg
    2.4 MB · Views: 61
  • 20220212_135916.jpg
    20220212_135916.jpg
    332.9 KB · Views: 61
  • 20220212_144554.jpg
    20220212_144554.jpg
    411.9 KB · Views: 57
I would ALWAYS have the lan cable enter from the bottom via a 25mm cable gland with a loop which makes it impossible for water integration. That conduit has zero IP rating.
 
  • Like
Reactions: sebastiantombs
There are compression connectors for EMT. A set screw connector, outside, is a definite NO-NO.
 
  • Like
Reactions: Smilingreen
Folks, a couple of questions:

1) I am having issues with the camera triggering when a moth or bug flies by at night and it's lit up by the IR. This results in a lot of spurious alerts. This is even with SMD turned on and the object filter set, and sensitivity set to low in SMD.

If I turn off motion detection and only use IVS, the camera seems to miss a lot where cars moving at a distance or other events are not captured.

Any tips on reducing these false

2) As part of debugging 1), I notice that I can't seem to turn off the object filter in the IVS menu. Is that normal? I was thinking if I turned off objection detection but motion detection off, that would improve my false positive situation, but that doesn't seem to be an option.

Thx
mike


Guys, any ideas here? With motion detection off, the IVS triggers very infrequently, even on people and cars (esp at a distance).
 
Guys, any ideas here? With motion detection off, the IVS triggers very infrequently, even on people and cars (esp at a distance).

We need to see field of view screenshots and screenshots of your settings.

Lot's of reasons why it can miss, but most are related to field of view and initial settings.
 
We need to see field of view screenshots and screenshots of your settings.

Lot's of reasons why it can miss, but most are related to field of view and initial settings.


Ok! Here are my settings (and you can see the filed of view in the images I think):


Setting1.png

Setting2.png

Setting3.png
Setting6.pngSetting5.png

For some reason, when I access the settings through chrome, it resets the track time 15 secs. But in IE I can reset it to the normal 240 secs I use.

The view you are seeing in screen caps is the default PTZ location. The street is a dead end on the other side, so this is the view that gets the most action.

Thanks!
 
Yes, that is normal with this camera - it is sensitive to using Explorer, especially for track time.

So the potential issue with that field of view is it is great for a fixed cam, but the subject is too close to the PTZ and the PTZ cannot respond fast enough to keep tracking it as it approaches the PTZ. Going away from the camera should be fine.

What zoom are you on - is that wide open or zoomed in a bit? If you have room to zoom out I would do that to give the camera more time to react and respond.

The ability for the IVS to pick up the top left motion will be problematic. IVS needs to see the object and identify it and determine if it is something that needs to be tracked. So it will not catching anything going left-right type motion and for objects coming towards the camera, it will take awhile in many instances before it identifies it to track it.

What are your other settings like brightness, exposure, etc. - are they on default or have you dialed them in to your setting and field of view? That can make a difference as well.
 
Yes, that is normal with this camera - it is sensitive to using Explorer, especially for track time.

So the potential issue with that field of view is it is great for a fixed cam, but the subject is too close to the PTZ and the PTZ cannot respond fast enough to keep tracking it as it approaches the PTZ. Going away from the camera should be fine.

What zoom are you on - is that wide open or zoomed in a bit? If you have room to zoom out I would do that to give the camera more time to react and respond.

The ability for the IVS to pick up the top left motion will be problematic. IVS needs to see the object and identify it and determine if it is something that needs to be tracked. So it will not catching anything going left-right type motion and for objects coming towards the camera, it will take awhile in many instances before it identifies it to track it.

What are your other settings like brightness, exposure, etc. - are they on default or have you dialed them in to your setting and field of view? That can make a difference as well.


Thanks for the feedback. There are limited locations available for mounting this camera, and I do have fixed cameras that look at other positions as well, albeit from the house and not on the other side of the fence, where this camera is located. I use the other cameras as spotter to direct the PTZ camera around to ameliorate the lack of the camera being able to track a vehicle constantly by itself. This camera is amounted about 15 ft on a tree that is close to the street. There is another tree further down, closer to the dead end that I could use, but is more complicated to get the cable to. That would see the entire street and may be better in terms of tracking vehicles etc...

The camera is wide open normally in this position. With the motion detection on, it's actually very good at picking up people and cars at a distance, and then locking on them, and even better when people or cars come from the other side and are moving away The day settings for exposure etc.. are close toe defaults, and they have worked pretty well. I have done a lot more tweaking of the nightime settings, and I am happy to send them, but if they don't work well in the daytime, I don't think the night will be any better.

Perhaps what I need to do is find a way for the motion detection to not react so much to trees and bushes swaying? What are my options for that?

Thanks very much for the feedback!!
 
Yeah I thought that field of view looked familiar, so I went back on this thread and saw that I have offered suggestions in the past.

It is a great location, but can be problematic as well, so the spotter cam can help get around the limitation of the great location.

I assume you are talking about the motion detection from within the camera and not BI? If so, you can go into each one of those and start eliminating the "blocks" that clearly would be trees and bushes moving, and just leave the motion blocks be just about the intrusion box.

That should eliminate over 95% of those problems. Then adjust the sensitivity if needed.
 
Yeah I thought that field of view looked familiar, so I went back on this thread and saw that I have offered suggestions in the past.

It is a great location, but can be problematic as well, so the spotter cam can help get around the limitation of the great location.

I assume you are talking about the motion detection from within the camera and not BI? If so, you can go into each one of those and start eliminating the "blocks" that clearly would be trees and bushes moving, and just leave the motion blocks be just about the intrusion box.

That should eliminate over 95% of those problems. Then adjust the sensitivity if needed.


Actually, I was talking about motion detection the in camera itself, video motion detection and smart motion detection, and I am using the ONVIF event detection for BI to pick it up. I need to see if the camera is actually locking on objects and tracking them but not generating an alert, or without motion detection on, the camera never tries to track the object in the first place. I suspect the latter, but can do a test to verify.